Significance Agricultural sensor is the key technology for developing modern agriculture. Agricultural sensor is a kind of detection device that can sense and convert physical signal, which is related to the agricultural environment, plants and animals, into an electrical signal. Agricultural sensors could be applied to monitor crops and livestock in different agricultural environments, including weather, water, atmosphere and soil. It is also an important driving force to promote the iterative upgrading of agricultural technology and change agricultural production methods. Progress The different agricultural sensors are categorized, the cutting-edge research trends of agricultural sensors are analyzed, and summarizes the current research status of agricultural sensors are summarized in different application scenarios. Moreover, a deep analysis and discussion of four major categories is conducted, which include agricultural environment sensors, animal and plant life information sensors, agricultural product quality and safety sensors, and agricultural machinery sensors. The process of research, development, the universality and limitations of the application of the four types of agricultural sensors are summarized. Agricultural environment sensors are mainly used for real-time monitoring of key parameters in agricultural production environments, such as the quality of water, gas, and soil. The soil sensors provide data support for precision irrigation, rational fertilization, and soil management by monitoring indicators such as soil humidity, pH, temperature, nutrients, microorganisms, pests and diseases, heavy metals and agricultural pollution, etc. Monitoring of dissolved oxygen, pH, nitrate content, and organophosphorus pesticides in irrigation and aquaculture water through water sensors ensures the rational use of water resources and water quality safety. The gas sensor monitors the atmospheric CO2, NH3, C2H2, CH4 concentration, and other information, which provides the appropriate environmental conditions for the growth of crops in greenhouses. The animal life information sensor can obtain the animal's growth, movement, physiological and biochemical status, which include movement trajectory, food intake, heart rate, body temperature, blood pressure, blood glucose, etc. The plant life information sensors monitor the plant's health and growth, such as volatile organic compounds of the leaves, surface temperature and humidity, phytohormones, and other parameters. Especially, the flexible wearable plant sensors provide a new way to measure plant physiological characteristics accurately and monitor the water status and physiological activities of plants non-destructively and continuously. These sensors are mainly used to detect various indicators in agricultural products, such as temperature and humidity, freshness, nutrients, and potentially hazardous substances (e.g., bacteria, pesticide residues, heavy metals, etc. Agricultural machinery sensors can achieve real-time monitoring and controlling of agricultural machinery to achieve real-time cultivation, planting, management, and harvesting, automated operation of agricultural machinery, and accurate application of pesticide, fertilizer. [Conclusions and Prospects In the challenges and prospects of agricultural sensors, the core bottlenecks of large-scale application of agricultural sensors at the present stage are analyzed in detail. These include low-cost, specialization, high stability, and adaptive intelligence of agricultural sensors. Furthermore, the concept of "ubiquitous sensing in agriculture" is proposed, which provides ideas and references for the research and development of agricultural sensor technology.
Objective The content of nitrogen (N) and potassium (K) in the soil directly affects crop yield, making it a crucial indicator in agricultural production processes. Insufficient levels of the two nutrients can impede crop growth and reduce yield, while excessive levels can result in environmental pollution. Rapidly quantifying the N and K content in soil is of great importance for agricultural production and environmental protection. Methods A rapid and quantitative method was proposed for detecting N and K nutrient ions in soil based on polydimethylsiloxane (PDMS) microfluidic chip electrophoresis and capacitively coupled contactless conductivity detection (C4D). Microfluidic chip electrophoresis enables rapid separation of multiple ions in soil. The electrophoresis microfluidic chips have a cross-shaped channel layout and were fabricated using soft lithography technology. The sample was introduced into the microfluidic chip by applying the appropriate injection voltage at both ends of the injection channel. This simple and efficient procedure ensured an accurate sample introduction. Subsequently, an electrophoretic voltage was applied at both ends of the separation channel, creating a capillary zone electrophoresis that enables the rapid separation of different ions. This process offered high separation efficiency, required a short processing time, and had a small sample volume requirement. This enabled the rapid processing and analysis of many samples. C4D enabled precise measurement of changes in conductivity. The sensing electrodes were separated from the microfluidic chips and printed onto a printed circuit board (PCB) using an immersion gold process. The ions separated under the action of an electric field and sequentially reach the sensing electrodes. The detection circuit, connected to the sensing electrodes, received and regulated the conductivity signal to reflect the variance in conductivity between the sample and the buffer solution. The sensing electrodes were isolated from the sample solution to prevent interference from the high-voltage electric field used for electrophoresis. Results and Discussions The voltage used for electrophoresis, as well as the operating frequency and excitation voltage of the excitation signal in the detection system, had a significant effect on separation and detection performance. Based on the response characteristics of the system output, the optimal operating frequency of 1 000 kHz, excitation voltage of 50 V, and electrophoresis voltage of 1.5 kV were determined. A peak overshoot was observed in the electrophoresis spectrum, which was associated with the operating frequency of the system. The total noise level of the system was approximately 0.091 mV. The detection limit (S/N = 3) for soil nutrient ions was determined by analyzing a series of standard sample solutions with varying concentrations. The detection limited for potassium (K+), ammonium (NH4+), and nitrate (NO3‒) standard solutions were 0.5, 0.1 and 0.4 mg/L, respectively. For the quantitative determination of soil nutrient ion concentration, the linear relationship between peak area and corresponding concentration was investigated under optimal experimental conditions. K+, NH4+, and NO3‒ exhibit a strong linear relationship in the range of 0.5~40 mg/L, with linear correlation coefficients (R2) of 0.994, 0.997, and 0.990, respectively, indicating that this method could accurately quantify N and K ions in soil. At the same time, to evaluate the repeatability of the system, peak height, peak area, and peak time were used as evaluation indicators in repeatability experiments. The relative standard deviation (RSD) was less than 4.4%, indicating that the method shows good repeatability. In addition, to assess the ability of the C4D microfluidic system to detect actual soil samples, four collected soil samples were tested using MES/His and PVP/PTAE as running buffers. K+, NH4+,Na+, Chloride (Cl‒), NO3‒, and sulfate (SO43‒) were separated sequentially within 1 min. The detection efficiency was significantly improved. To evaluate the accuracy of this method, spiked recovery experiments were performed on four soil samples. The recovery rates ranged from 81.74% to 127.76%, indicating the good accuracy of the method. Conclusions This study provides a simple and effective method for the rapid detection of N and K nutrient ions in soil. The method is highly accurate and reliable, and it can quickly and efficiently detect the contents of N and K nutrient ions in soil. This contactless measurement method reduced costs and improved economic efficiency while extending the service life of the sensing electrodes and reducing the frequency of maintenance and replacement. It provided strong support for long-term, continuous conductivity monitoring.
Objective The soil moisture content is a crucial factor that directly affected the growth and yield of crops. By using a soil measurement instrument to measure the soil's moisture content, lots of powerful data support for the development of agriculture can be provided. Furthermore, these data have guiding significance for the implementation of scientific irrigation and water-saving irrigation in farmland. In order to develop a reliable and efficient soil moisture sensor, a new capacitive soil moisture sensor based on microfabrication technology was proposed in this study. Capacitive moisture sensors have the advantages of low power consumption, good performance, long-term stability, and easy industrialization. Method The forked electrode array consists of multiple capacitors connected in parallel on the same plane. The ideal design parameters of 10 μm spacing and 75 pairs of forked electrodes were obtained by calculating the design of forked finger logarithms, forked finger spacing, forked finger width, forked finger length, and electrode thickness, and studying the influence of electrode parameters on capacitance sensitivity using COMSOL Multiphysics software. The size obtained an initial capacitance on the order of picofarads, and was not easily breakdown or failed. The sensor was constructed using microelectromechanical systems (MEMS) technology, where a 30 nm titanium adhesion layer was sputtered onto a glass substrate, followed by sputtering a 100 nm gold electrode to form a symmetrical structure of forked electrodes. Due to the strong adsorption capacity of water molecules of the MoS2 (molybdenum disulfide) layer, it exhibited high sensitivity to soil moisture and demonstrated excellent soil moisture sensing performance. The molybdenum disulfide was coated onto the completed electrodes as the humidity-sensitive material to create a humidity sensing layer. When the humidity changed, the dielectric constant of the electrode varied due to the moisture-absorbing characteristics of molybdenum disulfide, and the capacitance value of the device changed accordingly, thus enabling the measurement of soil moisture. Subsequently, the electrode was encapsulated with a polytetrafluoroethylene (PTFE) polymer film. The electrode encapsulated with the microporous film could be directly placed in the soil, which avoided direct contact between the soil/sand particles and the molybdenum disulfide on the device and allowed the humidity sensing unit to only capture the moisture in the soil for measuring humidity. This ensured the device's sensitivity to water moisture and improved its long-term stability. The method greatly reduced the size of the sensor, making it an ideal choice for on-site dynamic monitoring of soil moisture. Results and Discussions The surface morphology of molybdenum disulfide was characterized and analyzed using a Scanning Electron Microscope (SEM). It was observed that molybdenum disulfide nanomaterial exhibited a sheet-like two-dimensional structure, with smooth surfaces on the nanosheets. Some nanosheets displayed sharp edges or irregular shapes along the edges, and they were irregularly arranged with numerous gaps in between. The capacitive soil moisture sensor, which utilized molybdenum disulfide as the humidity-sensitive layer, exhibited excellent performance under varying levels of environmental humidity and soil moisture. At room temperature, a humidity generator was constructed using saturated salt solutions. Saturated solutions of lithium chloride, potassium acetate, magnesium chloride, copper chloride, sodium chloride, potassium chloride, and potassium sulfate were used to generate relative humidity levels of 11%, 23%, 33%, 66%, 75%, 84%, and 96%, respectively. The capacitance values of the sensor were measured at different humidity levels using an LCR meter (Agilent E4980A). The capacitance output of the sensor at a frequency of 200 Hz ranged from 12.13 pF to 187.42 nF as the relative humidity varied between 11% to 96%. The sensor exhibited high sensitivity and a wide humidity sensing range. Additionally, the frequency of the input voltage signal had a significant impact on the capacitance output of the sensor. As the testing frequency increased, the response of the sensor's system decreased. The humidity sensing performance of the sensor was tested in soil samples with moisture content of 8.66%, 13.91%, 22.02%, 31.11%, and 42.75%, respectively. As the moisture content in the soil increased from 8.66% to 42.75%, the capacitance output of the sensor at a frequency of 200 Hz increased from 119.51 nF to 377.98 nF, demonstrating a relatively high sensitivity. Similarly, as the frequency of the input voltage increased, the capacitance output of the sensor decreased. Additionally, the electrode exhibited good repeatability and the sensitivity of the sensor increased significantly as the testing frequency decreased. Conclusions The capacitive soil moisture sensor holds promise for effective and accurate monitoring of soil moisture levels, with its excellent performance, sensitivity, repeatability, and responsiveness to changes in humidity and soil moisture. The ultimate goal of this study is to achieve long-term monitoring of capacitance changes in capacitive soil moisture sensors, enabling monitoring of long-term changes in soil moisture. This will enable farmers to optimize irrigation systems, improve crop yields, and reduce water usage. In conclusion, the development of this innovative soil moisture sensor has the potential to promote agricultural modernization by providing accurate and reliable monitoring of soil moisture levels.
Objective Capacitive coupled contactless conductivity detection (C4D) plays an important role in agricultural soil nutrient ion detection. Effective identification of characteristic ion peaks in C4D signals is conducive to subsequent qualitative and quantitative analysis of characteristic ion peaks, which provides a basis for improving agricultural soil nutrient management. However, the detection of characteristic peaks in C4D signals still has shortcomings, such as the inability of automatic and accurate identification, complicated manual operation, and low efficiency. Methods In this study, an automatic spectral peak identification algorithm based on continuous wavelet transform combined with particle swarm optimization (PSO) and maximum interclass variance method (Otsu) was proposed to achieve accurate, efficient and automated identification of C4D signal peaks. Capillary electrophoresis (CE) combined with a C4D device (CE-C4D) was used to detect the standard ions and soil sample solutions to obtain the C4D ion signal spectra, which were simulated according to the characteristics of the real C4D signal spectra to obtain the C4D simulated signals containing single Gaussian peaks and multi-Gaussian peaks. The continuous wavelet transform was performed on the C4D spectrogram signal to obtain the wavelet transform coefficient matrix. The local maxima and local minima of the continuous wavelet transform coefficient matrix were searched by the staircase scanning method, and the local maxima and local minima were connected to form ridges and valleys. The wavelet coefficient matrix was converted to a gray-scale image by logistic mapping to visualize the data. The number of particle populations in PSO was set to 15, the gray scale threshold of 15 particles was set to a random integer within the gray scale level of 0~255, and the initial velocity of the particles was set to 5. The combination of PSO and Otsu calculated the fitness (variance value) of each particle, updated the individual best position and the global best position, further updated the velocities and positions of the particles to find the global best position (the maximum interclass variance), and defined the maximum interclass variance was defined as the optimal threshold value, used the optimal threshold value for background and target segmentation of the gray-scale image and extracted the ridges within the peak region segmented from the gray-scale image by the PSO-Otsu algorithm. A threshold was set according to the length of the ridge line; the ridge lines larger than the threshold were extracted; the valley lines on both sides of the ridge line were found according to the filtered ridge line; and the start and end points of the peak region were obtained from the valley lines. The filtered ridge lines were used to identify the peak location of the peak region, and the edge threshold was set to remove the false peaks due to continuous wavelet transform (CWT) located in the edge region of the C4D signal and to accurately identify the location of the true peak value. Results and Discussions The datasets containing 41, 61 and 102 peaks were tested, and the Receiver Operating Characteristic (ROC) curves and metric values were used as a guideline to evaluate the performance of the peak detection algorithms. Compared to the two methods, multi scale peak detection (MSPD) and CWT-based image segmentation (CWT-IS), the CWT combined with Particle Swarm Optimization based maximum spectral peaks automatic identification algorithm based on Continuous Wavelet Transform combined with Particle Swarm Optimization of Otsu to segment image (CWTSPSO) method of interclass variance segmentation (CWT-IS), the ROC curves of the three groups remained above 0.9. Testing the dataset containing 102 peaks, the ROC curves of MSPD and CWT-IS were also high only in the case of high false discovery rate. The highest metric values of CWTSPSO were 0.976, 0.915, and 0.969, respectively, and the highest metric values of 1 set of MSPD and CWT-IS were 0.754 and 0.505. The results showed that the ROC curves of CWTSPSO in the 3 sets of dataset were not high. Using ROC curves and metric values as a criterion comparison to evaluate the performance of peak detection algorithms, the characteristic peak recognition performance was outstanding, which could achieve a higher correct rate while maintaining a lower false discovery rate, effectively detected more weak and overlapping peaks while detecting fewer false peaks, which was conducive to the enhancement of the spectral peak recognition rate and accuracy of the C4D signals. Conclusions This study provided a fast and accurate method for the identification of characteristic peaks in the spectrograms of ion signals detected by contactless conductivity, CWTSPSO could accurately identify the weak and overlapping peaks in the spectrograms of ion signals detected by contactless conductivity. The automatic identification algorithm of the spectrogram peaks of CWTSPSO would provide powerful support for the characterization and quantification of the signals of nutrient ions detected by contactless conductivity in agricultural soils and would further strengthen the precision of soil fertilization and crop growth management fertilization and crop growth management.
Significance Crop production is related to national food security, economic development and social stability, so timely information on the growth of major crops is of great significance for strengthening the crop production management and ensuring food security. The traditional crop growth monitoring mainly judges the growth of crops by manually observing the shape, color and other appearance characteristics of crops through the external industry, which has better reliability and authenticity, but it will consume a lot of manpower, is inefficient and difficult to carry out monitoring of a large area. With the development of space technology, satellite remote sensing technology provides an opportunity for large area crop growth monitoring. However, the acquisition of optical remote sensing data is often limited by the weather during the peak crop growth season when rain and heat coincide. Synthetic aperture radar (SAR) compensates well for the shortcomings of optical remote sensing, and has a wide demand and great potential for application in crop growth monitoring. However, the current research on crop growth monitoring using SAR data is still relatively small and lacks systematic sorting and summarization. In this paper, the research progress of SAR inversion of crop growth parameters were summarized through comprehensive analysis of existing literature, clarify the main technical methods and application of SAR monitoring of crop growth, and explore the existing problems and look forward to its future research direction. Progress] The current research status of SAR crop growth monitoring were reviewed, the application of SAR technology had gone through several development stages: from the early single-polarization, single-band stage, gradually evolving to the mid-term multi-polarization, multi-band stage, and then to the stage of joint application of tight polarization and optical remote sensing. Then, the research progress and milestone achievements of crop growth monitoring based on SAR data were summarized in three aspects, namely, crop growth SAR remote sensing monitoring indexes, crop growth SAR remote sensing monitoring data and crop growth SAR remote sensing monitoring methods. First, the key parameters of crop growth were summarized, and the crop growth monitoring indexes were divided into morphological indicators, physiological and biochemical indicators, yield indicators and stress indicators. Secondly, the core principle of SAR monitoring of crop growth parameters was introduced, which was based on the interaction between SAR signals and vegetation, and then the specific scattering model and inversion algorithm were used to estimate the crop growth parameters. Then, a detailed summary and analysis of the radar indicators mainly applied to crop growth monitoring were also presented. Finally, SAR remote sensing methods for crop growth monitoring, including mechanistic modeling, empirical modeling, semi-empirical modeling, direct monitoring, and assimilation monitoring of crop growth models, were described, and their applicability and applications in growth monitoring were analyzed. Conclusions and Prospects Four challenges exist in SAR crop growth monitoring are proposed: 1) Compared with the methods of crop growth monitoring using optical remote sensing data, the methods of crop growth monitoring using SAR data are obviously relatively small. The reason may be that SAR remote sensing itself has some inherent shortcomings; 2) Insufficient mining of microwave scattering characteristics, at present, a large number of studies have applied the backward scattering intensity and polarization characteristics to crop growth monitoring, but few have applied the phase information to crop growth monitoring, especially the application study of polarization decomposition parameters to growth monitoring. The research on the application of polarization decomposition parameter to crop growth monitoring is still to be deepened; 3) Compared with the optical vegetation index, the radar vegetation index applied to crop growth monitoring is relatively less; 4 ) Crop growth monitoring based on SAR scattered intensity is mainly based on an empirical model, which is difficult to be extended to different regions and types of crops, and the existence of this limitation prevents the SAR scattering intensity-based technology from effectively realizing its potential in crop growth monitoring. Finally, future research should focus on mining microwave scattering features, utilizing SAR polarization decomposition parameters, developing and optimizing radar vegetation indices, and deepening scattering models for crop growth monitoring.
Objective The investigation of plant photosynthetic phenotypes is essential for unlocking insights into plant physiological characteristics and dissecting morphological traits. However, traditional two-dimensional chlorophyll fluorescence imaging methods struggle to capture the complex three-dimensional spatial variations inherent in plant photosynthetic processes. To boost the efficacy of plant phenotyping and meet the increasingly demand for high-throughput analysis of photosynthetic phenotypes, the development and validation of a novel plant photosynthetic phenotype imaging system was explored, which uniquely combines three-dimensional structured light techniques with chlorophyll fluorescence technology. Methods The plant photosynthetic phenotype imaging system was composed of three primary parts: A tailored light source and projector, a camera, and a motorized filter wheel fitted with filters of various bandwidths, in addition to a terminal unit equipped with a development board and a touchscreen interface. The system was based on the principles and unique characteristics of chlorophyll fluorescence and structured light phase-shifted streak 3D reconstruction techniques. It utilized the custom-designed light source and projector, together with the camera's capability to choose specific wavelength bands, to its full potential. The system employed low-intensity white light within the 400–700 nm spectrum to elicit stable fluorescence, with blue light in the 440–450 nm range optimally triggering the fluorescence response. A projector was used to project dual-frequency, twelve-step phase-shifted stripes onto the plant, enabling the capture of both planar and stripe images, which were essential for the reconstruction of the plant's three-dimensional structure. An motorized filter wheel containing filters for red, green, blue, and near-infrared light, augmented by a filter less wheel for camera collaboration, facilitated the collection of images of plants at different wavelengths under varying lighting conditions. When illuminated with white light, filters corresponding to the red, green, and blue bands were applied to capture multiband images, resulting in color photographs that provides a comprehensive documentation of the plant's visual features. Upon exposure to blue light, the near-infrared filter was employed to capture near-infrared images, yielding data on chlorophyll fluorescence intensity. During the structured light streak projection, no filter was applied to obtain both planar and streak images of the plant, which were then employed in the 3D morphological reconstruction of the plant. The terminal, incorporating a development board and a touch screen, served as the control hub for the data acquisition and subsequent image processing within the plant photosynthetic phenotypic imaging system. It enabled the switching of light sources and the selection of camera bands through a combination of command and serial port control circuits. Following image acquisition, the data were transmitted back to the development board for analysis, processing, storage, and presentation. To validate the accuracy of 3D reconstruction and the reliability of photosynthetic efficiency assessments by the system, a prototype of the plant photosynthetic phenotypic imaging system was developed using 3D structured light and chlorophyll fluorescence technology, in accordance with the aforementioned methods, serving as an experimental validation platform. The accuracy of 3D reconstruction and the effectiveness of photosynthetic analysis capabilities of this imaging system were further confirmed through the analysis and processing of the experimental results, with comparative evaluations conducted against conventional 3D reconstruction methods and traditional chlorophyll fluorescence-based photosynthetic efficiency analyses. Results and Discussions The imaging system utilized for plant photosynthetic phenotypes incorporates a dual-frequency phase-shift algorithm to facilitate the reconstruction of three-dimensional (3D) plant phenotypes. Simultaneously, plant chlorophyll fluorescence images were employed to evaluate the plant's photosynthetic efficiency. This method enabled the analysis of the distribution of photosynthetic efficiency within a 3D space, offering a significant advancement over traditional plant photosynthetic imaging techniques. The 3D phenotype reconstructed using this method exhibits high precision, with an overall reconstruction accuracy of 96.69%. The total error was merely 3.31%, and the time required for 3D reconstruction was only 1.11 s. A comprehensive comparison of the 3D reconstruction approach presented with conventional methods had validated the accuracy of this technique, laying a robust foundation for the precise estimation of a plant's 3D photosynthetic efficiency. In the realm of photosynthetic efficiency analysis, the correlation coefficient between the photosynthetic efficiency values inferred from the chlorophyll fluorescence image analysis and those determined by conventional analysis exceeded 0.9. The experimental findings suggest a significant correlation between the photosynthetic efficiency values obtained using the proposed method and those from traditional methods, which could be characterized by a linear relationship, thereby providing a basis for more precise predictions of plant photosynthetic efficiency. Conclusions The method melds the 3D phenotype of plants with an analysis of photosynthetic efficiency, allowing for a more holistic assessment of the spatial heterogeneity in photosynthetic efficiency among plants by examining the pseudo-color images of chlorophyll fluorescence's spatial distribution. This approach elucidates the discrepancies in photosynthetic efficiency across various regions. The plant photosynthetic phenotype imaging system affords an intuitive and comprehensive view of the photosynthetic efficiency in plants under diverse stress conditions. Additionally, It provides technical support for the analysis of the spatial heterogeneity of high-throughput photosynthetic efficiency in plants.
Objective Brassinolide is an important endogenous plant hormone. In this work, an electrochemical immunosensor for in situ detection of brassinolide was constructed using screen-printed electrode (SPE). Methods Au nanoparticles (AuNPs) were firstly electrodeposited on the surface of SPE electrode by electrochemical workstation, and then CuCl2 nanowires (CuCl2 NWs) were added to the electrode, which can not only increase the conductivity of the electrode, but also Cu2+ can be used as a REDOX probe for the sensor. Finally, Mxene and polydopamine nanocomposite (Mxene@PDA) were selected as the modification materials for SPE electrodes because Mxene has the advantages of large surface area and good electrical conductivity, which can further amplify Cu2+ signals. However, Mxene is easily oxidized and unstable in air. Polydopamine (PDA) contains a large number of catechol and amino groups, which are coated on the surface of Mxene after self-polymerization by dopamine, cutting off the path of oxygen penetration, making Mxene difficult to be oxidized. Mxene@PDA can also be used as a coupling agent to fix more antibodies on the electrode surface, improving the overall biocompatibility, and improve the overall biocompatibility. Results and Discussions The sensor has a wide linear detection range: 0.1 pg/mL to 1 mg/mL, and the detection limit was 0.015 pg/ml (S/N=3). In addition, the content of endogenous brassinolide in wheat was detected by SPE electrodes in vitro and the recovery rate was 98.13% to 104.74%.While verifying the accuracy of the sensor, it also demonstrated its superior stability and sensitivity. Besides, the sensor also showed excellent application potential in the in situ detection of brassinosteroids from wheat leaves. Compared with other brassinolide detection methods, the immunosensor developed in this study has better analytical performance. Conclusions An electrochemical immunosensor for in situ detection of brassinolide was developed for the first time, providing a good electrochemical platform for in situ determination of brassinolide in plant leaves, which has great application potential in precision agriculture.
Objective Traditional methods for detecting mold are time-consuming, labor-intensive, and vulnerable to environmental influences, highlighting the need for a swift, precise, and dependable detection approach. Researchers have utilized visible-near infrared (NIR) spectroscopy for the non-destructive, rapid assessment of wheat moisture content, crude protein content, concealed pests, starch content, dry matter, weight, hardness, origin, and other attributes. However, most of these studies rely on research-grade Visible-NIR spectrometers typically found in laboratories. While these spectrometers offer superior detection accuracy and stability, their bulky size, lack of portability, and high cost hinder their widespread use and adoption across various agricultural product distribution channels. Methods A low-resolution Visible-NIR spectrometer (VNIAPD, with a resolution of 1.6 nm) was utilized to gather wheat data. The aim was to enhance the accuracy of moldy wheat detection by identifying suitable spectral data preprocessing methods using corresponding algorithms. A high-resolution Visible-NIR spectrometer (SINO2040, with a resolution of 0.19 nm) served as a control to validate the instrument and method's effectiveness. The Zhoumai (No. 22) wheat variety was adopted, with a total of 100 samples prepared. The spectra of fresh wheat were scanned and then placed in a constant temperature chamber at 35 °C to replicate the appropriate conditions for mold growth, thereby accelerating the reproduction of naturally occurring mold in the wheat. The degree of mold was categorized based on the cultivation time in the constant temperature chamber, with wheat classified as mildly, moderately, or severely moldy after 3, 6, and 9 days of cultivation, respectively. A total of 400 wheat spectral data points were collected, including 100 samples each of fresh wheat, wheat cultured for 3 days, wheat cultured for 6 days, and wheat cultured for 9 days. Preprocessing methods such as standard deviation normalization (SDN), standard normal variation (SNV), mean centrality (MC), first-order derivatives (1ST), Savitzky-Golay smoothing (SG), and multiple scattering correction (MSC) were applied to the spectral data. Outliers were identified and eliminated using the local outlier factor (LOF) method. Following this, the sequential projection algorithm (SPA) and Least absolute shrinkage and selection operator (LASSO) were used to extract characteristic wavelengths from the preprocessed spectra. Subsequently, six algorithms, including k-nearest neighbors (KNN), support vector machines (SVM), random forests (RF), Naïve-Bayes, back propagation neural networks (BPNN), and deep neural networks (DNN), were employed to model and analyze the feature wavelength spectra, differentiating moldy wheat and classifying the degree of mold. Evaluation criteria encompassed accuracy, modeling time, and model size to aid in selecting the most suitable model for specific application scenarios. Results and discussions Regarding accuracy, even when utilizing the computationally slower and more memory-demanding neural network models BPNN and DNN, both the VNIAPD and SINO2040 achieved a perfect 100% accuracy in the binary classification task of distinguishing between fresh and moldy wheat. They also maintained a faultless 100% accuracy in the ternary classification task that differentiates three varying levels of mold growth. Adopting faster and more memory-efficient shallow models such as KNN, SVM, RF, and Naïve-Bayes, the VNIAPD yielded a top test set accuracy of 97.72% when combined with RF for binary classification. Conversely, SINO2040 achieved 100% accuracy using Naïve-Bayes. In the ternary classification scenario, the VNIAPD hit the mark at 100% accuracy with both KNN and RF, while SINO2040 demonstrated 97.72% accuracy with KNN and SVM. Regarding modeling speed, the shallow machine learning algorithms, including KNN, SVM, RF, and Naïve-Bayes, exhibited quicker training times, with Naïve-Bayes being the swiftest at just 3 ms. In contrast, the neural network algorithms BPNN and DNN required more time for training, taking 3 293 and 18 614 ms, respectively. Regarding memory footprint, BPNN had the largest model size, occupying 4 028 kb, whereas SVM was the most memory-efficient, with a size of only 4 kb. Overall, the VNIAPD matched the SINO2040 in detection accuracy despite having lower optical parameters: A slightly lesser optical resolution of 1.6 nm compared to the SINO2040's 0.19 nm—and a lower cost, highlighting its efficiency and cost-effectiveness in the given context. Conclusions In this study, by comparing different preprocessing methods for spectral data, the optimal data optimization choices for corresponding algorithms were identified. As a result, the low-resolution spectrometer VNIAPD was able to achieve performance on par with the high-resolution spectrometer SINO2040 in detecting moldy wheat, providing a new option for low-cost, non-destructive detection of wheat mold and the degree of moldiness based on Visible-NIR spectroscopy.
Objective The use of pesticides is one of the root causes of food safety problems. Pesticide exposure and pesticide residues can not only lead to environmental pollution issues but also seriously affect human health. In order to meet the rapid and sensitive detection needs of pesticide residues in agricultural products, a method based on lemon juice reduction to prepare silver nanoparticles (AgNPs) is reported in this research. Methods First, fresh lemon juice was filtered through filter paper and diluted to a 2% lemon juice aqueous solution. Then, a certain concentration of AgNO3 solution, 50 mm NaOH solution were prepared and stored at room temperature. Then, 10 mL ddH2O, 2 mL NaOH, 2 mL 2% lemon juice, and 5 mL AgNO3 solution were mixed. When the solution turned to a clear yellow color, the solution was centrifuged to obtain AgNPs. The morphology and structure of AgNPs were observed by transmission electron microscopy (TEM). In order to verify the successful synthesis of the nanoparticles and the distribution characteristics of the nanoparticles, ultraviolet spectroscopy was used for measurement and analysis, and 4-ATP was used as a SERS probe to preliminarily verify the SERS enhancement performance of AgNPs. Furthermore, the content of the main reducing components in lemon juice, namely ascorbic acid, glucose, and fructose was analyzed. The content of ascorbic acid in lemon juice was determined by high-performance liquid chromatography, and the content of glucose and fructose in lemon juice was determined by UV-visible spectrophotometry. To verify the stability and uniformity of the SERS signal of the nanoparticles, 4-ATP was used as an surface enhancement of raman scattering (SERS) probe for detection analysis. The stability of the SERS performance of the colloidal substrate within 41 days and the SERS performance at temperatures ranging from 0-80 °C were analyzed. Using 4-ATP as the SERS probe, the experimental conditions were optimized for the preparation of AgNPs by the lemon juice method, including pH and AgNO3 concentration. To validate the practical usability of the nanoparticles, the solutions of paraquat and carbendazim and the detection limits of pesticide residues on different fruits and vegetables were detected by SERS. Results and discussions The method for preparing AgNPs has the advantages of simple operation, green and easy synthesis. The particle morphology and size of the prepared AgNPs were basically uniform, with a size of about 20 nm. The ultraviolet-visible spectrum of AgNPs solution showed that the absorption peak was about 400 nm and the peak shape was narrow, indicating that the colloidal solution had good homogeneity. The detection limit of 4-ATP as the SERS probe was 10-14 M, indicating that the nanoparticle had a good SERS. In addition, the content of ascorbic acid, the main reducing ingredient, in lemon juice measured by high-performance liquid chromatography (HPLC) was 395.76 μg/mL. The contents of glucose and fructose, which were the main reducing components in lemon juice, were 5.95 and 5.90 mg/mL, respectively. Furthermore, the characterization and analysis results of the AgNPs prepared by the mixed reducing solution prepared according to the concentration data of each component showed that the AgNPs obtained were also uniform in morphology and size, with a diameter of about 20 nm, but the SERS enhancement performance was not as good as that of the AgNPs reduced by lemon juice. The SERS signal uniformity of the AgNPs reduced by lemon juice analyzed results showed that the peak intensity of the SERS spectral of 4-ATP at different sites at the same concentration was not significantly changed for 15 times, and its standard deviation RSD=5.03%, which was much lower than the intersubstrate RSD value (<16%) of the qualified new SERS active substrate for quantitative analysis. The temporal stability and temperature stability of AgNPs analyzed results showed that the nanoparticles still had SERS enhanced performance after 41 days of storage, and had SERS enhanced performance stability over a wide temperature range (0~80 °C). In addition, the optimization results of experimental conditions showed that the optimal pH for the preparation of AgNPs was around 7.5, and the optimal range of AgNO3 concentration used was 1.76×10-4~3.33×10-4 mol/L. Finally, using AgNPs prepared by lemon juice reduction method for pesticide residue SERS detection on the surface of fruits and vegetables, the detection limits for paraquat and carbendazim in solution were as low as 10-14 and 10-10 M, respectively, and the concentration of pesticides showed a good linear relationship with Raman spectral intensity. The lowest detection limits for paraquat and carbendazim residues on different fruits and vegetables were as low as 3.90 ng/kg and 0.22 µg/kg, respectively. Conclusions This work provides a green and convenient method for preparing SERS materials for rapid detection of pesticide residues on fruits and vegetables. This method has practical value for universal operation. The prepared AgNPs can be used for trace pesticide residue detection, providing a pathway for rapid and sensitive detection of pesticide residues in agricultural products.
Objective During the operation of the silage machine, the inclusion of ferrous metal foreign objects such as stray iron wires can inflict severe damage to the machine's critical components and livestock organs. To safeguard against that, a metal detection system with superior performance was developed in this research to enable precise and efficient identification of metal foreign bodies during field operations, ensuring the integrity of the silage process and the well-being of the animals. Methods The ferrous metal detection principle of silage machine was firstly analyzed. The detection coil is the probe of the metal detection system. After being connected in parallel with a capacitor, it is connected to the detection module. The detection coil received the alternating signal generated by the detection module to generate an alternating magnetic field. After the metal object entered the magnetic field, it affects the equivalent resistance and equivalent inductance of the detection coil. The detection module detected the change of the equivalent resistance and equivalent inductance, and then transmited the signal to the control module through the serial peripheral interface (SPI). The control module filtered the signal and transmited it to the display terminal through the serial port. The display terminal could set the threshold. When the data exceeded the threshold, the system performed sound and light alarm and other processing. Hardware part of the metal detection system of silage machine were firstly design. The calculation of the planar spiral coil and the cylindrical coil was carried out and the planar spiral coil was selected as the research object. By using the nondominated sorting genetic algorithm-Ⅱ (NSGA-II) combined with the method of finite element simulation analysis, the wire diameter, inner diameter, outer diameter, layer number and frequency of the coil were determined, and the calculation of the bent coil and the unbent coil and the array coil was carried out. The hardware system was integrated. The software system for the metal detection system was also designed, utilizing an STM32 microcontroller as the control module and LabView for writing the primary program on the upper computer. The system continuously displayed the read data and time-equivalent impedance graph in real-time, allowing for the setting of upper and lower alarm thresholds. When a metal foreign object was detected, the warning light turned red and an alarm sound was emitted, causing the feed roll to stop. To simulate the scenario of metal detection during the operation of a silage machine, a test bench was set up to validate the performance of the metal detection system. Results and Discussions The test results of the metal detection function showed that for a metal wire with a diameter of 0.6 mm and a length of 20 mm, as the inner diameter of the detection coil increased, the maximum alarm distance increased first and then decreased. The maximum alarm distance occured when the inner diameter was 35 mm, which was consistent with the optimization result. The maximum alarm distance was the largest when the detection coil was two layers, and there was no data readout when it was three layers. Therefore, the optimal thickness of the detection coil for this metal detection system was two layers. When the detection distance was greater than 80 mm, the alarm rate began to decrease, and the detection effect was weakened. When the detection distance was within 70 mm, the metal detection system could achieve a 100% alarm rate. The test results of the system response time showed that the average system response time was 0.105 0 s, which was less than the safe transportation time of 0.202 0 s. The system can give an alarm before the metal foreign object reaches the cutter, so the system is safe and effective. Conclusion In this study, a metal detection system for silage machines was designed. A set of optimization methods for metal detection coils was proposed, and the corresponding metal detection software and hardware systems were developed, and the functions of the metal detection system were verified through experiments, which could provide strong technical support for the safe operation of silage machines.
Objective Knowledge services in agricultural scenarios have the characteristics of long periodicity and prolonged activity time. Traditional recommendation models cannot effectively mine hidden information in agricultural scenarios, in order to improve the quality of agricultural knowledge recommendation services, agricultural contextual information based on agricultural time should be fully considered. To address these issues, a Time-aware and filter-enhanced sequential recommendation model for agricultural knowledge (TiFSA) was proposed, integrating temporal perception and enhanced filtering. Methods First, based on the temporal positional embedding, combining the temporal information of farmers' interactions with positional embedding based on time perception, it helped to learn project relevance based on agricultural season in agricultural contexts. A multi-head self-attention network recommendation algorithm based on time-awareness was proposed for the agricultural knowledge recommendation task, which extracted different interaction time information in the user interaction sequence and introduced it into the multi-head self-attention network to calculate the attention weight, which encoded the user's periodic interaction information based on the agricultural time, and also effectively captured the user's dynamic preference information over time. Then, through the temporal positional embedding, a filter filtering algorithm was introduced to adaptively attenuate the noise in farmers' situational data adaptively. The filtering algorithm was introduced to enhance the filtering module to effectively filter the noisy information in the agricultural dataset and alleviate the overfitting problem due to the poorly normalized and sparse agricultural dataset. By endowing the model with lower time complexity and adaptive noise attenuation capability. The applicability of this method in agricultural scenarios was improved. Next, a multi-head self attention network with temporal information was constructed to achieve unified modeling of time, projects, and features, and represent farmers' preferences of farmers over time in context, thereby providing reliable recommendation results for users. Finally, the AdamW optimizer was used to update and compute the model parameters. AdamW added L2 regularization and an appropriate penalty mechanism for larger weights, which could update all weights more smoothly and alleviate the problem of falling into local minima. Applied in the field of agricultural recommendation, it could further improve the training effect of the model. The experimental data came from user likes, comments, and corresponding time information in the "National Agricultural Knowledge Intelligent Service Cloud Platform", and the dataset ml-1m in the movie recommendation scenario was selected as an auxiliary validation of the performance of this model. Results and Discussions According to the user interaction sequence datasets in the "National Agricultural Knowledge Intelligent Service Cloud Platform", from the experimental results, it could be learned that TiFSA outperforms the other models on two different datasets, in which the enhancement was more obvious on the Agriculture dataset, where HR and NDCG were improved by 14.02% and 16.19%, respectively, compared to the suboptimal model, TiSASRec; while on the ml-1m dataset compared to the suboptimal model, SASRec, HR and NDCG were improved by 1.90% and 2.30%, respectively. In summary, the TiFSA model proposed in this paper has a large improvement compared with other models, which verifies verified the effectiveness of the TiFSA model and showed that the time interval information of farmer interaction and the filtering algorithm play an important role in the improvement of the model performance in the agricultural context. From the results of the ablation experiments, it could be seen that when the time-aware and enhanced filtering modules were removed, the values of the two metrics HR@10 and NDCG@10 were 0.293 6 and 0.203 9, respectively, and the recommended performance was poor. When only the time-aware module and only the augmentation filtering module were removed, the experimental results had different degrees of improvement compared to TiFSA-tf, and the TiFSA model proposed in this paper achieved the optimal performance in the two evaluation metrics. When only the multi-head self-attention network was utilized for recommendation, both recommendation metrics of the model were lower, indicating that the traditional sequence recommendation method that only considered the item number was not applicable to agricultural scenarios. When the augmented filtering module was introduced without the time-aware module, the model performance was improved, but still failed to achieve the ideal recommendation effect. When only the time-aware module was introduced without the augmented filtering module, there was a significant improvement in the model effect, which proved that the time-aware module was more applicable to agricultural scenarios and can effectively improve the model performance of the sequence recommendation task. When both time-aware and augmented filtering modules were introduced, the model performance was further improved, which on the one hand illustrated the dependence of the augmented filtering module on the time-aware module, and on the other hand verified the necessity of adopting the augmented filtering to the time-aware self-attention network model. Conclusions This research proposes an agricultural knowledge recommendation model that integrates time-awareness and augmented filtering, which introduces the user's interaction time interval into the embedded information, so that the model effectively learns the information of agricultural time in the agricultural scene, and the prediction of the user's interaction time and the object is more closely related to the actual scene; augmented filtering algorithms are used to attenuate the noise in the agricultural data. At the same time, the enhanced filtering algorithm is used to attenuate the noise in the agricultural data, and can be effectively integrated into the model for use, further improving the recommendation performance of the model. The experimental results show the effectiveness of the proposed TiFSA model on the agricultural dataset. The ablation experiments confirm the positive effect of time-awareness and enhanced filtering modules on the improvement of recommendation performance.
Objective Flax, characterized by its short growth cycle and strong adaptability, is one of the major cash crops in northern China. Due to its versatile uses and unique quality, it holds a significant position in China's oil and fiber crops. The quality of flax seeds directly affects the yield of the flax plant. Seed evaluation is a crucial step in the breeding process of flax. Common parameters used in the seed evaluation process of flax include circumference, area, length axis, and 1 000-seed weight. To ensure the high-quality production of flax crops, it is of great significance to understand the phenotypic characteristics of flax seeds, select different resources as parents based on breeding objectives, and adopt other methods for the breeding, cultivation, and evaluation of seed quality and traits of flax. Methods In response to the high error rates and low efficiency issues observed during the automated seed testing of flax seeds, the measurement methods were explored of flax seed contours based on machine vision research. The flax seed images were preprocessed, and the collected color images were converted to grayscale. A filtering and smoothing process was applied to obtain binary images. To address the issues of flax seed overlap and adhesion, a contour fitting image segmentation method based on fused corner features was proposed. This method incorporated adaptive threshold selection during edge detection of the image contour. Only multi-seed target areas that met certain criteria were subjected to image segmentation processing, while single-seed areas bypassed this step and were directly summarized for seed testing data. After obtaining the multi-seed adhesion target areas, the flax seeds underwent contour approximation, corner extraction, and contour fitting. Based on the provided image contour information, the image contour shape was approximated to another contour shape with fewer vertices, and the original contour curve was simplified to a more regular and compact line segment or polygon, minimizing computational complexity. All line shape characteristics in the image were marked as much as possible. Since the pixel intensity variations in different directions of image corners were significant, the second derivative matrix based on pixel grayscale values was used to detect image corners. Based on the contour approximation algorithm, contour corner detection was performed to obtain the coordinates of each corner. The resulting contour points and corners were used as outputs to further improve the accuracy and precision of subsequent contour fitting methods, resulting in a two-dimensional discrete point dataset of the image contour. Using the contour point dataset as an input, the geometric moments of the image contour were calculated, and the optimal solution for the ellipse parameters was obtained through numerical optimization based on the least squares method and the geometric features of the ellipse shape. Ultimately, the optimal contour was fitted to the given image, achieving the segmentation and counting of flax seed images. Meanwhile, each pixel in the digital image was a uniform small square in size and shape, so the circumference, area, and major and minor axes of the flax seeds could be represented by the total number of pixels occupied by the seeds in the image. The weight of a single seed could be calculated by dividing the total weight of the seeds by the total number of seeds detected by the contour, thereby obtaining the weight of the individual seed and converting it accordingly. Through the pixelization of the 1 yuan and 1 jiao coins from the fifth iteration of the 2019 Renminbi, a summary of the circumference, area, major axis, minor axis, and 1 000-seed weight of the flax seeds was achieved. Additionally, based on the aforementioned method, this study designed an automated real-time analysis system for flax seed testing data, realizing the automation of flax seed testing research. Experiments were conducted on images of flax seeds captured by an industrial camera. Results and Discussions The proposed automated seed identification method achieved an accuracy rate of 97.28% for statistically distinguishing different varieties of flax seeds. The average processing time for 100 seeds was 69.58 ms. Compared to the extreme erosion algorithm and the watershed algorithm based on distance transformation, the proposed method improved the average calculation accuracy by 19.6% over the extreme erosion algorithm and required a shorter average computation time than the direct use of the watershed algorithm. Considering the practical needs of automated seed identification, this method did not employ methods such as dilation or erosion for image morphology processing, thereby preserving the original features of the image to the greatest extent possible. Additionally, the flax seed automated seed identification data real-time analysis system could process image information in batches. By executing data summarization functions, it automatically generated corresponding data table folders, storing the corresponding image data summary tables. Conclusions The proposed method exhibits superior computational accuracy and processing speed, with shorter operation time and robustness. It is highly adaptable and able to accurately acquire the morphological feature parameters of flax seeds in bulk, ensuring measurement errors remain within 10%, which could provide technical support for future flax seed evaluation and related industrial development.
Objective To effectively tackle the unique attributes of wheat leaf pests and diseases in their native environment, a high-caliber and efficient pest detection model named YOLOv8-SS (You Only Look Once Version 8-SS) was proposed. This innovative model is engineered to accurately identify pests, thereby providing a solid scientific foundation for their prevention and management strategies. Methods A total of 3 639 raw datasets of images of wheat leaf pests and diseases were collected from 6 different wheat pests and diseases in various farmlands in the Yuchong County area of Gansu Province, at different periods of time, using mobile phones. This collection demonstrated the team's proficiency and commitment to advancing agricultural research. The dataset was meticulously constructed using the LabelImg software to accurately label the images with targeted pest species. To guarantee the model's superior generalization capabilities, the dataset was strategically divided into a training set and a test set in an 8:2 ratio. The dataset includes thorough observations and recordings of the wheat leaf blade's appearance, texture, color, as well as other variables that could influence these characteristics. The compiled dataset proved to be an invaluable asset for both training and validation activities. Leveraging the YOLOv8 algorithm, an enhanced lightweight convolutional neural network, ShuffleNetv2, was selected as the basis network for feature extraction from images. This was accomplished by integrating a 3×3 Depthwise Convolution (DWConv) kernel, the h-swish activation function, and a Squeeze-and-Excitation Network (SENet) attention mechanism. These enhancements streamlined the model by diminishing the parameter count and computational demands, all while sustaining high detection precision. The deployment of these sophisticated methodologies exemplified the researchers' commitment and passion for innovation. The YOLOv8 model employs the SEnet attention mechanism module within both its Backbone and Neck components, significantly reducing computational load while bolstering accuracy. This method exemplifies the model's exceptional performance, distinguishing it from other models in the domain. By integrating a dedicated small target detection layer, the model's capabilities have been augmented, enabling more efficient and precise pest and disease detection. The introduction of a new detection feature map, sized 160×160 pixels, enables the network to concentrate on identifying small-targeted pests and diseases, thereby enhancing the accuracy of pest and disease recognition. Results and Discussion The YOLOv8-SS wheat leaf pests and diseases detection model has been significantly improved to accurately detect wheat leaf pests and diseases in their natural environment. By employing the refined ShuffleNet V2 within the DarkNet-53 framework, as opposed to the conventional YOLOv8, under identical experimental settings, the model exhibited a 4.53% increase in recognition accuracy and a 4.91% improvement in F1-Score, compared to the initial model. Furthermore, the incorporation of a dedicated small target detection layer led to a subsequent rise in accuracy and F1-Scores of 2.31% and 2.16%, respectively, despite a minimal upsurge in the number of parameters and computational requirements. The integration of the SEnet attention mechanism module into the YOLOv8 model resulted in a detection accuracy rate increase of 1.85% and an F1-Score enhancement of 2.72%. Furthermore, by swapping the original neural network architecture with an enhanced ShuffleNet V2 and appending a compact object detection sublayer (namely YOLOv8-SS), the resulting model exhibited a heightened recognition accuracy of 89.41% and an F1-Score of 88.12%. The YOLOv8-SS variant substantially outperformed the standard YOLOv8, showing a remarkable enhancement of 10.11% and 9.92% in accuracy, respectively. This outcome strikingly illustrates the YOLOv8-SS's prowess in balancing speed with precision. Moreover, it achieves convergence at a more rapid pace, requiring approximately 40 training epochs, to surpass other renowned models such as Faster R-CNN, MobileNetV2, SSD, YOLOv5, YOLOX, and the original YOLOv8 in accuracy. Specifically, the YOLOv8-SS boasted an average accuracy 23.01%, 15.13%, 11%, 25.21%, 27.52%, and 10.11% greater than that of the competing models, respectively. In a head-to-head trial involving a public dataset (LWDCD 2020) and a custom-built dataset, the LWDCD 2020 dataset yielded a striking accuracy of 91.30%, outperforming the custom-built dataset by a margin of 1.89% when utilizing the same network architecture, YOLOv8-SS. The AI Challenger 2018-6 and Plant-Village-5 datasets did not perform as robustly, achieving accuracy rates of 86.90% and 86.78% respectively. The YOLOv8-SS model has shown substantial improvements in both feature extraction and learning capabilities over the original YOLOv8, particularly excelling in natural environments with intricate, unstructured backdrops. Conclusion The YOLOv8-SS model is meticulously designed to deliver unmatched recognition accuracy while consuming a minimal amount of storage space. In contrast to conventional detection models, this groundbreaking model exhibits superior detection accuracy and speed, rendering it exceedingly valuable across various applications. This breakthrough serves as an invaluable resource for cutting-edge research on crop pest and disease detection within natural environments featuring complex, unstructured backgrounds. Our method is versatile and yields significantly enhanced detection performance, all while maintaining a lean model architecture. This renders it highly appropriate for real-world scenarios involving large-scale crop pest and disease detection.
[Significance] Big Models, or Foundation Models, have offered a new paradigm in smart agriculture. These models, built on the Transformer architecture, incorporate numerous parameters and have undergone extensive training, often showing excellent performance and adaptability, making them effective in addressing agricultural issues where data is limited. Integrating big models in agriculture promises to pave the way for a more comprehensive form of agricultural intelligence, capable of processing diverse inputs, making informed decisions, and potentially overseeing entire farming systems autonomously. [Progress] The fundamental concepts and core technologies of big models are initially elaborated from five aspects: the generation and core principles of the Transformer architecture, scaling laws of extending big models, large-scale self-supervised learning, the general capabilities and adaptions of big models, and the emerging capabilities of big models. Subsequently, the possible application scenarios of the big model in the agricultural field are analyzed in detail, the development status of big models is described based on three types of the models: Large language models (LLMs), large vision models (LVMs), and large multi-modal models (LMMs). The progress of applying big models in agriculture is discussed, and the achievements are presented. [Conclusions and Prospects] The challenges and key tasks of applying big models technology in agriculture are analyzed. Firstly, the current datasets used for agricultural big models are somewhat limited, and the process of constructing these datasets can be both expensive and potentially problematic in terms of copyright issues. There is a call for creating more extensive, more openly accessible datasets to facilitate future advancements. Secondly, the complexity of big models, due to their extensive parameter counts, poses significant challenges in terms of training and deployment. However, there is optimism that future methodological improvements will streamline these processes by optimizing memory and computational efficiency, thereby enhancing the performance of big models in agriculture. Thirdly, these advanced models demonstrate strong proficiency in analyzing image and text data, suggesting potential future applications in integrating real-time data from IoT devices and the Internet to make informed decisions, manage multi-modal data, and potentially operate machinery within autonomous agricultural systems. Finally, the dissemination and implementation of these big models in the public agricultural sphere are deemed crucial. The public availability of these models is expected to refine their capabilities through user feedback and alleviate the workload on humans by providing sophisticated and accurate agricultural advice, which could revolutionize agricultural practices.
[Significance] The crop phenotype is the visible result of the complex interplay between crop genes and the environment. It reflects the physiological, ecological, and dynamic aspects of crop growth and development, serving as a critical component in the realm of advanced breeding techniques. By systematically analyzing crop phenotypes, researchers can gain valuable insights into gene function and identify genetic factors that influence important crop traits. This information can then be leveraged to effectively harness germplasm resources and develop breakthrough varieties. Utilizing data-driven, intelligent, dynamic, and non-invasive methods for measuring crop phenotypes allows researchers to accurately capture key growth traits and parameters, providing essential data for breeding and selecting superior crop varieties throughout the entire growth cycle. This article provides an overview of intelligent identification technologies for crop agronomic traits and morphological structural phenotypes. [Progress] Crop phenotype acquisition equipment serves as the essential foundation for acquiring, analyzing, measuring, and identifying crop phenotypes. This equipment enables detailed monitoring of crop growth status. The article presents an overview of the functions, performance, and applications of the leading high-throughput crop phenotyping platforms, as well as an analysis of the characteristics of various sensing and imaging devices used to obtain crop phenotypic information. The rapid advancement of high-throughput crop phenotyping platforms and sensory imaging equipment has facilitated the integration of cutting-edge imaging technology, spectroscopy technology, and deep learning algorithms. These technologies enable the automatic and high-throughput acquisition of yield, resistance, quality, and other relevant traits of large-scale crops, leading to the generation of extensive multi-dimensional, multi-scale, and multi-modal crop phenotypic data. This advancement supports the rapid progression of crop phenomics. The article also discusses the research progress of intelligent recognition technologies for agronomic traits such as crop plant height acquisition, crop organ detection, and counting, as well as crop ideotype recognition, crop morphological information measurement, and crop three-dimensional reconstruction for morphological structure intelligent recognition. Furthermore, this article outlines the main challenges faced in this field, including: difficulties in data collection in complex environments, high requirements for data scale, diversity, and preprocessing, the need to improve the lightweight nature and generalization ability of models, as well as the high cost of data collection equipment and the need to enhance practicality. [Conclusions and Prospects] Finally, this article puts forward the development directions of crop phenotype intelligent recognition technology, including: developing new and low cost intelligent field equipment for acquiring and analyzing crop phenotypes, enhancing the standardization and consistency of field crop phenotype acquisition, strengthening the generality of intelligent crop phenotype recognition models, researching crop phenotype recognition methods that involve multi-perspective, multimodal, multi-point continuous analysis, and spatiotemporal feature fusion, as well as improving model interpretability.
[Objective] In recent years, there has been a significant increase in the severity of leaf diseases in maize, with a noticeable trend of mixed occurrence. This poses a serious threat to the yield and quality of maize. However, there is a lack of studies that combine the identification of different types of leaf diseases and their severity classification, which cannot meet the needs of disease prevention and control under the mixed occurrence of different diseases and different severities in actual maize fields. [Methods] A method was proposed for identifying the types of typical leaf diseases in maize and classifying their severity using hyperspectral technology. Hyperspectral data of three leaf diseases of maize: northern corn leaf blight (NCLB), southern corn leaf blight (SCLB) and southern corn rust (SCR), were obtained through greenhouse pathogen inoculation and natural inoculation. The spectral data were preprocessed by spectral standardization, SG filtering, sensitive band extraction and vegetation index calculation, to explore the spectral characteristics of the three leaf diseases of maize. Then, the inverse frequency weighting method was utilized to balance the number of samples to reduce the overfitting phenomenon caused by sample imbalance. Relief-F and variable selection using random forests (VSURF) method were employed to optimize the sensitive spectral features, including band features and vegetation index features, to construct models for disease type identification based on the full stages of disease development (including all disease severities) and for individual disease severities using several representative machine learning approaches, demonstrating the effectiveness of the research method. Furthermore, the study individual occurrence severity classification models were also constructed for each single maize leaf disease, including the NCLB, SCLB and SCR severity classification models, respectively, aiming to achieve full-process recognition and disease severity classification for different leaf diseases. Overall accuracy (OA) and Macro F1 were used to evaluate the model accuracy in this study. Results and Discussion The research results showed significant spectrum changes of three kinds of maize leaf diseases primarily focusing on the visible (550-680 nm), red edge (740-760 nm), near-infrared (760-1 000 nm) and shortwave infrared (1 300-1 800 nm) bands. Disease-specific spectral features, optimized based on disease spectral response rules, effectively identified disease species and classify their severity. Moreover, vegetation index features were more effective in identifying disease-specific information than sensitive band features. This was primarily due to the noise and information redundancy present in the selected hyperspectral sensitive bands, whereas vegetation index could reduce the influence of background and atmospheric noise to a certain extent by integrating relevant spectral signals through band calculation, so as to achieve higher precision in the model. Among several machine learning algorithms, the support vector machine (SVM) method exhibited better robustness than random forest (RF) and decision tree (DT). In the full stage of disease development, the optimal overall accuracy (OA) of the disease classification model constructed by SVM based on vegetation index reached 77.51%, with a Macro F1 of 0.77, representing a 28.75% increase in OA and 0.30 higher of Macro F1 compared to the model based on sensitive bands. Additionally, the accuracy of the disease classification model with a single severity of the disease increased with the severity of the disease. The accuracy of disease classification during the early stage of disease development (OA=70.31%) closely approached that of the full disease development stage (OA=77.51%). Subsequently, in the moderate disease severity stage, the optimal accuracy of disease classification (OA=80.00%) surpassed the optimal accuracy of disease classification in the full disease development stage. Furthermore, the optimal accuracy of disease classification under severe severity reached 95.06%, with a Macro F1 of 0.94. This heightened accuracy during the severity stage can be attributed to significant changes in pigment content, water content and cell structure of the diseased leaves, intensifying the spectral response of each disease and enhancing the differentiation between different diseases. In disease severity classification model, the optimal accuracy of the three models for maize leaf disease severity all exceeded 70%. Among the three kinds of disease severity classification results, the NCLB severity classification model exhibited the best performance. The NCLB severity classification model, utilizing SVM based on the optimal vegetation index features, achieved an OA of 86.25%, with a Macro F1 of 0.85. In comparison, the accuracy of the SCLB severity classification model (OA=70.35%, Macro F1=0.70) and SCR severity classification model (OA=71.39%, Macro F1=0.69) were lower than that of NCLB. [Conclusions] The aforementioned results demonstrate the potential to effectively identify and classify the types and severity of common leaf diseases in maize using hyperspectral data. This lays the groundwork for research and provides a theoretical basis for large-scale crop disease monitoring, contributing to precision prevention and control as well as promoting green agriculture.
[Objective] The widespread prevalence of sclerotinia disease poses a significant challenge to the cultivation and supply of oilseed rape, not only results in substantial yield losses and decreased oil content in infected plant seeds but also severely impacts crop productivity and quality, leading to significant economic losses. To solve the problems of complex operation, environmental pollution, sample destruction and low detection efficiency of traditional chemical detection methods, a Bi-directional Gate Recurrent Unit (Bi-GRU) model based on space-spectrum feature fusion was constructed to achieve hyperspectral images (HSIs) segmentation of oilseed rape sclerotinia infected area. [Methods] The spectral characteristics of sclerotinia disease from a spectral perspective was initially explored. Significantly varying spectral reflectance was notably observed around 550 nm and within the wavelength range of 750-1 000 nm at different locations on rapeseed leaves. As the severity of sclerotinia infection increased, the differences in reflectance at these wavelengths became more pronounced. Subsequently, a rapeseed leaf sclerotinia disease dataset comprising 400 HSIs was curated using an intelligent data annotation tool. This dataset was divided into three subsets: a training set with 280 HSIs, a validation set with 40 HSIs, and a test set with 80 HSIs. Expanding on this, a 7×7 pixel neighborhood was extracted as the spatial feature of the target pixel, incorporating both spatial and spectral features effectively. Leveraging the Bi-GRU model enabled simultaneous feature extraction at any point within the sequence data, eliminating the impact of the order of spatial-spectral data fusion on the model's performance. The model comprises four key components: an input layer, hidden layers, fully connected layers, and an output layer. The Bi-GRU model in this study consisted of two hidden layers, each housing 512 GRU neurons. The forward hidden layer computed sequence information at the current time step, while the backward hidden layer retrieves the sequence in reverse, incorporating reversed-order information. These two hidden layers were linked to a fully connected layer, providing both forward and reversed-order information to all neurons during training. The Bi-GRU model included two fully connected layers, each with 1 000 neurons, and an output layer with two neurons representing the healthy and diseased classes, respectively. [Results and Discussions] To thoroughly validate the comprehensive performance of the proposed Bi-GRU model and assess the effectiveness of the spatial-spectral information fusion mechanism, relevant comparative analysis experiments were conducted. These experiments primarily focused on five key parameters—ClassAP(1), ClassAP(2), mean average precision (mAP), mean intersection over union (mIoU), and Kappa coefficient—to provide a comprehensive evaluation of the Bi-GRU model's performance. The comprehensive performance analysis revealed that the Bi-GRU model, when compared to mainstream convolutional neural network (CNN) and long short-term memory (LSTM) models, demonstrated superior overall performance in detecting rapeseed sclerotinia disease. Notably, the proposed Bi-GRU model achieved an mAP of 93.7%, showcasing a 7.1% precision improvement over the CNN model. The bidirectional architecture, coupled with spatial-spectral fusion data, effectively enhanced detection accuracy. Furthermore, the study visually presented the segmentation results of sclerotinia disease-infected areas using CNN, Bi-LSTM, and Bi-GRU models. A comparison with the Ground-Truth data revealed that the Bi-GRU model outperformed the CNN and Bi-LSTM models in detecting sclerotinia disease at various infection stages. Additionally, the Dice coefficient was employed to comprehensively assess the actual detection performance of different models at early, middle, and late infection stages. The dice coefficients for the Bi-GRU model at these stages were 83.8%, 89.4% and 89.2%, respectively. While early infection detection accuracy was relatively lower, the spatial-spectral data fusion mechanism significantly enhanced the effectiveness of detecting early sclerotinia infections in oilseed rape. [Conclusions] This study introduces a Bi-GRU model that integrates spatial and spectral information to accurately and efficiently identify the infected areas of oilseed rape sclerotinia disease. This approach not only addresses the challenge of detecting early stages of sclerotinia infection but also establishes a basis for high-throughput non-destructive detection of the disease.
[Objective] It is of great significance to improve the efficiency and accuracy of crop pest detection in complex natural environments, and to change the current reliance on expert manual identification in the agricultural production process. Targeting the problems of small target size, mimicry with crops, low detection accuracy, and slow algorithm reasoning speed in crop pest detection, a complex scene crop pest target detection algorithm named YOLOv8-Entend was proposed in this research. [Methods] Firstly, the GSConv was introduecd to enhance the model's receptive field, allowing for global feature aggregation. This mechanism enables feature aggregation at both node and global levels simultaneously, obtaining local features from neighboring nodes through neighbor sampling and aggregation operations, enhancing the model's receptive field and semantic understanding ability. Additionally, some Convs were replaced with lightweight Ghost Convolutions and HorBlock was utilized to capture longer-term feature dependencies. The recursive gate convolution employed gating mechanisms to remember and transmit previous information, capturing long-term correlations. Furthermore, Concat was replaced with BiFPN for richer feature fusion. The bidirectional fusion of depth features from top to bottom and from bottom to top enhances the transmission of feature information acrossed different network layers. Utilizing the VoVGSCSP module, feature maps of different scales were connected to create longer feature map vectors, increasing model diversity and enhancing small object detection. The convolutional block attention module (CBAM) attention mechanism was introduced to strengthen features of field pests and reduce background weights caused by complexity. Next, the Wise IoU dynamic non-monotonic focusing mechanism was implemented to evaluate the quality of anchor boxes using "outlier" instead of IoU. This mechanism also included a gradient gain allocation strategy, which reduced the competitiveness of high-quality anchor frames and minimizes harmful gradients from low-quality examples. This approach allowed WIoU to concentrate on anchor boxes of average quality, improving the network model's generalization ability and overall performance. Subsequently, the improved YOLOv8-Extend model was compared with the original YOLOv8 model, YOLOv5, YOLOv8-GSCONV, YOLOv8-BiFPN, and YOLOv8-CBAM to validate the accuracy and precision of model detection. Finally, the model was deployed on edge devices for inference verification to confirm its effectiveness in practical application scenarios. [Results and Discussions] The results indicated that the improved YOLOv8-Extend model achieved notable improvements in accuracy, recall, mAP@0.5, and mAP@0.5:0.95 evaluation indices. Specifically, there were increases of 2.6%, 3.6%, 2.4% and 7.2%, respectively, showcasing superior detection performance. YOLOv8-Extend and YOLOv8 run respectively on the edge computing device JETSON ORIN NX 16 GB and were accelerated by TensorRT, mAP@0.5 improved by 4.6%, FPS reached 57.6, meeting real-time detection requirements. The YOLOv8-Extend model demonstrated better adaptability in complex agricultural scenarios and exhibited clear advantages in detecting small pests and pests sharing similar growth environments in practical data collection. The accuracy in detecting challenging data saw a notable increased of 11.9%. Through algorithm refinement, the model showcased improved capability in extracting and focusing on features in crop pest target detection, addressing issues such as small targets, similar background textures, and challenging feature extraction. [Conclusions] The YOLOv8-Extend model introduced in this study significantly boosts detection accuracy and recognition rates while upholding high operational efficiency. It is suitable for deployment on edge terminal computing devices to facilitate real-time detection of crop pests, offering technological advancements and methodologies for the advancement of cost-effective terminal-based automatic pest recognition systems. This research can serve as a valuable resource and aid in the intelligent detection of other small targets, as well as in optimizing model structures.
[Objective] In recent years, there has been a steady increase in the occurrence and fatality rates of shrimp diseases, causing substantial impacts in shrimp aquaculture. These diseases are marked by their swift onset, high infectivity, complex control requirements, and elevated mortality rates. With the continuous growth of shrimp factory farming, traditional manual detection approaches are no longer able to keep pace with the current requirements. Hence, there is an urgent necessity for an automated solution to identify shrimp diseases. The main goal of this research is to create a cost-effective inspection method using computer vision that achieves a harmonious balance between cost efficiency and detection accuracy. The improved YOLOv8 (You Only Look Once) network and multiple features were employed to detect shrimp diseases. [Methods] To address the issue of surface foam interference, the improved YOLOv8 network was applied to detect and extract surface shrimps as the primary focus of the image. This target detection approach accurately recognizes objects of interest in the image, determining their category and location, with extraction results surpassing those of threshold segmentation. Taking into account the cost limitations of platform computing power in practical production settings, the network was optimized by reducing parameters and computations, thereby improving detection speed and deployment efficiency. Additionally, the Farnberck optical flow method and gray level co-occurrence matrix (GLCM) were employed to capture the movement and image texture features of shrimp video clips. A dataset was created using these extracted multiple feature parameters, and a Support Vector Machine (SVM) classifier was trained to categorize the multiple feature parameters in video clips, facilitating the detection of shrimp health. [Results and Discussions] The improved YOLOv8 in this study effectively enhanced detection accuracy without increasing the number of parameters and flops. According to the results of the ablation experiment, replacing the backbone network with FasterNet lightweight backbone network significantly reduces the number of parameters and computation, albeit at the cost of decreased accuracy. However, after integrating the efficient multi-scale attention (EMA) on the neck, the mAP0.5 increased by 0.3% compared to YOLOv8s, while mAP0.95 only decreased by 2.1%. Furthermore, the parameter count decreased by 45%, and FLOPs decreased by 42%. The improved YOLOv8 exhibits remarkable performance, ranking second only to YOLOv7 in terms of mAP0.5 and mAP0.95, with respective reductions of 0.4% and 0.6%. Additionally, it possesses a significantly reduced parameter count and FLOPS compared to YOLOv7, matching those of YOLOv5. Despite the YOLOv7-Tiny and YOLOv8-VanillaNet models boasting lower parameters and Flops, their accuracy lags behind that of the improved YOLOv8. The mAP0.5 and mAP0.95 of YOLOv7-Tiny and YOLOv8-VanillaNet are 22.4%, 36.2%, 2.3%, and 4.7% lower than that of the improved YOLOv8, respectively. Using a support vector machine (SVM) trained on a comprehensive dataset incorporating multiple feature, the classifier achieved an impressive accuracy rate of 97.625%. The 150 normal fragments and the 150 diseased fragments were randomly selected as test samples. The classifier exhibited a detection accuracy of 89% on this dataset of the 300 samples. This result indicates that the combination of features extracted using the Farnberck optical flow method and GLCM can effectively capture the distinguishing dynamics of movement speed and direction between infected and healthy shrimp. In this research, the majority of errors stem from the incorrect recognition of diseased segments as normal segments, accounting for 88.2% of the total error. These errors can be categorized into three main types: 1) The first type occurs when floating foam obstructs the water surface, resulting in a small number of shrimp being extracted from the image. 2) The second type is attributed to changes in water movement. In this study, nanotubes were used for oxygenation, leading to the generation of sprays on the water surface, which affected the movement of shrimp. 3) The third type of error is linked to video quality. When the video's pixel count is low, the difference in optical flow between diseased shrimp and normal shrimp becomes relatively small. Therefore, it is advisable to adjust the collection area based on the actual production environment and enhance video quality. [Conclusions] The multiple features introduced in this study effectively capture the movement of shrimp, and can be employed for disease detection. The improved YOLOv8 is particularly well-suited for platforms with limited computational resources and is feasible for deployment in actual production settings. However, the experiment was conducted in a factory farming environment, limiting the applicability of the method to other farming environments. Overall, this method only requires consumer-grade cameras as image acquisition equipment and has lower requirements on the detection platform, and can provide a theoretical basis and methodological support for the future application of aquatic disease detection methods.
[Objective] Accurate identification of insect pests is crucial for the effective prevention and control of crop infestations. However, existing pest identification methods primarily rely on traditional machine learning or deep learning techniques that are trained on seen classes. These methods falter when they encounter unseen pest species not included in the training set, due to the absence of image samples. An innovative method was proposed to address the zero-shot recognition challenge for pests. [Methods] The novel zero-shot learning (ZSL) method proposed in this study was capable of identifying unseen pest species. First, a comprehensive pest image dataset was assembled, sourced from field photography conducted around Beijing over several years, and from web crawling. The final dataset consisted of 2 000 images across 20 classes of adult Lepidoptera insects, with 100 images per class. During data preprocessing, a semantic dataset was manually curated by defining attributes related to color, pattern, size, and shape for six parts: antennae, back, tail, legs, wings, and overall appearance. Each image was annotated to form a 65-dimensional attribute vector for each class, resulting in a 20×65 semantic attribute matrix with rows representing each class and columns representing attribute values. Subsequently, 16 classes were designated as seen classes, and 4 as unseen classes. Next, a novel zero-shot pest recognition method was proposed, focusing on synthesizing high-quality pseudo-visual features aligned with semantic information using a generator. The wasserstein generative adversarial networks (WGAN) architecture was strategically employed as the fundamental network backbone. Conventional generative adversarial networks (GANs) have been known to suffer from training instabilities, mode collapse, and convergence issues, which can severely hinder their performance and applicability. The WGAN architecture addresses these inherent limitations through a principled reformulation of the objective function. In the proposed method, the contrastive module was designed to capture highly discriminative visual features that could effectively distinguish between different insect classes. It operated by creating positive and negative pairs of instances within a batch. Positive pairs consisted of different views of the same class, while negative pairs were formed from instances belonging to different classes. The contrastive loss function encouraged the learned representations of positive pairs to be similar while pushing the representations of negative pairs apart. Tightly integrated with the WGAN structure, this module substantially improved the generation quality of the generator. Furthermore, the visual-semantic alignment module enforced consistency constraints from both visual and semantic perspectives. This module constructed a cross-modal embedding space, mapping visual and semantic features via two projection layers: One for mapping visual features into the cross-modal space, and another for mapping semantic features. The visual projection layer took the synthesized pseudo-visual features from the generator as input, while the semantic projection layer ingested the class-level semantic vectors. Within this cross-modal embedding space, the module enforced two key constraints: Maximizing the similarity between same-class visual-semantic pairs and minimizing the similarity between different-class pairs. This was achieved through a carefully designed loss function that encourages the projected visual and semantic representations to be closely aligned for instances belonging to the same class, while pushing apart the representations of different classes. The visual-semantic alignment module acted as a regularizer, preventing the generator from producing features that deviated from the desired semantic information. This regularization effect complemented the discriminative power gained from the contrastive module, resulting in a generator that produces high-quality, diverse, and semantically aligned pseudo-visual features. [Results and Discussions] The proposed method was evaluated on several popular ZSL benchmarks, including CUB, AWA, FLO, and SUN. The results demonstrated that the proposed method achieved state-of-the-art performance across these datasets, with a maximum improvement of 2.8% over the previous best method, CE-GZSL. This outcome fully demonstrated the method's broad effectiveness in different benchmarks and its outstanding generalization ability. On the self-constructed 20-class insect dataset, the method also exhibited exceptional recognition accuracy. Under the standard ZSL setting, it achieved a precise recognition rate of 77.4%, outperforming CE-GZSL by 2.1%. Under the generalized ZSL setting, it achieved a harmonic mean accuracy of 78.3%, making a notable 1.2% improvement. This metric provided a balanced assessment of the model's performance across seen and unseen classes, ensuring that high accuracy on unseen classes does not come at the cost of forgetting seen classes. These results on the pest dataset, coupled with the performance on public benchmarks, firmly validated the effectiveness of the proposed method. [Conclusions] The proposed zero-shot pest recognition method represents a step forward in addressing the challenges of pest management. It effectively generalized pest visual features to unseen classes, enabling zero-shot pest recognition. It can facilitate pests identification tasks that lack training samples, thereby assisting in the discovery and prevention of novel crop pests. Future research will focus on expanding the range of pest species to further enhance the model's practical applicability.
[Objective] With the development of agricultural informatization, a large amount of information about agricultural diseases exists in the form of text. However, due to problems such as nested entities and confusion of entity types, traditional named entities recognition (NER) methods often face challenges of low accuracy when processing agricultural disease text. To address this issue, this study proposes a new agricultural disease NER method called RoFormer-PointerNet, which combines the RoFormer pre-trained model with the PointerNet baseline model. The aim of this method is to improve the accuracy of entity recognition in agricultural disease text, providing more accurate data support for intelligent analysis, early warning, and prevention of agricultural diseases. [Methods] This method first utilized the RoFormer pre-trained model to perform deep vectorization processing on the input agricultural disease text. This step was a crucial foundation for the subsequent entity extraction task. As an advanced natural language processing model, the RoFormer pre-trained model's unique rotational position embedding approach endowed it with powerful capabilities in capturing textual positional information. In agricultural disease text, due to the diversity of terminology and the existence of polysemy, traditional entity recognition methods often faced challenges in confusing entity types. However, through its unique positional embedding mechanism, the RoFormer model was able to incorporate more positional information into the vector representation, effectively enriching the feature information of words. This characteristic enabled the model to more accurately distinguish between different entity types in subsequent entity extraction tasks, reducing the possibility of type confusion. After completing the vectorization representation of the text, this study further emploied a pointer network for entity extraction. The pointer network was an advanced sequence labeling approach that utilizes head and tail pointers to annotate entities within sentences. This labeling method was more flexible compared to traditional sequence labeling methods as it was not restricted by fixed entity structures, enabling the accurate extraction of all types of entities within sentences, including complex entities with nested relationships. In agricultural disease text, entity extraction often faced the challenge of nesting, such as when multiple different entity types are nested within a single disease symptom description. By introducing the pointer network, this study effectively addressed this issue of entity nesting, improving the accuracy and completeness of entity extraction. [Results and Discussions] To validate the performance of the RoFormer-PointerNet method, this study constructed an agricultural disease dataset, which comprised 2 867 annotated corpora and a total of 10 282 entities, including eight entity types such as disease names, crop names, disease characteristics, pathogens, infected areas, disease factors, prevention and control methods, and disease stages. In comparative experiments with other pre-trained models such as Word2Vec, BERT, and RoBERTa, RoFormer-PointerNet demonstrated superiority in model precision, recall, and F1-Score, achieving 87.49%, 85.76% and 86.62%, respectively. This result demonstrated the effectiveness of the RoFormer pre-trained model. Additionally, to verify the advantage of RoFormer-PointerNet in mitigating the issue of nested entities, this study compared it with the widely used bidirectional long short-term memory neural network (BiLSTM) and conditional random field (CRF) models combined with the RoFormer pre-trained model as decoding methods. RoFormer-PointerNet outperformed the RoFormer-BiLSTM, RoFormer-CRF, and RoFormer-BiLSTM-CRF models by 4.8%, 5.67% and 3.87%, respectively. The experimental results indicated that RoFormer-PointerNet significantly outperforms other models in entity recognition performance, confirming the effectiveness of the pointer network model in addressing nested entity issues. To validate the superiority of the RoFormer-PointerNet method in agricultural disease NER, a comparative experiment was conducted with eight mainstream NER models such as BiLSTM-CRF, BERT-BiLSTM-CRF, and W2NER. The experimental results showed that the RoFormer-PointerNet method achieved precision, recall, and F1-Score of 87.49%, 85.76% and 86.62%, respectively in the agricultural disease dataset, reaching the optimal level among similar methods. This result further verified the superior performance of the RoFormer-PointerNet method in agricultural disease NER tasks. [Conclusions] The agricultural disease NER method RoFormer-PointerNet, proposed in this study and based on the RoFormer pre-trained model, demonstrates significant advantages in addressing issues such as nested entities and type confusion during the entity extraction process. This method effectively identifies entities in Chinese agricultural disease texts, enhancing the accuracy of entity recognition and providing robust data support for intelligent analysis, early warning, and prevention of agricultural diseases. This research outcome holds significant importance for promoting the development of agricultural informatization and intelligence.
[Objective] There's a growing demand among plant cultivators and breeders for efficient methods to acquire plant phenotypic traits at high throughput, facilitating the establishment of mappings from phenotypes to genotypes. By integrating mobile phenotyping platforms with improved instance segmentation techniques, researchers have achieved a significant advancement in the automation and accuracy of phenotypic data extraction. Addressing the need for rapid extraction of leaf age and canopy width phenotypes in strawberry plants cultivated in controlled environments, this study introduces a novel high-throughput phenotyping extraction approach leveraging a mobile phenotyping platform and instance segmentation technology. [Methods] Data acquisition was conducted using a compact mobile phenotyping platform equipped with an array of sensors, including an RGB sensor, and edge control computers, capable of capturing overhead images of potted strawberry plants in greenhouses. Targeted adjustments to the network structure were made to develop an enhanced convolutional neural network (Mask R-CNN) model for processing strawberry plant image data and rapidly extracting plant phenotypic information. The model initially employed a split-attention networks (ResNeSt) backbone with a group attention module, replacing the original network to improve the precision and efficiency of image feature extraction. During training, the model adopted the Mosaic method, suitable for instance segmentation data augmentation, to expand the dataset of strawberry images. Additionally, it optimized the original cross-entropy classification loss function with a binary cross-entropy loss function to achieve better detection accuracy of plants and leaves. Based on this, the improved Mask R-CNN description involves post-processing of training results. It utilized the positional relationship between leaf and plant masks to statistically count the number of leaves. Additionally, it employed segmentation masks and image calibration against true values to calculate the canopy width of the plant. [Results and Discussions] This research conducted a thorough evaluation and comparison of the performance of an improved Mask R-CNN model, underpinned by the ResNeSt-101 backbone network. This model achieved a commendable mask accuracy of 80.1% and a detection box accuracy of 89.6%. It demonstrated the ability to efficiently estimate the age of strawberry leaves, demonstrating a high plant detection rate of 99.3% and a leaf count accuracy of 98.0%. This accuracy marked a significant improvement over the original Mask R-CNN model and meeting the precise needs for phenotypic data extraction. The method displayed notable accuracy in measuring the canopy widths of strawberry plants, with errors falling below 5% in about 98.1% of cases, highlighting its effectiveness in phenotypic dimension evaluation. Moreover, the model operated at a speed of 12.9 frames per second (FPS) on edge devices, effectively balancing accuracy and operational efficiency. This speed proved adequate for real-time applications, enabling rapid phenotypic data extraction even on devices with limited computational capabilitie. [Conclusions] This study successfully deployed a mobile phenotyping platform combined with instance segmentation techniques to analyze image data and extract various phenotypic indicators of strawberry plant. Notably, the method demonstrates remarkable robustness. The seamless fusion of mobile platforms and advanced image processing methods not only enhances efficiency but also ignifies a shift towards data-driven decision-making in agriculture.
[Objective] Currently, the lack of computerized systems to monitor the quality of cabbage transplants is a notable shortcoming in the agricultural industry, where transplanting operations play a crucial role in determining the overall yield and quality of the crop. To address this problem, a lightweight and efficient algorithm was developed to monitor the status of cabbage transplants in a natural environment. [Methods] First, the cabbage image dataset was established, the cabbage images in the natural environment were collected, the collected image data were filtered and the transplanting status of the cabbage was set as normal seedling (upright and intact seedling), buried seedling (whose stems and leaves were buried by the soil) and exposed seedling (whose roots were exposed), and the dataset was manually categorized and labelled using a graphical image annotation tool (LabelImg) so that corresponding XML files could be generated. And the dataset was pre-processed with data enhancement methods such as flipping, cropping, blurring and random brightness mode to eliminate the scale and position differences between the cabbages in the test and training sets and to improve the imbalance of the data. Then, a cabbage transplantation state detection model based on YOLOv8s (You Only Look Once Version 8s) was designed. To address the problem that light and soil have a large influence on the identification of the transplantation state of cabbage in the natural environment, a multi-scale attention mechanism was embedded to increase the number of features in the model, and a multi-scale attention mechanism was embedded to increase the number of features in the model. Embedding the multi-scale attention mechanism to increase the algorithm's attention to the target region and improve the network's attention to target features at different scales, so as to improve the model's detection efficiency and target recognition accuracy, and reduce the leakage rate; by combining with deformable convolution, more useful target information was captured to improve the model's target recognition and convergence effect, and the model complexity increased by C3-layer convolution was reduced, which further reduced the model complexity. Due to the unsatisfactory localization effect of the algorithm, the focal extended intersection over union loss (Focal-EIoU Loss) was introduced to solve the problem of violent oscillation of the loss value caused by low-quality samples, and the influence weight of high-quality samples on the loss value was increased while the influence of low-quality samples was suppressed, so as to improve the convergence speed and localization accuracy of the algorithm. [Results and Discussions] Eventually, the algorithm was put through a stringent testing phase, yielding a remarkable recognition accuracy of 96.2% for the task of cabbage transplantation state. This was an improvement of 2.8% over the widely used YOLOv8s. Moreover, when benchmarked against other prominent target detection models, the algorithm emerged as a clear winner. It showcased a notable enhancement of 3% and 8.9% in detection performance compared to YOLOv3-tiny. Simultaneously, it also managed to achieve a 3.7% increase in the recall rate, a metric that measured the efficiency of the algorithm in identifying actual targets among false positives. On a comparative note, the algorithm outperformed YOLOv5 in terms of recall rate by 1.1%, 2% and 1.5%, respectively. When pitted against the robust faster region-based convolutional neural network (Faster R-CNN), the algorithm demonstrated a significant boost in recall rate by 20.8% and 11.4%, resulting in an overall improvement of 13%. A similar trend was observed when the algorithm was compared to the single shot multibox detector (SSD) model, with a notable 9.4% and 6.1% improvement in recall rate. The final experimental results show that when the enhanced model was compared with YOLOv7-tiny, the recognition accuracy was increased by 3%, and the recall rate was increased by 3.5%. These impressive results validated the superiority of the algorithm in terms of accuracy and localization ability within the target area. The algorithm effectively eliminates interferenced factors such as soil and background impurities, thereby enhancing its performance and making it an ideal choice for tasks such as cabbage transplantation state recognition. [Conclusions] The experimental results show that the proposed cabbage transplantation state detection method can meet the accuracy and real-time requirements for the identification of cabbage transplantation state, and the detection accuracy and localization accuracy of the improved model perform better when the target is smaller and there are weeds and other interferences in the background. Therefore, the method proposed in this study can improve the efficiency of cabbage transplantation quality measurement, reduce the time and labor, and improve the automation of field transplantation quality survey.
[Objective] The fresh cut rose industry has shown a positive growth trend in recent years, demonstrating sustained development. Considering the current fresh cut roses grading process relies on simple manual grading, which results in low efficiency and accuracy, a new model named Flower-YOLOv8s was proposed for grading detection of fresh cut roses. [Methods] The flower head of a single rose against a uniform background was selected as the primary detection target. Subsequently, fresh cut roses were categorized into four distinct grades: A, B, C, and D. These grades were determined based on factors such as color, size, and freshness, ensuring a comprehensive and objective grading system. A novel dataset contenting 778 images was specifically tailored for rose fresh-cut flower grading and detection was constructed. This dataset served as the foundation for our subsequent experiments and analysis. To further enhance the performance of the YOLOv8s model, two cutting-edge attention convolutional block attention module (CBAM) and spatial attention module (SAM) were introduced separately for comparison experiments. These modules were seamlessly integrated into the backbone network of the YOLOv8s model to enhance its ability to focus on salient features and suppressing irrelevant information. Moreover, selecting and optimizing the SAM module by reducing the number of convolution kernels, incorporating a depth-separable convolution module and reducing the number of input channels to improve the module's efficiency and contribute to reducing the overall computational complexity of the model. The convolution layer (Conv) in the C2f module was replaced by the depth separable convolution (DWConv), and then combined with Optimized-SAM was introduced into the C2f structure, giving birth to the Flower-YOLOv8s model. Precision, recall and F1 score were used as evaluation indicators. [Results and Discussions] Ablation results showed that the Flower-YOLOv8s model proposed in this study, namely YOLOv8s+DWConv+Optimized-SAM, the recall rate was 95.4%, which was 3.8% higher and the average accuracy, 0.2% higher than that of YOLOv8s with DWConv alone. When compared to the baseline model YOLOv8s, the Flower-YOLOv8s model exhibited a remarkable 2.1% increase in accuracy, reaching a peak of 97.4%. Furthermore, mAP was augmented by 0.7%, demonstrating the model's superior performance across various evaluation metrics. The effectiveness of adding Optimized-SAM was proved. From the overall experimental results, the number of parameters of Flower-YOLOv8s was reduced by 2.26 M compared with the baseline model YOLOv8s, and the reasoning time was also reduced from 15.6 to 5.7 ms. Therefore, the Flower-YOLOv8s model was superior to the baseline model in terms of accuracy rate, average accuracy, number of parameters, detection time and model size. The performances of Flower-YOLOv8s network were compared with other target detection algorithms of Fast-RCNN, Faster-RCNN and first-stage target detection models of SSD, YOLOv3, YOLOv5s and YOLOv8s to verify the superiority under the same condition and the same data set. The average precision values of the Flower-YOLOv8s model proposed in this study were 2.6%, 19.4%, 6.5%, 1.7%, 1.9% and 0.7% higher than those of Fast-RCNN, Faster-RCNN, SSD, YOLOv3, YOLOv5s and YOLOv8s, respectively. Compared with YOLOv8s with higher recall rate, Flower-YOLOv8s reduced model size, inference time and parameter number by 4.5 MB, 9.9 ms and 2.26 M, respectively. Notably, the Flower-YOLOv8s model achieved these improvements while simultaneously reducing model parameters and computational complexity. [Conclusions] The Flower-YOLOv8s model not only demonstrated superior detection accuracy but also exhibited a reduction in model parameters and computational complexity. This lightweight yet powerful model is highly suitable for real-time applications, making it a promising candidate for flower grading and detection tasks in the agricultural and horticultural industries.
[Objective] Wheat lodging is one of the key isuess threatening stable and high yields. Lodging detection technology based on deep learning generally limited to identifying lodging at a single growth stage of wheat, while lodging may occur at various stages of the growth cycle. Moreover, the morphological characteristics of lodging vary significantly as the growth period progresses, posing a challenge to the feature capturing ability of deep learning models. The aim is exploring a deep learning-based method for detecting wheat lodging boundaries across multiple growth stages to achieve automatic and accurate monitoring of wheat lodging. [Methods] A model called Lodging2Former was proposed, which integrates the innovative hierarchical interactive feature pyramid network (HI-FPN ) on top of the advanced segmentation model Mask2Former. The key focus of this network design lies in enhancing the fusion and interaction between feature maps at adjacent hierarchical levels, enabling the model to effectively integrate feature information at different scales. Building upon this, even in complex field backgrounds, the Lodging2Former model significantly enhances the recognition and capturing capabilities of wheat lodging features at multiple growth stages. [Results and Discussions] The Lodging2Former model demonstrated superiority in mean average precision (mAP) compared to several mainstream algorithms such as mask region-based convolutional neural network (Mask R-CNN), segmenting objects by locations (SOLOv2), and Mask2Former. When applied to the scenario of detecting lodging in mixed growth stage wheat, the model achieved mAP values of 79.5%, 40.2%, and 43.4% at thresholds of 0.5, 0.75, and 0.5 to 0.95, respectively. Compared to Mask2Former, the performance of the improved model was enhanced by 1.3% to 4.3%. Compared to SOLOv2, a growth of 9.9% to 30.7% in mAP was achieved; and compared to the classic Mask R-CNN, a significant improvement of 24.2% to 26.4% was obtained. Furthermore, regardless of the IoU threshold standard, the Lodging2Former exhibited the best detection performance, demonstrating good robustness and adaptability in the face of potential influencing factors such as field environment changes. [Conclusions] The experimental results indicated that the proposed HI-FPN network could effectively utilize contextual semantics and detailed information in images. By extracting rich multi-scale features, it enabled the Lodging2Former model to more accurately detect lodging areas of wheat across different growth stages, confirming the potential and value of HI-FPN in detecting lodging in multi-growth-stage wheat.
[Objective] The daylily, a perennial herb in the lily family, boasts a rich nutritional profile. Given its economic importance, enhancing its yield is a crucial objective. However, current research on daylily cultivation is limited, especially regarding three-dimensional dynamic growth simulation of daylily plants. In order to establish a technological foundation for improved cultivation management, growth dynamics prediction, and the development of plant variety types in daylily crops, this study introduces an innovative three-dimensional dynamic growth and yield simulation model for daylily plants. [Methods] The open-source GroIMP software platform was used to simulate and visualize three-dimensional scenes. With Datong daylily, the primary cultivated variety of daylily in the Datong area, as the research subject, a field experiment was conducted from March to September 2022, which covered the growth season of daylily. Through actual cultivation experiment measurements, morphological data and leaf photosynthetic physiological parameters of daylily leaves, flower stems, flower buds, and other organs were collected. The functional-structural plant model (FSPM) platform's three-dimensional modeling technology was employed to establish the Cloud Cover-based solar radiation models (CSRMs) and the Farquhar, von Camerer, and Berry model (FvCB model) suitable for daylily. Moreover, based on the source-sink relationship of daylily, the carbon allocation model of daylily photosynthetic products was developed. By using the β growth function, the growth simulation model of daylily organs was constructed, and the daily morphological data of daylily during the growth period were calculated, achieving the three-dimensional dynamic growth and yield simulation of daylily plants. Finally, the model was validated with measured data. [Results and Discussions] The coefficient of determination (R2) between the measured and simulated outdoor surface solar radiation was 0.87, accompanied by a Root Mean Squared Error (RMSE) of 28.52 W/m2. For the simulated model of each organ of the daylily plant, the R2 of the measured against the predicted values ranged from 0.896 to 0.984, with an RMSE varying between 1.4 and 17.7 cm. The R2 of the average flower bud yield simulation was 0.880, accompanied by an RMSE of 0.5 g. The overall F-value spanned from 82.244 to 1 168.533, while the Sig. value was consistently below the 0.05 significance level, suggesting a robust fit and statistical significance for the aforementioned models. Subsequently, a thorough examination of the light interaction, temperature influences, and photosynthetic attributes of daylily leaves throughout their growth cycle was carried out. The findings revealed that leaf nutrition growth played a pivotal role in the early phase of daylily's growth, followed by the contribution of leaf and flower stem nutrition in the middle stage, and finally the growth of daylily flower buds, which is the crucial period for yield formation, in the later stages. Analyzing the photosynthetic traits of daylily leaves comprehensively, it was observed that the photosynthetic rate was relatively low in the early spring as the new leaves were initially emerging and reached a plateau during the summer. Considering real-world climate conditions, the actual net photosynthetic rate was marginally lower than the rate verified under optimal conditions, with the simulated net assimilation rate typically ranging from 2 to 4 μmol CO2/(m2·s). [Conclusions] The three-dimensional dynamic growth model of daylily plants proposed in this study can faithfully articulate the growth laws and morphological traits of daylily plants across the three primary growth stages. This model not only illustrates the three-dimensional dynamic growth of daylily plants but also effectively mimics the yield data of daylily flower buds. The simulation outcomes concur with actual conditions, demonstrating a high level of reliability.
[Significance] The mechanization, automation and intelligentization of agricultural equipment are key factors to improve operation efficiency, free up labor force and promote the sustainable development of agriculture. It is also the hot spot of research and development of agricultural machinery industry in the future. In China, hills and mountains serves as vital production bases for agricultural products, accounting for about 70% of the country's land area. In addition, these regions face various environmental factors such as steep slopes, narrow road, small plots, complex terrain and landforms, as well as harsh working environment. Moreover, there is a lack of reliable agricultural machinery support across various production stages, along with a shortage of theoretical frameworks to guide the research and development of agricultural machinery tailored to hilly and mountainous locales. [Progress] This article focuses on the research advances of tractor leveling and anti-overturning systems in hilly and mountainous areas, including tractor body, cab and seat leveling technology, tractor rear suspension and implement leveling slope adaptive technology, and research progress on tractor anti-overturning protection devices and warning technology. The vehicle body leveling mechanism can be roughly divided into five types based on its different working modes: parallel four bar, center of gravity adjustable, hydraulic differential high, folding and twisting waist, and omnidirectional leveling. These mechanisms aim to address the issue of vehicle tilting and easy overturning when traversing or working on sloping or rugged roads. By keeping the vehicle body posture horizontal or adjusting the center of gravity within a stable range, the overall driving safety of the vehicle can be improved to ensure the accuracy of operation. Leveling the driver's cab and seats can mitigate the lateral bumps experienced by the driver during rough or sloping operations, reducing driver fatigue and minimizing strain on the lumbar and cervical spine, thereby enhancing driving comfort. The adaptive technology of tractor rear suspension and implement leveling on slopes can ensure that the tractor maintains consistent horizontal contact with the ground in hilly and mountainous areas, avoiding changes in the posture of the suspended implement with the swing of the body or the driving path, which may affect the operation effect. The tractor rollover protection device and warning technology have garnered significant attention in recent years. Prioritizing driver safety, rollover warning system can alert the driver in advance of the dangerous state of the tractor, automatically adjust the vehicle before rollover, or automatically open the rollover protection device when it is about to rollover, and timely send accident reports to emergency contacts, thereby ensuring the safety of the driver to the greatest extent possible. [Conclusions and Prospects] The future development directions of hill and mountain tractor leveling, anti-overturning early warning, unmanned, automatic technology were looked forward: Structure optimization, high sensitivity, good stability of mountain tractor leveling system research; Study on copying system of agricultural machinery with good slope adaptability; Research on anti-rollover early warning technology of environment perception and automatic interference; Research on precision navigation technology, intelligent monitoring technology and remote scheduling and management technology of agricultural machinery; Theoretical study on longitudinal stability of sloping land. This review could provide reference for the research and development of high reliability and high safety mountain tractor in line with the complex working environment in hill and mountain areas.
[Significance] Soil stands as the fundamental pillar of agricultural production, with its quality being intrinsically linked to the efficiency and sustainability of farming practices. Historically, the intensive cultivation and soil erosion have led to a marked deterioration in some arable lands, characterized by a sharp decrease in soil organic matter, diminished fertility, and a decline in soil's structural integrity and ecological functions. In the strategic framework of safeguarding national food security and advancing the frontiers of smart and precision agriculture, the march towards agricultural modernization continues apace, intensifying the imperative for meticulous soil quality management. Consequently, there is an urgent need for the rrapid acquisition of soil's physical and chemical parameters. Interdisciplinary scholars have delved into soil monitoring research, achieving notable advancements that promise to revolutionize the way we understand and manage soil resource. [Progress] Utilizing the the Web of Science platform, a comprehensive literature search was conducted on the topic of "soil," further refined with supplementary keywords such as "electrochemistry", "spectroscopy", "electromagnetic", "ground-penetrating radar", and "satellite". The resulting literature was screened, synthesized, and imported into the CiteSpace visualization tool. A keyword emergence map was yielded, which delineates the trajectory of research in soil physical and chemical parameter detection technology. Analysis of the keyword emergence map reveals a paradigm shift in the acquisition of soil physical and chemical parameters, transitioning from conventional indoor chemical and spectrometry analyses to outdoor, real-time detection methods. Notably, soil sensors integrated into drones and satellites have garnered considerable interest. Additionally, emerging monitoring technologies, including biosensing and terahertz spectroscopy, have made their mark in recent years. Drawing from this analysis, the prevailing technologies for soil physical and chemical parameter information acquisition in agricultural fields have been categorized and summarized. These include: 1. Rapid Laboratory Testing Techniques: Primarily hinged on electrochemical and spectrometry analysis, these methods offer the dual benefits of time and resource efficiency alongside high precision; 2. Rapid Near-Ground Sensing Techniques: Leveraging electromagnetic induction, ground-penetrating radar, and various spectral sensors (multispectral, hyperspectral, and thermal infrared), these techniques are characterized by their high detection accuracy and swift operation. 3. Satellite Remote Sensing Techniques: Employing direct inversion, indirect inversion, and combined analysis methods, these approaches are prized for their efficiency and extensive coverage. 4. Innovative Rapid Acquisition Technologies: Stemming from interdisciplinary research, these include biosensing, environmental magnetism, terahertz spectroscopy, and gamma spectroscopy, each offering novel avenues for soil parameter detection. An in-depth examination and synthesis of the principles, applications, merits, and limitations of each technology have been provided. Moreover, a forward-looking perspective on the future trajectory of soil physical and chemical parameter acquisition technology has been offered, taking into account current research trends and hotspots. [Conclusions and Prospects] Current advancements in the technology for rapaid acquiring soil physical and chemical parameters in agricultural fields have been commendable, yet certain challenges persist. The development of near-ground monitoring sensors is constrained by cost, and their reliability, adaptability, and specialization require enhancement to effectively contend with the intricate and varied conditions of farmland environments. Additionally, remote sensing inversion techniques are confronted with existing limitations in data acquisition, processing, and application. To further develop the soil physical and chemical parameter acquisition technology and foster the evolution of smart agriculture, future research could beneficially delve into the following four areas: Designing portable, intelligent, and cost-effective near-ground soil information acquisition systems and equipment to facilitate rapid on-site soil information detection; Enhancing the performance of low-altitude soil information acquisition platforms and refine the methods for data interpretation to ensure more accurate insights; Integrating multifactorial considerations to construct robust satellite remote sensing inversion models, leveraging diverse and open cloud computing platforms for in-depth data analysis and mining; Engaging in thorough research on the fusion of multi-source data in the acquisition of soil physical and chemical parameter information, developing soil information sensing algorithms and models with strong generalizability and high reliability to achieve rapaid, precise, and intelligent acquisition of soil parameters.
[Objective] To fully utilize and protect farmland and lay a solid foundation for the sustainable use of land, it is particularly important to obtain real-time and precise information regarding farmland area, distribution, and other factors. Leveraging remote sensing technology to obtain farmland data can meet the requirements of large-scale coverage and timeliness. However, the current research and application of deep learning methods in remote sensing for cultivated land identification still requires further improvement in terms of depth and accuracy. The objective of this study is to investigate the potential application of deep learning methods in remote sensing for identifying cultivated land in the hilly areas of Southwest China, to provide insights for enhancing agricultural land utilization and regulation, and for harmonizing the relationship between cultivated land and the economy and ecology. [Methods] Santai county, Mianyang city, Sichuan province, China (30°42'34"~31°26'35"N, 104°43'04"~105°18'13"E) was selected as the study area. High-resolution imagery from two scenes captured by the Gaofen-6 (GF-6) satellite served as the primary image data source. Additionally, 30-meter resolution DEM data from the United States National Aeronautics and Space Administration (NASA) in 2020 was utilized. A land cover data product, SinoLC-1, was also incorporated for comparative evaluation of the accuracy of various extraction methods' results. Four deep learning models, namely Unet, PSPNet, DeeplabV3+, and Unet++, were utilized for remote sensing land identification research in cultivated areas. The study also involved analyzing the identification accuracy of cultivated land in high-resolution satellite images by combining the results of the random forest (RF) algorithm along with the deep learning models. A validation dataset was constructed by randomly generating 1 000 vector validation points within the research area. Concurrently, Google Earth satellite images with a resolution of 0.3 m were used for manual visual interpretation to determine the land cover type of the pixels where the validation points are located. The identification results of each model were compared using a confusion matrix to compute five accuracy evaluation metrics: Overall accuracy (OA), intersection over union (IoU), mean intersection over union (MIoU), F1-Score, and Kappa Coefficient to assess the cultivated land identification accuracy of different models and data products. [Results and Discussions] The deep learning models displayed significant advances in accuracy evaluation metrics, surpassing the performance of traditional machine learning approaches like RF and the latest land cover product, SinoLC-1 Landcover. Among the models assessed, the UNet++ model performed the best, its F1-Score, IoU, MIoU, OA, and Kappa coefficient values were 0.92, 85.93%, 81.93%, 90.60%, and 0.80, respectively. DeeplabV3+, UNet, and PSPNet methods followed suit. These performance metrics underscored the superior accuracy of the UNet++ model in precisely identifying and segmenting cultivated land, with a remarkable increase in accuracy of nearly 20% than machine learning methods and 50% for land cover products. Four typical areas of town, water body, forest land and contiguous cultivated land were selected to visually compare the results of cultivated land identification results. It could be observed that the deep learning models generally exhibited consistent distribution patterns with the satellite imageries, accurately delineating the boundaries of cultivated land and demonstrating overall satisfactory performance. However, due to the complex features in remote sensing images, the deep learning models still encountered certain challenges of omission and misclassification in extracting cultivated land. Among them, the UNet++ model showed the closest overall extraction results to the ground truth and exhibited advantages in terms of completeness of cultivated land extraction, discrimination between cultivated land and other land classes, and boundary extraction compared to other models. Using the UNet++ model with the highest recognition accuracy, two types of images constructed with different features—solely spectral features and spectral combined with terrain features—were utilized for cultivated land extraction. Based on the three metrics of IoU, OA, and Kappa, the model incorporating both spectral and terrain features showed improvements of 0.98%, 1.10%, and 0.01% compared to the model using only spectral features. This indicated that fusing spectral and terrain features can achieve information complementarity, further enhancing the identification effectiveness of cultivated land. [Conclusions] This study focuses on the practicality and reliability of automatic cultivated land extraction using four different deep learning models, based on high-resolution satellite imagery from the GF-6 in Santai county in China. Based on the cultivated land extraction results in Santai county and the differences in network structures among the four deep learning models, it was found that the UNet++ model, based on UNet, can effectively improve the accuracy of cultivated land extraction by introducing the mechanism of skip connections. Overall, this study demonstrates the effectiveness and practical value of deep learning methods in obtaining accurate farmland information from high-resolution remote sensing imagery.
[Objective] The accurate estimation of terraced field areas is crucial for addressing issues such as slope erosion control, water retention, soil conservation, and increasing food production. The use of high-resolution remote sensing imagery for terraced field information extraction holds significant importance in these aspects. However, as imaging sensor technologies continue to advance, traditional methods focusing on shallow features may no longer be sufficient for precise and efficient extraction in complex terrains and environments. Deep learning techniques offer a promising solution for accurately extracting terraced field areas from high-resolution remote sensing imagery. By utilizing these advanced algorithms, detailed terraced field characteristics with higher levels of automation can be better identified and analyzed. The aim of this research is to explore a proper deep learning algorithm for accurate terraced field area extraction in high-resolution remote sensing imagery. [Methods] Firstly, a terraced dataset was created using high-resolution remote sensing images captured by the Gaofen-6 satellite during fallow periods. The dataset construction process involved data preprocessing, sample annotation, sample cropping, and dataset partitioning with training set augmentation. To ensure a comprehensive representation of terraced field morphologies, 14 typical regions were selected as training areas based on the topographical distribution characteristics of Yuanyang county. To address misclassifications near image edges caused by limited contextual information, a sliding window approach with a size of 256 pixels and a stride of 192 pixels in each direction was utilized to vary the positions of terraced fields in the images. Additionally, geometric augmentation techniques were applied to both images and labels to enhance data diversity, resulting in a high-resolution terraced remote sensing dataset. Secondly, an improved DeepLab v3+ model was proposed. In the encoder section, a lightweight MobileNet v2 was utilized instead of Xception as the backbone network for the semantic segmentation model. Two shallow features from the 4th and 7th layers of the MobileNet v2 network were extracted to capture relevant information. To address the need for local details and global context simultaneously, the multi-scale feature fusion (MSFF) module was employed to replace the atrous spatial pyramid pooling (ASPP) module. The MSFF module utilized a series of dilated convolutions with increasing dilation rates to handle information loss. Furthermore, a coordinate attention mechanism was applied to both shallow and deep features to enhance the network's understanding of targets. This design aimed to lightweight the DeepLab v3+ model while maintaining segmentation accuracy, thus improving its efficiency for practical applications. [Results and Discussions] The research findings reveal the following key points: (1) The model trained using a combination of near-infrared, red, and green (NirRG) bands demonstrated the optimal overall performance, achieving precision, recall, F1-Score, and intersection over union (IoU) values of 90.11%, 90.22%, 90.17% and 82.10%, respectively. The classification results indicated higher accuracy and fewer discrepancies, with an error in reference area of only 12 hm2. (2) Spatial distribution patterns of terraced fields in Yuanyang county were identified through the deep learning model. The majority of terraced fields were found within the slope range of 8º to 25º, covering 84.97% of the total terraced area. Additionally, there was a noticeable concentration of terraced fields within the altitude range of 1 000 m to 2 000 m, accounting for 95.02% of the total terraced area. (3) A comparison with the original DeepLab v3+ network showed that the improved DeepLab v3+ model exhibited enhancements in terms of precision, recall, F1-Score, and IoU by 4.62%, 2.61%, 3.81% and 2.81%, respectively. Furthermore, the improved DeepLab v3+ outperformed UNet and the original DeepLab v3+ in terms of parameter count and floating-point operations. Its parameter count was only 28.6% of UNet and 19.5% of the original DeepLab v3+, while the floating-point operations were only 1/5 of UNet and DeepLab v3+. This not only improved computational efficiency but also made the enhanced model more suitable for resource-limited or computationally less powerful environments. The lightweighting of the DeepLab v3+ network led to improvements in accuracy and speed. However, the slection of the NirGB band combination during fallow periods significantly impacted the model's generalization ability. [Conclusions] The research findings highlights the significant contribution of the near-infrared (NIR) band in enhancing the model's ability to learn terraced field features. Comparing different band combinations, it was evident that the NirRG combination resulted in the highest overall recognition performance and precision metrics for terraced fields. In contrast to PSPNet, UNet, and the original DeepLab v3+, the proposed model showcased superior accuracy and performance on the terraced field dataset. Noteworthy improvements were observed in the total parameter count, floating-point operations, and the Epoch that led to optimal model performance, outperforming UNet and DeepLab v3+. This study underscores the heightened accuracy of deep learning in identifying terraced fields from high-resolution remote sensing imagery, providing valuable insights for enhanced monitoring and management of terraced landscapes.
[Objective] Farmland consolidation for agricultural mechanization in hilly and mountainous areas can alter the landscape pattern, elevation, slope and microgeomorphology of cultivated land. It is of great significance to assess the ecological risk of cultivated land to provide data reference for the subsequent farmland consolidation for agricultural mechanization. This study aims to assess the ecological risk of cultivated land before and after farmland consolidation for agricultural mechanization in hilly and mountainous areas, and to explore the relationship between cultivated land ecological risk and cultivated land slope. [Methods] Twenty counties in Tongnan district of Chongqing city was selected as the assessment units. Based on the land use data in 2010 and 2020 as two periods, ArcGIS 10.8 and Excel software were used to calculate landscape pattern indices. The weights for each index were determined by entropy weight method, and an ecological risk assessment model was constructed, which was used to reveal the temporal and spatial change characteristics of ecological risk. Based on the principle of mathematical statistics, the correlation analysis between cultivated land ecological risk and cultivated land slope was carried out, which aimed to explore the relationship between cultivated land ecological risk and cultivated land slope. [Results and Discussions] Comparing to 2010, patch density (PD), division (D), fractal dimension (FD), and edge density (ED) of cultivated land all decreased in 2020, while meant Patch Size (MPS) increased, indicating an increase in the contiguity of cultivated land. The mean shape index (MSI) of cultivated land increased, indicating that the shape of cultivated land tended to be complicated. The landscape disturbance index (U) decreased from 0.97 to 0.94, indicating that the overall resistance to disturbances in cultivated land has increased. The landscape vulnerability index (V) increased from 2.96 to 3.20, indicating that the structure of cultivated land become more fragile. The ecological risk value of cultivated land decreased from 3.10 to 3.01, indicating the farmland consolidation for agricultural mechanization effectively improved the landscape pattern of cultivated land and enhanced the safety of the agricultural ecosystem. During the two periods, the ecological risk areas were primarily composed of low-risk and relatively low-risk zones. The area of low-risk zones increased by 6.44%, mainly expanding towards the northern part, while the area of relatively low-risk zones increased by 6.17%, primarily spreading towards the central-eastern and southeastern part. The area of moderate-risk zones increased by 24.4%, mainly extending towards the western and northwestern part, while the area of relatively high-risk zones decreased by 60.70%, with some new additions spreading towards the northeastern part. The area of high-risk zones increased by 16.30%, with some new additions extending towards the northwest part. Overall, the ecological safety zones of cultivated relatively increased. The cultivated land slope was primarily concentrated in the range of 2° to 25°. On the one hand, when the cultivated land slope was less than 15°, the proportion of the slope area was negatively correlated with the ecological risk value. On the other hand, when the slope was above 15°, the proportion of the slope area was positively correlated with the ecological risk value. In 2010, there was a highly significant correlation between the proportion of slope area and ecological risk value for cultivated land slope within the ranges of 5° to 8°, 15° to 25°, and above 25°, with corresponding correlation coefficients of 0.592, 0.609, and 0.849, respectively. In 2020, there was a highly significant correlation between the proportion of slope area and ecological risk value for cultivated land slope within the ranges of 2° to 5°, 5° to 8°, 15° to 25°, and above 25°, with corresponding correlation coefficients of 0.534, 0.667, 0.729, and 0.839, respectively. [Conclusions] The assessment of cultivated land ecological risk in Tongnan district of Chongqing city before and after the farmland consolidation for agricultural mechanization, as well as the analysis of the correlation between ecological risk and cultivated land slope, demonstrate that the farmland consolidation for agricultural mechanization can reduce cultivated land ecological risk, and the proportion of cultivated land slope can be an important basis for precision guidance in the farmland consolidation for agricultural mechanization. Considering the occurrence of moderate sheet erosion from a slope of 5° and intense erosion from a slope of 10° to 15°, and taking into account the reduction of ecological risk value and the actual topographic conditions, the subsequent farmland consolidation for agricultural mechanization in Tongnan district should focus on areas with cultivated land slope ranging from 5° to 8° and 15° to 25°.
[Objective] The technology of multi-machine convoy driving has emerged as a focal point in the field of agricultural mechanization. By organizing multiple agricultural machinery units into convoys, unified control and collaborative operations can be achieved. This not only enhances operational efficiency and reduces costs, but also minimizes human labor input, thereby maximizing the operational potential of agricultural machinery. In order to solve the problem of communication delay in cooperative control of multi-vehicle formation and its compensation strategy, the trajectory control method of multi-vehicle formation was proposed based on model predictive control (MPC) delay compensator. [Methods] The multi-vehicle formation cooperative control strategy was designed, which introduced the four-vehicle formation cooperative scenario in three lanes, and then introduced the design of the multi-vehicle formation cooperative control architecture, which was respectively enough to establish the kinematics and dynamics model and equations of the agricultural machine model, and laied down a sturdy foundation for solving the formation following problem later. The testing and optimization of automatic driving algorithms based on real vehicles need to invest too much time and economic costs, and were subject to the difficulties of laws and regulations, scene reproduction and safety, etc. Simulation platform testing could effectively solve the above question. For the agricultural automatic driving multi-machine formation scenarios, the joint simulation platform Carsim and Simulink were used to simulate and validate the formation driving control of agricultural machines. Based on the single-machine dynamics model of the agricultural machine, a delay compensation controller based on MPC was designed. Feedback correction first detected the actual output of the object and then corrected the model-based predicted output with the actual output and performed a new optimization. Based on the above model, the nonlinear system of kinematics and dynamics was linearized and discretized in order to ensure the real-time solution. The objective function was designed so that the agricultural machine tracks on the desired trajectory as much as possible. And because the operation range of the actuator was limited, the control increment and control volume were designed with corresponding constraints. Finally, the control increment constraints were solved based on the front wheel angle constraints, front wheel angle increments, and control volume constraints of the agricultural machine. [Results and Discussions] Carsim and MATLAB/Simulink could be effectively compatible, enabling joint simulation of software with external solvers. When the delay step size d=5 was applied with delay compensation, the MPC response was faster and smoother; the speed error curve responded faster and gradually stabilized to zero error without oscillations. Vehicle 1 effectively changed lanes in a short time and maintains the same lane as the lead vehicle. In the case of a longer delay step size d =10, controllers without delay compensation showed more significant performance degradation. Even under higher delay conditions, MPC with delay compensation applied could still quickly respond with speed error and longitudinal acceleration gradually stabilizing to zero error, avoiding oscillations. The trajectory of Vehicle 1 indicated that the effectiveness of the delay compensation mechanism decreased under extreme delay conditions. The simulation results validated the effectiveness of the proposed formation control algorithm, ensuring that multiple vehicles could successfully change lanes to form queues while maintaining specific distances and speeds. Furthermore, the communication delay compensation control algorithm enables vehicles with added delay to effectively complete formation tasks, achieving stable longitudinal and lateral control. This confirmed the feasibility of the model predictive controller with delay compensation proposed. [Conclusions] At present, most of the multi-machine formation coordination is based on simulation platform for verification, which has the advantages of safety, economy, speed and other aspects, however, there's still a certain gap between the idealized model in the simulation platform and the real machine experiment. Therefore, multi-machine formation operation of agricultural equipment still needs to be tested on real machines under sound laws and regulations.
[Objective] The traditional predictive control approach usually employs a fixed time horizon and often overlooks the impact of changes in curvature and road bends. This oversight leads to subpar tracking performance and inadequate adaptability of robots for navigating curves and paths. Although extending the time horizon of the standard fixed time horizon model predictive control (MPC) can improve curve path tracking accuracy, it comes with high computational costs, making it impractical in situations with restricted computing resources. Consequently, an adaptive time horizon MPC controller was developed to meet the requirements of complex tasks such as autonomous mowing. [Methods] Initially, it was crucial to establish a kinematic model for the mowing robot, which required employing Taylor linearization and Euler method discretization techniques to ensure accurate path tracking. The prediction equation for the error model was derived after conducting a comprehensive analysis of the robot's kinematics model employed in mowing. Second, the size of the previewing area was determined by utilizing the speed data and reference path information gathered from the mowing robot. The region located a certain distance ahead of the robot's current position, was identified to as the preview region, enabling a more accurate prediction of the robot's future traveling conditions. Calculations for both the curve factor and curve change factor were carried out within this preview region. The curvature factor represented the initial curvature of the path, while the curvature change factor indicated the extent of curvature variation in this region. These two variables were then fed into a fuzzy controller, which adjusted the prediction time horizon of the MPC. The integration enabled the mowing robot to promptly adjust to changes in the path's curvature, thereby improving its accuracy in tracking the desired trajectory. Additionally, a novel technique for triggering MPC execution was developed to reduce computational load and improve real-time performance. This approach ensured that MPC activation occurred only when needed, rather than at every time step, resulting in reduced computational expenses especially during periods of smooth robot motion where unnecessary computation overhead could be minimized. By meeting kinematic and dynamic constraints, the optimization algorithm successfully identified an optimal control sequence, ultimately enhancing stability and reliability of the control system. Consequently, these set of control algorithms facilitated precise path tracking while considering both kinematic and dynamic limitations in complex environments. [Results and Discussion] The adaptive time-horizon MPC controller effectively limited the maximum absolute heading error and maximum absolute lateral error to within 0.13 rad and 11 cm, respectively, surpassing the performance of the MPC controller in the control group. Moreover, compared to both the first and fourth groups, the adaptive time-horizon MPC controller achieved a remarkable reduction of 75.39% and 57.83% in mean values for lateral error and heading error, respectively (38.38% and 31.84%, respectively). Additionally, it demonstrated superior tracking accuracy as evidenced by its significantly smaller absolute standard deviation of lateral error (0.025 6 m) and course error (0.025 5 rad), outperforming all four fixed time-horizon MPC controllers tested in the study. Furthermore, this adaptive approach ensured precise tracking and control capabilities for the mowing robot while maintaining a remarkably low average solution time of only 0.004 9 s, notably faster than that observed with other control data sets-reducing computational load by approximately 10.9 ms compared to maximum time-horizon MPC. [Conclusions] The experimental results demonstrated that the adaptive time-horizon MPC tracking approach effectively addressed the trade-off between control accuracy and computational complexity encountered in fixed time-horizon MPC. By dynamically adjusting the time horizon length the and performing MPC calculations based on individual events, this approach can more effectively handle scenarios with restricted computational resources, ensuring superior control precision and stability. Furthermore, it achieves a balance between control precision and real-time performance in curve route tracking for mowing robots, offering a more practical and reliable solution for their practical application.
[Objective] High-precision localization technology serves as the crucial foundation in enabling the autonomous navigation operations of intelligent agricultural robots. However, the traditional global navigation satellite system (GNSS) localization method faces numerous limitations, such as tree shadow, electromagnetic interference, and other factors in the agricultural environment brings challenges to the accuracy and reliability of localization technology. To address the deficiencies and achieve precise localization of agricultural robots independent of GNSS, a localization method was proposed based on the fusion of three-dimensional light detection and ranging (LiDAR) data and inertial measurement unit (IMU) information to enhance localization accuracy and reliability. [Methods] LiDAR was used to obtain point cloud data in the agricultural environment and realize self-localization via point cloud matching. By integrating real-time motion parameter measurements from the IMU with LiDAR data, a high-precision localization solution for agricultural robots was achieved through a specific fusion algorithm. Firstly, the LiDAR-obtained point cloud data was preprocessed and the depth map was used to save the data. This approach could reduce the dimensionality of the original LiDAR point cloud, and eliminate the disorder of the original LiDAR point cloud arrangement, facilitating traversal and clustering through graph search. Given the presence of numerous distinct crops like trees in the agricultural environment, an angle-based clustering method was adopted. Specific angle-based clustering criteria were set to group the point cloud data, leading to the segmentation of different clusters of points, and obvious crops in the agricultural environment was effectively perceived. Furthermore, to improve the accuracy and stability of positioning, an improved three-dimensional normal distribution transform (3D-NDT) localization algorithm was proposed. This algorithm operated by matching the LiDAR-scanned point cloud data in real time with the pre-existing down sampled point cloud map to achieve real-time localization. Considering that direct down sampling of LiDAR point clouds in the agricultural environment could result in the loss of crucial environmental data, a point cloud clustering operation was used in place of down sampling operation, thereby improving matching accuracy and positioning precision. Secondly, to address potential constraints and shortcomings of using a single sensor for robot localization, a multi-sensor information fusion strategy was deployed to improve the localization accuracy. Specifically, the extended Kalman filter algorithm (EKF) was chosen to fuse the localization data from LiDAR point cloud and the IMU odometer information. The IMU provided essential motion parameters such as acceleration and angular velocity of the agricultural robot, and by combining with the LiDAR-derived localization information, the localization of the agricultural robot could be more accurately estimated. This fusion approach maximized the advantages of different sensors, compensated for their individual limitations, and improved the overall localization accuracy of the agricultural robot. [Results and Discussions] A series of experimental results in the Gazebo simulation environment of the robot operating system (ROS) and real operation scenarios showed that the fusion localization method proposed had significant advantages. In the simulation environment, the average localization errors of the proposed multi-sensor data fusion localization method were 1.7 and 1.8 cm, respectively, while in the experimental scenario, these errors were 3.3 and 3.3 cm, respectively, which were significantly better than the traditional 3D-NDT localization algorithm. These findings showed that the localization method proposed in this study could achieve high-precision localization in the complex agricultural environment, and provide reliable localization assistance for the autonomous functioning of agricultural robots. [Conclusions] The proposed localization method based on the fusion of LiDAR data and IMU information provided a novel localization solution for the autonomous operation of agricultural robots in areas with limited GNSS reception. Through the comprehensive utilization of multi-sensor information and adopting advanced data processing and fusion algorithms, the localization accuracy of agricultural robots could be significantly improved, which could provide a new reference for the intelligence and automation of agricultural production.
[Objective] The pneumatic rapeseed seeder easily inhales the dust generated during live broadcast operations into the seeding pipe, and then releases it along with the rapeseed. Consequently, when monitoring the rapeseed flow, dust interference can affect the flow, detection sensitivity and accuracy. This research aims to develop a dust removal rapeseed seeding rate monitoring system suitable for air-assisted pneumatic rapeseed seeders to improve the transparency and intelligence of the sowing process. [Methods] The monitoring system comprised a dust removal rapeseed seed flow detection device and a sowing monitoring terminal, which could be adjusted for different seeders widths by altering the number of detection devices. The rapeseed stream sensing structure operated on the principle of photoelectric induction. A delicate light layer was generated using an LED light source and a narrow slit structure. The convex lens condenses the light and directed it onto the sensing area of the silicon photovoltaic cell. When the rapeseed seeds pass through the sensing light layer, the silicon photovoltaic cell produced a voltage change signal. The signal converts it into a pulse signal that can be recognized by the microcontroller. A dust removal mechanism was designed by analyzing dust sources in the seeding system during normal field operation of the air-assisted rapeseed seeding machine and understanding the impact mechanism of the dust detection device on the accuracy of rapeseed flow monitoring. This mechanism employed a transparent plate to protect the photoelectric induction device in a relatively enclosed space and used a stepper motor screw mechanism to generate friction between the transparent plate and the dust removal cloth for effective dust removal. The appropriate size of the dust shield was determined by comparing its movement stroke with other structural dimensions of the detection device. The relationship between the silicon photocells voltage and detection accuracy was established through experiments at seeding frequencies of 10‒40 Hz. To ensure that the real-time detection accuracy was not less than 90%, the dust removal control threshold was set at 82% of the initial voltage value. In order to prevent congestion and data loss during data transmission and improve the scalability and compatibility of the monitoring system, data transmission between the detection device and the monitoring terminal was implemented based on the CAN2.0A communication protocol. The structural framework and monitoring terminal functions of the rapeseed sowing monitoring system were outlined. Software functions of the detection device were designed to meet the dust removal, communication, and rapeseed flow detection needs. The program execution process of the detection device was explained. In order to provide data support for the dust flow rate that should be controlled at various seeding frequencies during the bench test, experiments were conducted in the field to obtain theoretical data. [Results and Discussions] The comparison bench test of the detection device indicates that with the average seeding frequency ranging from 12.4 to 36.3 Hz and the average dust flow rate ranging from 252 to 386 mg/s, the detection accuracy after two dust removal cycles without a dustproof and dust removal detection device was not higher than 80.2%. The dust detection device with dust removal got an accuracy rate of not less than 90.2%, and the average detection accuracy rate within a single dust removal cycle was not less than 93.6%. The seeding amount monitoring bench test showed that when the average seeding frequency was no higher than 37.6 Hz, the seeding rate monitoring accuracy was not less than 92.2%. Furthermore, the field sowing experiment results demonstrated that at a normal operating speed (2.8‒4.6 km/h) of the rapeseed direct seeder, with a field sowing frequency of 14.8‒31.1 Hz, the accuracy of sowing quantity monitoring was not less than 93.1%. [Conclusions] The rapeseed sowing quality monitoring system provides effective support for precise detection even when operating in dusty conditions with the pneumatic rapeseed direct seeder. In the future, by integrating positioning data, sowing information, and fertilization monitoring data through CAN bus technology, a comprehensive field sowing and fertilization status map can be created to further enhance the system's capabilities.
[Objective] The identification of agricultural activities plays a crucial role for greenhouse vegetables production, particularly in the precise management of cucumber cultivation. By monitoring and analyzing the timing and procedures of agricultural operations, effective guidance can be provided for agricultural production, leading to increased crop yield and quality. However, in practical applications, the recognition of agricultural activities in cucumber cultivation faces significant challenges. The complex and ever-changing growing environment of cucumbers, including dense foliage and internal facility structures that may obstruct visibility, poses difficulties in recognizing agricultural activities. Additionally, agricultural tasks involve various stages such as planting, irrigation, fertilization, and pruning, each with specific operational intricacies and skill requirements. This requires the recognition system to accurately capture the characteristics of various complex movements to ensure the accuracy and reliability of the entire recognition process. To address the complex challenges, an innovative algorithm: SlowFast-SMC-ECA (SlowFast-Spatio-Temporal Excitation, Channel Excitation, Motion Excitation-Efficient Channel Attention) was proposed for the recognition of agricultural activity behaviors in cucumber cultivation within facilities. [Methods] This algorithm represents a significant enhancement to the traditional SlowFast model, with the goal of more accurately capturing hand motion features and crucial dynamic information in agricultural activities. The fundamental concept of the SlowFast model involved processing video streams through two distinct pathways: the Slow Pathway concentrated on capturing spatial detail information, while the Fast Pathway emphasized capturing temporal changes in rapid movements. To further improve information exchange between the Slow and Fast pathways, lateral connections were incorporated at each stage. Building upon this foundation, the study introduced innovative enhancements to both pathways, improving the overall performance of the model. In the Fast Pathway, a multi-path residual network (SMC) concept was introduced, incorporating convolutional layers between different channels to strengthen temporal interconnectivity. This design enabled the algorithm to sensitively detect subtle temporal variations in rapid movements, thereby enhancing the recognition capability for swift agricultural actions. Meanwhile, in the Slow Pathway, the traditional residual block was replaced with the ECA-Res structure, integrating an effective channel attention mechanism (ECA) to improve the model's capacity to capture channel information. The adaptive adjustment of channel weights by the ECA-Res structure enriched feature expression and differentiation, enhancing the model's understanding and grasp of key spatial information in agricultural activities. Furthermore, to address the challenge of class imbalance in practical scenarios, a balanced loss function (Smoothing Loss) was developed. By introducing regularization coefficients, this loss function could automatically adjust the weights of different categories during training, effectively mitigating the impact of class imbalance and ensuring improved recognition performance across all categories. [Results and Discussions] The experimental results significantly demonstrated the outstanding performance of the improved SlowFast-SMC-ECA model on a specially constructed agricultural activity dataset. Specifically, the model achieved an average recognition accuracy of 80.47%, representing an improvement of approximately 3.5% compared to the original SlowFast model. This achievement highlighted the effectiveness of the proposed improvements. Further ablation studies revealed that replacing traditional residual blocks with the multi-path residual network (SMC) and ECA-Res structures in the second and third stages of the SlowFast model leads to superior results. This highlighted that the improvements made to the Fast Pathway and Slow Pathway played a crucial role in enhancing the model's ability to capture details of agricultural activities. Additional ablation studies also confirmed the significant impact of these two improvements on improving the accuracy of agricultural activity recognition. Compared to existing algorithms, the improved SlowFast-SMC-ECA model exhibited a clear advantage in prediction accuracy. This not only validated the potential application of the proposed model in agricultural activity recognition but also provided strong technical support for the advancement of precision agriculture technology. In conclusion, through careful refinement and optimization of the SlowFast model, it was successfully enhanced the model's recognition capabilities in complex agricultural scenarios, contributing valuable technological advancements to precision management in greenhouse cucumber cultivation. [Conclusions] By introducing advanced recognition technologies and intelligent algorithms, this study enhances the accuracy and efficiency of monitoring agricultural activities, assists farmers and agricultural experts in managing and guiding the operational processes within planting facilities more efficiently. Moreover, the research outcomes are of immense value in improving the traceability system for agricultural product quality and safety, ensuring the reliability and transparency of agricultural product quality.
[Objective] Kale is an important bulk vegetable crop worldwide, its main growth characteristics are outer leaves and leaf bulbs. The traits of leaf bulb kale are crucial for adjusting water and fertilizer parameters in the field to achieve maximum yield. However, various factors such as soil quality, light exposure, leaf overlap, and shading can affect the growth of in practical field conditions. The similarity in color and texture between leaf bulbs and outer leaves complicates the segmentation process for existing recognition models. In this paper, the segmentation of kale outer leaves and leaf bulbs in complex field background was proposed, using pixel values to determine leaf bulb size for intelligent field management. A semantic segmentation algorithm, UperNet-ESA was proposed to efficiently and accurately segment nodular kale outer leaf and leaf bulb in field scenes using the morphological features of the leaf bulbs and outer leaves of nodular kale to realize the intelligent management of nodular kale in the field. [Methods] The UperNet-ESA semantic segmentation algorithm, which uses the unified perceptual parsing network (UperNet) as an efficient semantic segmentation framework, is more suitable for extracting crop features in complex environments by integrating semantic information across different scales. The backbone network was improved using ConvNeXt, which is responsible for feature extraction in the model. The similarity between kale leaf bulbs and outer leaves, along with issues of leaf overlap affecting accurate target contour localization, posed challenges for the baseline network, leading to low accuracy. ConvNeXt effectively combines the strengths of convolutional neural networks (CNN) and Transformers, using design principles from Swin Transformer and building upon ResNet50 to create a highly effective network structure. The simplicity of the ConvNeXt design not only enhances segmentation accuracy with minimal model complexity, but also positions it as a top performer among CNN architectures. In this study, the ConvNeXt-B version was chosen based on considerations of computational complexity and the background characteristics of the knotweed kale image dataset. To enhance the model's perceptual acuity, block ratios for each stage were set at 3:3:27:3, with corresponding channel numbers of 128, 256, 512 and 1 024, respectively. Given the visual similarity between kale leaf bulbs and outer leaves, a high-efficiency channel attention mechanism was integrated into the backbone network to improve feature extraction in the leaf bulb region. By incorporating attention weights into feature mapping through residual inversion, attention parameters were cyclically trained within each block, resulting in feature maps with attentional weights. This iterative process facilitated the repeated training of attentional parameters and enhanced the capture of global feature information. To address challenges arising from direct pixel addition between up-sampling and local features, potentially leading to misaligned context in feature maps and erroneous classifications at kale leaf boundaries, a feature alignment module and feature selection module were introduced into the feature pyramid network to refine target boundary information extraction and enhance model segmentation accuracy. [Results and Discussions] The UperNet-ESA semantic segmentation model outperforms the current mainstream UNet model, PSPNet model, DeepLabV3+ model in terms of segmentation accuracy, where mIoU and mPA reached 92.45% and 94.32%, respectively, and the inference speed of up to 16.6 frames per second (fps). The mPA values were better than that of the UNet model, PSPNet model, ResNet-50 based, MobilenetV2, and DeepLabV3+ model with Xception as the backbone, showing improvements of 11.52%, 13.56%, 8.68%, 4.31%, and 6.21%, respectively. Similarly, the mIoU exhibited improvements of 12.21%, 13.04%, 10.65%, 3.26% and 7.11% compared to the mIoU of the UNet-based model, PSPNet model, and DeepLabV3+ model based on the ResNet-50, MobilenetV2, and Xception backbones, respectively. This performance enhancement can be attributed to the introduction of the ECA module and the improvement made to the feature pyramid network in this model, which strengthen the judgement of the target features at each stage to obtain effective global contextual information. In addition, although the PSPNet model had the fastest inference speed, the overall accuracy was too low to for developing kale semantic segmentation models. On the contrary, the proposed model exhibited superior inference speed compared to all other network models. [Conclusions] The experimental results showed that the UperNet-ESA semantic segmentation model proposed in this study outperforms the original network in terms of performance. The improved model achieves the best accuracy-speed balance compared to the current mainstream semantic segmentation networks. In the upcoming research, the current model will be further optimized and enhanced, while the kale dataset will be expanded to include a wider range of samples of nodulated kale leaf bulbs. This expansion is intended to provide a more robust and comprehensive theoretical foundation for intelligent kale field management.
[Objective] Camellia oleifera is one of the four major woody oil plants in the world. Diseases is a significant factor leading to the decline in quality of Camellia oleifera and the financial loss of farmers. Among these diseases, anthracnose is a common and severe disease in Camellia oleifera forests, directly impacting yields and production rates. Accurate disease assessment can improve the prevention and control efficiency and safeguarding the farmers' profit. In this study, an improved You Only Look at CoefficienTs (YOLACT) based method was proposed to realize automatic and efficient grading of the severity of Camellia oleifera leaf anthracnose. [Methods] High-resolution images of Camellia oleifera anthracnose leaves were collected using a smartphone at the National Camellia oleifera Seed Base of Jiangxi Academy of Forestry, and finally 975 valid images were retained after a rigorous screening process. Five data enhancement means were applied, and a data set of 5 850 images was constructed finally, which was divided into training, validation, and test sets in a ratio of 7:2:1. For model selection, the Camellia-YOLACT model was proposed based on the YOLACT instance segmentation model, and by introducing improvements such as Swin-Transformer, weighted bi-directional feature pyramid network, and HardSwish activation function. The Swin Transformer was utilized for feature extraction in the backbone network part of YOLACT, leveraging the global receptive field and shift window properties of the self-attention mechanism in the Transformer architecture to enhance feature extraction capabilities. Additionally, a weighted bidirectional feature pyramid network was introduced to fuse feature information from different scales to improve the detection ability of the model for objects at different scales, thereby improving the detection accuracy. Furthermore, to increase the the model's robustness against the noise in the input data, the HardSwish activation function with stronger nonlinear capability was adopted to replace the ReLu activation function of the original model. Since images in natural environments usually have complex background and foreground information, the robustness of HardSwish helped the model better handling these situations and further improving the detection accuracy. With the above improvements, the Camellia-YOLACT model was constructed and experimentally validated by testing the Camellia oleifera anthracnose leaf image dataset. [Results and Discussions] A transfer learning approach was used for experimental validation on the Camellia oleifera anthracnose severity grading dataset, and the results of the ablation experiments showed that the mAP75 of Camellia-YOLACT proposed in this study was 86.8%, mAPall was 78.3%, mAR was 91.6% which were 5.7%, 2.5% and 7.9% higher than YOLACT model. In the comparison experiments, Camellia-YOLACT performed better than Segmenting Objects by Locations (SOLO) in terms of both accuracy and speed, and its detection speed was doubled compared to Mask R-CNN algorithm. Therefore, the Camellia-YOLACT algorithm was suitable in Camellia oleifera gardens for anthracnose real-time segmentation. In order to verify the outdoors detection performance of Camellia-YOLACT model, 36 groups of Camellia oleifera anthracnose grading experiments were conducted. Experimental results showed that the grading correctness of Camellia oleifera anthracnose injection severity reached 94.4%, and the average absolute error of K-value was 1.09%. Therefore, the Camellia-YOLACT model proposed in this study has a better performance on the grading of the severity of Camellia oleifera anthracnose. [Conclusions] The Camellia-YOLACT model proposed got high accuracy in leaf and anthracnose segmentation of Camellia oleifera, on the basis of which it can realize automatic grading of the severity of Camellia oleifera anthracnose. This research could provide technical support for the precise control of Camellia oleifera diseases.
[Objective] As a critical component of agricultural product supply chain management, cold chain logistics demand prediction encounters challenges such as inadequate feature extraction, high nonlinearity of data, and the propensity for algorithms to become trapped in local optima during the digital transformation process. To address these issues and enhance the accuracy of demand prediction, achieve intelligent management of the agricultural product supply chain, a combined forecasting model that integrates grey relational analysis (GRA), the wild horse optimizer (WHO), and temporal convolutional networks (TCN) is proposed in this research. [Methods] Firstly, a cold chain logistics indicator system was established for the data of Zhejiang province, China, spanning the years 2000 to 2020. This system covered four key aspects: the economic scale of agricultural products, logistics transportation, digital technology, and agricultural product supply. Then, the GRA was applied to identify relevant indicators of cold chain logistics for agricultural products in Zhejiang province, with 17 indicators selected that had a correlation degree higher than 0.75. Sliding window technology, a problem-solving approach for data structures and algorithms, suitable for reducing the time complexity of data to a better level and improving the execution efficiency of algorithms, was used to partition the selected indicators. Secondly, the TCN model was employed to extract features of different scales by stacking multiple convolutional layers. Each layer utilized different-sized convolutional kernels to capture features within different time ranges. By utilizing the dilated convolutional module of TCN, temporal and spatial relationships within economic data were effectively mined, considering the temporal characteristics of socio-economic data and logistics information in the agricultural supply chain, and exploring the temporal and spatial features of economic data. Simultaneously, the WHO algorithm was applied to optimize five hyperparameters of the TCN model, including the number of TCN layers, the number of filters, residual blocks, Dense layers, and neurons within the Dense layer. Finally, the optimized GRA-WHO-TCN model was used to extract and analyze features from highly nonlinear multidimensional economic data, ultimately facilitating the prediction of cold chain logistics demand. [Results and Discussions] For comparative analysis of the superiority of the GRA-WHO-TCN model, the 17 selected indicators were input into long short-term memory (LSTM), TCN, WHO-LSTM, and WHO-TCN models. The parameters optimized by the WHO algorithm for the TCN model were set respectively: 2 TCN layer was, 2 residual blocks, 1 dense layer, 60 filters, and 16 neurons in the dense layer. The optimized GRA-WHO-TCN temporal model can effectively extract the temporal and spatial features of multidimensional data, fully explore the implicit relationships among indicator factors, and demonstrating good fitting effects. Compared to GRA-LSTM and GRA-TCN models, the GRA-TCN model exhibited superior performance, with a lower root mean square error of 37.34 and a higher correlation coefficient of 0.91, indicating the advantage of the TCN temporal model in handling complex nonlinear data. Furthermore, the GRA-WHO-LSTM and GRA-WHO-TCN models optimized by the WHO algorithm had improved prediction accuracy and stability compared to GRA-LSTM and GRA-TCN models, illustrating that the WHO algorithm effectively optimized model parameters to enhance the effectiveness of model fitting. When compared to the GRA-WHO-LSTM model, the GRA-WHO-TCN model displayed a lower root mean square error of 11.3 and an effective correlation coefficient of 0.95, predicting cold chain logistics demand quantities in Zhejiang province for the years 2016-2020 as 29.8, 30.46, 24.87, 26.45, and 27.99 million tons, with relative errors within 0.6%, achieving a high level of prediction accuracy. This achievement showcases a high level of prediction accuracy and underscores the utility of the GRA-WHO-TCN model in forecasting complex data scenarios. [Conclusions] The proposed GRA-WHO-TCN model demonstrated superior parameter optimization capabilities and predictive accuracy compared to the GRA-LSTM and GRA-TCN models. The predicted results align well with the development of cold chain logistics of agricultural products in Zhejiang province. This provides a scientific prediction foundation and practical reference value for the development of material flow and information flow in the agricultural supply chain under the digital economy context.
[Significance] The beef cattle industry plays a pivotal role in the development of China's agricultural economy and the enhancement of people's dietary structure. However, there exists a substantial disparity in feeding management practices and economic efficiency of beef cattle industry compared to developed countries. While the beef cattle industry in China is progressing towards intensive, modern, and large-scale development, it encounters challenges such as labor shortage and rising labor costs that seriously affect its healthy development. The determination of animal physiological indicators plays an important role in monitoring animal welfare and health status. Therefore, leveraging data collected from various sensors as well as technologies like machine learning, data mining, and modeling analysis enables automatic acquisition of meaningful information on beef cattle physiological indicators for intelligent management of beef cattle. In this paper, the intelligent monitoring technology of physiological indicators in beef cattle breeding process and its application value are systematically summarized, and the existing challenges and future prospects of intelligent beef cattle breeding process in China are prospected. [Progress] The methods of obtaining information on beef cattle physiological indicators include contact sensors worn on the body and non-contact sensors based on various image acquisitions. Monitoring the exercise behavior of beef cattle plays a crucial role in disease prevention, reproduction monitoring, and status assessment. The three-axis accelerometer sensor, which tracks the amount of time that beef cattle spend on lying, walking, and standing, is a widely used technique for tracking the movement behavior of beef cattle. Through machine vision analysis, individual recognition of beef cattle and identification of standing, lying down, and straddling movements can also be achieved, with the characteristics of non-contact, stress-free, low cost, and generating high data volume. Body temperature in beef cattle is associated with estrus, calving, and overall health. Sensors for monitoring body temperature include rumen temperature sensors and rectal temperature sensors, but there are issues with their inconvenience. Infrared temperature measurement technology can be utilized to detect beef cattle with abnormal temperatures by monitoring eye and ear root temperatures, although the accuracy of the results may be influenced by environmental temperature and monitoring distance, necessitating calibration. Heart rate and respiratory rate in beef cattle are linked to animal diseases, stress, and pest attacks. Monitoring heart rate can be accomplished through photoelectric volume pulse wave measurement and monitoring changes in arterial blood flow using infrared emitters and receivers. Respiratory rate monitoring can be achieved by identifying different nostril temperatures during inhalation and exhalation using thermal infrared imaging technology. The ruminating behavior of beef cattle is associated with health and feed nutrition. Currently, the primary tools used to detect rumination behavior are pressure sensors and three-axis accelerometer sensors positioned at various head positions. Rumen acidosis is a major disease in the rapid fattening process of beef cattle, however, due to limitations in battery life and electrode usage, real-time pH monitoring sensors placed in the rumen are still not widely utilized. Changes in animal physiology, growth, and health can result in alterations in specific components within body fluids. Therefore, monitoring body fluids or surrounding gases through biosensors can be employed to monitor the physiological status of beef cattle. By processing and analyzing the physiological information of beef cattle, indicators such as estrus, calving, feeding, drinking, health conditions, and stress levels can be monitored. This will contribute to the intelligent development of the beef cattle industry and enhance management efficiency. While there has been some progress made in developing technology for monitoring physiological indicators of beef cattle, there are still some challenges that need to be addressed. Contact sensors consume more energy which affects their lifespan. Various sensors are susceptible to environmental interference which affects measurement accuracy. Additionally, due to a wide variety of beef cattle breeds, it is difficult to establish a model database for monitoring physiological indicators under different feeding conditions, breeding stages, and breeds. Furthermore, the installation cost of various intelligent monitoring devices is relatively high, which also limits its utilization coverage. [Conclusion and Prospects] The application of intelligent monitoring technology for beef cattle physiological indicators is highly significance in enhancing the management level of beef cattle feeding. Intelligent monitoring systems and devices are utilized to acquire physiological behavior data, which are then analyzed using corresponding data models or classified through deep learning techniques to promptly monitor subtle changes in physiological indicators. This enables timely detection of sick, estrus, and calving cattle, facilitating prompt measures by production managers, reducing personnel workload, and improving efficiency. The future development of physiological indicators monitoring technologies in beef cattle primarily focuses on the following three aspects: (1) Enhancing the lifespan of contact sensors by reducing energy consumption, decreasing data transmission frequency, and improving battery life. (2) Integrating and analyzing various monitoring data from multiple perspectives to enhance the accuracy and utility value. (3) Strengthening research on non-contact, high-precision and automated analysis technologies to promote the precise and intelligent development within the beef cattle industry.
[Objective] The detection of lameness in dairy cows is an important issue that needs to be solved urgently in the process of large-scale dairy farming. Timely detection and effective intervention can reduce the culling rate of young dairy cows, which has important practical significance for increasing the milk production of dairy cows and improving the economic benefits of pastures. Due to the low efficiency and low degree of automation of traditional manual detection and contact sensor detection, the mainstream cow lameness detection method is mainly based on computer vision. The detection perspective of existing computer vision-based cow lameness detection methods is mainly side view, but the side view perspective has limitations that are difficult to eliminate. In the actual detection process, there are problems such as cows blocking each other and difficulty in deployment. The cow lameness detection method from the top view will not be difficult to use on the farm due to occlusion problems. The aim is to solve the occlusion problem under the side view. [Methods] In order to fully explore the movement undulations of the trunk of the cow and the movement information in the time dimension during the walking process of the cow, a cow lameness detection method was proposed from a top view based on fused spatiotemporal flow features. By analyzing the height changes of the lame cow in the depth video stream during movement, a spatial stream feature image sequence was constructed. By analyzing the instantaneous speed of the lame cow's body moving forward and swaying left and right when walking, optical flow was used to capture the instantaneous speed of the cow's movement, and a time flow characteristic image sequence was constructed. The spatial flow and time flow features were combined to construct a fused spatiotemporal flow feature image sequence. Different from traditional image classification tasks, the image sequence of cows walking includes features in both time and space dimensions. There would be a certain distinction between lame cows and non-lame cows due to their related postures and walking speeds when walking, so using video information analysis was feasible to characterize lameness as a behavior. The video action classification network could effectively model the spatiotemporal information in the input image sequence and output the corresponding category in the predicted result. The attention module Convolutional Block Attention Module (CBAM) was used to improve the PP-TSMv2 video action classification network and build the Cow-TSM cow lameness detection model. The CBAM module could perform channel weighting on different modes of cows, while paying attention to the weights between pixels to improve the model's feature extraction capabilities. Finally, cow lameness experiments were conducted on different modalities, different attention mechanisms, different video action classification networks and comparison of existing methods. The data was used for cow lameness included a total of 180 video streams of cows walking. Each video was decomposed into 100‒400 frames. The ratio of the number of video segments of lame cows and normal cows was 1:1. For the feature extraction of cow lameness from the top view, RGB images had less extractable information, so this work mainly used depth video streams. [Results and Discussions] In this study, a total of 180 segments of cow image sequence data were acquired and processed, including 90 lame cows and 90 non-lame cows with a 1:1 ratio of video segments, and the prediction accuracy of automatic detection method for dairy cow lameness based on fusion of spatiotemporal stream features reaches 88.7%, the model size was 22 M, and the offline inference time was 0.046 s. The prediction accuracy of the common mainstream video action classification models TSM, PP-TSM, SlowFast and TimesFormer models on the data set of automatic detection method for dairy cow lameness based on fusion of spatiotemporal stream features reached 66.7%, 84.8%, 87.1% and 85.7%, respectively. The comprehensive performance of the improved Cow-TSM model in this paper was the most. At the same time, the recognition accuracy of the fused spatiotemporal flow feature image was improved by 12% and 4.1%, respectively, compared with the temporal mode and spatial mode, which proved the effectiveness of spatiotemporal flow fusion in this method. By conducting ablation experiments on different attention mechanisms of SE, SK, CA and CBAM, it was proved that the CBAM attention mechanism used has the best effect on the data of automatic detection method for dairy cow lameness based on fusion of spatiotemporal stream features. The channel attention in CBAM had a better effect on fused spatiotemporal flow data, and the spatial attention could also focus on the key spatial information in cow images. Finally, comparisons were made with existing lameness detection methods, including different methods from side view and top view. Compared with existing methods in the side-view perspective, the prediction accuracy of automatic detection method for dairy cow lameness based on fusion of spatiotemporal stream features was slightly lower, because the side-view perspective had more effective cow lameness characteristics. Compared with the method from the top view, a novel fused spatiotemporal flow feature detection method with better performance and practicability was proposed. [Conclusions] This method can avoid the occlusion problem of detecting lame cows from the side view, and at the same time improves the prediction accuracy of the detection method from the top view. It is of great significance for reducing the incidence of lameness in cows and improving the economic benefits of the pasture, and meets the needs of large-scale construction of the pasture.
[Objective] Real-time monitoring of cow ruminant behavior is of paramount importance for promptly obtaining relevant information about cow health and predicting cow diseases. Currently, various strategies have been proposed for monitoring cow ruminant behavior, including video surveillance, sound recognition, and sensor monitoring methods. However, the application of edge device gives rise to the issue of inadequate real-time performance. To reduce the volume of data transmission and cloud computing workload while achieving real-time monitoring of dairy cow rumination behavior, a real-time monitoring method was proposed for cow ruminant behavior based on edge computing. [Methods] Autonomously designed edge devices were utilized to collect and process six-axis acceleration signals from cows in real-time. Based on these six-axis data, two distinct strategies, federated edge intelligence and split edge intelligence, were investigated for the real-time recognition of cow ruminant behavior. Focused on the real-time recognition method for cow ruminant behavior leveraging federated edge intelligence, the CA-MobileNet v3 network was proposed by enhancing the MobileNet v3 network with a collaborative attention mechanism. Additionally, a federated edge intelligence model was designed utilizing the CA-MobileNet v3 network and the FedAvg federated aggregation algorithm. In the study on split edge intelligence, a split edge intelligence model named MobileNet-LSTM was designed by integrating the MobileNet v3 network with a fusion collaborative attention mechanism and the Bi-LSTM network. [Results and Discussions] Through comparative experiments with MobileNet v3 and MobileNet-LSTM, the federated edge intelligence model based on CA-MobileNet v3 achieved an average Precision rate, Recall rate, F1-Score, Specificity, and Accuracy of 97.1%, 97.9%, 97.5%, 98.3%, and 98.2%, respectively, yielding the best recognition performance. [Conclusions] It is provided a real-time and effective method for monitoring cow ruminant behavior, and the proposed federated edge intelligence model can be applied in practical settings.
[Objective] The monitoring of livestock grazing in natural pastures is a key aspect of the transformation and upgrading of large-scale breeding farms. In order to meet the demand for large-scale farms to achieve accurate real-time detection of a large number of sheep, a high-precision and easy-to-deploy small-target detection model: CSD-YOLOv8s was proposed to realize the real-time detection of small-targeted individual sheep under the high-altitude view of the unmanned aerial vehicle (UAV). [Methods] Firstly, a UAV was used to acquire video data of sheep in natural grassland pastures with different backgrounds and lighting conditions, and together with some public datasets downloaded formed the original image data. The sheep detection dataset was generated through data cleaning and labeling. Secondly, in order to solve the difficult problem of sheep detection caused by dense flocks and mutual occlusion, the SPPFCSPC module was constructed with cross-stage local connection based on the you only look once (YOLO)v8 model, which combined the original features with the output features of the fast spatial pyramid pooling network, fully retained the feature information at different stages of the model, and effectively solved the problem of small targets and serious occlusion of the sheep, and improved the detection performance of the model for small sheep targets. In the Neck part of the model, the convolutional block attention module (CBAM) convolutional attention module was introduced to enhance the feature information capture based on both spatial and channel aspects, suppressing the background information spatially and focusing on the sheep target in the channel, enhancing the network's anti-jamming ability from both channel and spatial dimensions, and improving the model's detection performance of multi-scale sheep under complex backgrounds and different illumination conditions. Finally, in order to improve the real-time and deploy ability of the model, the standard convolution of the Neck network was changed to a lightweight convolutional C2f_DS module with a changeable kernel, which was able to adaptively select the corresponding convolutional kernel for feature extraction according to the input features, and solved the problem of input scale change in the process of sheep detection in a more flexible way, and at the same time, the number of parameters of the model was reduced and the speed of the model was improved. [Results and Discussions] The improved CSD-YOLOv8s model exhibited excellent performance in the sheep detection task. Compared with YOLO, Faster R-CNN and other classical network models, the improved CSD-YOLOv8s model had higher detection accuracy and frames per second (FPS) of 87 f/s in the flock detection task with comparable detection speed and model size. Compared with the YOLOv8s model, Precision was improved from 93.0% to 95.2%, mAP was improved from 91.2% to 93.1%, and it had strong robustness to sheep targets with different degree of occlusion and different scales, which effectively solved the serious problems of missed and misdetection of sheep in the grassland pasture UAV-on-ground sheep detection task due to the small sheep targets, large background noise, and high degree of densification. misdetection serious problems. Validated by the PASCAL VOC 2007 open dataset, the CSD-YOLOv8s model proposed in this study improved the detection accuracy of 20 different objects, including transportation vehicles, animals, etc., especially in sheep detection, the detection accuracy was improved by 9.7%. [Conclusions] This study establishes a sheep dataset based on drone images and proposes a model called CSD-YOLOv8s for detecting grazing sheep in natural grasslands. The model addresses the serious issues of missed detections and false alarms in sheep detection under complex backgrounds and lighting conditions, enabling more accurate detection of grazing livestock in drone images. It achieves precise detection of targets with varying degrees of clustering and occlusion and possesses good real-time performance. This model provides an effective detection method for detecting sheep herds from the perspective of drones in natural pastures and offers technical support for large-scale livestock detection in breeding farms, with wide-ranging potential applications.
[Objective] Currently, pig farming facilities mainly rely on manual counting for tracking slaughtered and stored pigs. This is not only time-consuming and labor-intensive, but also prone to counting errors due to pig movement and potential cheating. As breeding operations expand, the periodic live asset inventories put significant strain on human, material and financial resources. Although methods based on electronic ear tags can assist in pig counting, these ear tags are easy to break and fall off in group housing environments. Most of the existing methods for counting pigs based on computer vision require capturing images from a top-down perspective, necessitating the installation of cameras above each hogpen or even the use of drones, resulting in high installation and maintenance costs. To address the above challenges faced in the group pig counting task, a high-efficiency and low-cost pig counting method was proposed based on improved instance segmentation algorithm and WeChat public platform. [Methods] Firstly, a smartphone was used to collect pig image data in the area from a human view perspective, and each pig's outline in the image was annotated to establish a pig count dataset. The training set contains 606 images and the test set contains 65 images. Secondly, an efficient global attention module was proposed by improving convolutional block attention module (CBAM). The efficient global attention module first performed a dimension permutation operation on the input feature map to obtain the interaction between its channels and spatial dimensions. The permuted features were aggregated using global average pooling (GAP). One-dimensional convolution replaced the fully connected operation in CBAM, eliminating dimensionality reduction and significantly reducing the model's parameter number. This module was integrated into the YOLOv8 single-stage instance segmentation network to build the pig counting model YOLOv8x-Ours. By adding an efficient global attention module into each C2f layer of the YOLOv8 backbone network, the dimensional dependencies and feature information in the image could be extracted more effectively, thereby achieving high-accuracy pig counting. Lastly, with a focus on user experience and outreach, a pig counting WeChat mini program was developed based on the WeChat public platform and Django Web framework. The counting model was deployed to count pigs using images captured by smartphones. [Results and Discussions] Compared with existing methods of Mask R-CNN, YOLACT(Real-time Instance Segmentation), PolarMask, SOLO and YOLOv5x, the proposed pig counting model YOLOv8x-Ours exhibited superior performance in terms of accuracy and stability. Notably, YOLOv8x-Ours achieved the highest accuracy in counting, with errors of less than 2 and 3 pigs on the test set. Specifically, 93.8% of the total test images had counting errors of less than 3 pigs. Compared with the two-stage instance segmentation algorithm Mask R-CNN and the YOLOv8x model that applies the CBAM attention mechanism, YOLOv8x-Ours showed performance improvements of 7.6% and 3%, respectively. And due to the single-stage design and anchor-free architecture of the YOLOv8 model, the processing speed of a single image was only 64 ms, 1/8 of Mask R-CNN. By embedding the model into the WeChat mini program platform, pig counting was conducted using smartphone images. In cases where the model incorrectly detected pigs, users were given the option to click on the erroneous location in the result image to adjust the statistical outcomes, thereby enhancing the accuracy of pig counting. [Conclusions] The feasibility of deep learning technology in the task of pig counting was demonstrated. The proposed method eliminates the need for installing hardware equipment in the breeding area of the pig farm, enabling pig counting to be carried out effortlessly using just a smartphone. Users can promptly spot any errors in the counting results through image segmentation visualization and easily rectify any inaccuracies. This collaborative human-machine model not only reduces the need for extensive manpower but also guarantees the precision and user-friendliness of the counting outcomes.
[Objective] The body size parameter of cattle is a key indicator reflecting the physical development of cattle, and is also a key factor in the cattle selection and breeding process. In order to solve the demand of measuring body size of beef cattle in the complex environment of large-scale beef cattle ranch, an image acquisition device and an automatic measurement algorithm of body size were designed. [Methods] Firstly, the walking channel of the beef cattle was established, and when the beef cattle entered the restraining device through the channel, the RGB and depth maps of the image on the right side of the beef cattle were acquired using the Inter RealSense D455 camera. Secondly, in order to avoid the influence of the complex environmental background, an improved instance segmentation network based on Mask2former was proposed, adding CBAM module and CA module, respectively, to improve the model's ability to extract key features from different perspectives, extracting the foreground contour from the 2D image of the cattle, partitioning the contour, and comparing it with other segmentation algorithms, and using curvature calculation and other mathematical methods to find the required body size measurement points. Thirdly, in the processing of 3D data, in order to solve the problem that the pixel point to be measured in the 2D RGB image was null when it was projected to the corresponding pixel coordinates in the depth-valued image, resulting in the inability to calculate the 3D coordinates of the point, a series of processing was performed on the point cloud data, and a suitable point cloud filtering and point cloud segmentation algorithm was selected to effectively retain the point cloud data of the region of the cattle's body to be measured, and then the depth map was 16. Then the depth map was filled with nulls in the field to retain the integrity of the point cloud in the cattle body region, so that the required measurement points could be found and the 2D data could be returned. Finally, an extraction algorithm was designed to combine 2D and 3D data to project the extracted 2D pixel points into a 3D point cloud, and the camera parameters were used to calculate the world coordinates of the projected points, thus automatically calculating the body measurements of the beef cattle. [Results and Discussions] Firstly, in the part of instance segmentation, compared with the classical Mask R-CNN and the recent instance segmentation networks PointRend and Queryinst, the improved network could extract higher precision and smoother foreground images of cattles in terms of segmentation accuracy and segmentation effect, no matter it was for the case of occlusion or for the case of multiple cattles. Secondly, in three-dimensional data processing, the method proposed in the study could effectively extract the three-dimensional data of the target area. Thirdly, the measurement error of body size was analysed, among the four body size measurement parameters, the smallest average relative error was the height of the cross section, which was due to the more prominent position of the cross section, and the different standing positions of the cattle have less influence on the position of the cross section, and the largest average relative error was the pipe circumference, which was due to the influence of the greater overlap of the two front legs, and the higher requirements for the standing position. Finally, automatic body measurements were carried out on 137 beef cattle in the ranch, and the automatic measurements of the four body measurements parameters were compared with the manual measurements, and the results showed that the average relative errors of body height, cross section height, body slant length, and tube girth were 4.32%, 3.71%, 5.58% and 6.25%, respectively, which met the needs of the ranch. The shortcomings were that fewer body-size parameters were measured, and the error of measuring circumference-type body-size parameters was relatively large. Later studies could use a multi-view approach to increase the number of body rule parameters to be measured and improve the accuracy of the parameters in the circumference category. [Conclusions] The article designed an automatic measurement method based on two-dimensional and three-dimensional contactless body measurements of beef cattle. Moreover, the innovatively proposed method of measuring tube girth has higher accuracy and better implementation compared with the current research on body measurements in beef cattle. The relative average errors of the four body tape parameters meet the needs of pasture measurements and provide theoretical and practical guidance for the automatic measurement of body tape in beef cattle.
[Objective] Nowadays most no contact body size measurement studies are based on point cloud segmentation method, they use a trained point cloud segmentation neural network to segment point cloud of pigs, then locate measurement points based on them. But point cloud segmentation neural network always need a larger graphics processing unit (GPU) memory, moreover, the result of the measurement key point still has room of improvement. This study aims to design a key point generating neural network to extract measurement key points from pig's point cloud. Reducing the GPU memory usage and improve the result of measurement points at the same time, improve both the efficiency and accuracy of the body size measurement. [Methods] A neural network model was proposed using improved Transformer attention mechanic called Pig Back Transformer for generating key points and back orientation points which were related to pig body dimensions. In the first part of the network, it was introduced an embedding structure for initial feature extraction and a Transformer encoder structure with edge attention which was a self-attention mechanic improved from Transformer's encoder. The embedding structure using two shared multilayer perceptron (MLP) and a distance embedding algorithm, it takes a set of points from the edge of pig back's point cloud as input and then extract information from the edge points set. In the encoder part, information about the offset distances between edge points and mass point which were their feature that extracted by the embedding structure mentioned before incorporated. Additionally, an extraction algorithm for back edge point was designed for extracting edge points to generate the input of the neural network model. In the second part of the network, it was proposed a Transformer encoder with improved self-attention called back attention. In the design of back attention, it also had an embedding structure before the encoder structure, this embedding structure extracted features from offset values, these offset values were calculated by the points which are none-edge and down sampled by farthest point sampling (FPS) to both the relative centroid point and model generated global key point from the first part that introduced before. Then these offset values were processed with max pooling with attention generated by the extracted features of the points' axis to extract more information that the original Transformer encoder couldn't extract with the same number of parameters. The output part of the model was designed to generate a set of offsets of the key points and points for back direction fitting, than add the set offset to the global key point to get points for pig body measurements. At last, it was introduced the methods for calculating body dimensions which were length, height, shoulder width, abdomen width, hip width, chest circumference and abdomen circumference using key points and back direction fitting points. [Results and Discussions] In the task of generating key points and points for back direction fitting, the improved Pig Back Transformer performed the best in the accuracy wise in the models tested with the same size of parameters, and the back orientation points generated by the model were evenly distributed which was a good preparation for a better body length calculation. A melting test for edge detection part with two attention mechanic and edge trim method both introduced above had being done, when the edge detection and the attention mechanic got cut off, the result had been highly impact, it made the model couldn't perform as well as before, when the edge trim method of preprocessing part had been cut off, there's a moderate impact on the trained model, but it made the loss of the model more inconsistence while training than before. When comparing the body measurement algorithm with human handy results, the relative error in length was 0.63%, which was an improvement compared to other models. On the other hand, the relative error of shoulder width, abdomen width and hip width had edged other models a little but there was no significant improvement so the performance of these measurement accuracy could be considered negligible, the relative error of chest circumference and abdomen circumference were a little bit behind by the other methods existed, it's because the calculate method of circumferences were not complicated enough to cover the edge case in the dataset which were those point cloud that have big holes in the bottom of abdomen and chest, it impacted the result a lot. [Conclusions] The improved Pig Back Transformer demonstrates higher accuracy in generating key points and is more resource-efficient, enabling the calculation of more accurate pig body measurements. And provides a new perspective for non-contact livestock body size measurements.
[Objective] There exist a high genetic correlation among various morphological characteristics of Mongolian horses. Utilizing advanced technology to obtain body structure parameters related to athletic performance could provide data support for breeding institutions to develop scientific breeding plans and establish the groundwork for further improvement of Mongolian horse breeds. However, traditional manual measurement methods are time-consuming, labor-intensive, and may cause certain stress responses in horses. Therefore, ensuring precise and effective measurement of Mongolian horse body dimensions is crucial for formulating early breeding plans. [Method] Video images of 50 adult Mongolian horses in the suitable breeding stage at the Inner Mongolia Agricultural University Horse Breeding Technical Center was first collected. Fifty images per horse were captured to construct the training and validation sets, resulting in a total of 2 500 high-definition RGB images of Mongolian horses, with an equal ratio of images depicting horses in motion and at rest. To ensure the model's robustness and considering issues such as angles, lighting, and image blurring during actual image capture, a series of enhancement algorithms were applied to the original dataset, expanding the Mongolian horse image dataset to 4 000 images. The YOLOv8n-pose was employed as the foundational keypoint detection model. Through the design of the C2f_DCN module, deformable convolution (DCNV2) was integrated into the C2f module of the Backbone network to enhance the model's adaptability to different horse poses in real-world scenes. Besides, an SA attention module was added to the Neck network to improve the model's focus on critical features. The original loss function was replaced with SCYLLA-IoU (SIoU) to prioritize major image regions, and a cosine annealing method was employed to dynamically adjust the learning rate during model training. The improved model was named DSS-YOLO (DCNv2-SA-SIoU-YOLO) network model. Additionally, a test set comprising 30 RGB-D images of mature Mongolian horses was selected for constructing body dimension measurement tasks. DSS-YOLO was used for keypoint detection of body dimensions. The 2D keypoint coordinates from RGB images were fused with corresponding depth values from depth images to obtain 3D keypoint coordinates, and Mongolian horse's point cloud information was transformed. Point cloud processing and analysis were performed using pass-through filtering, random sample consensus (RANSAC) shape fitting, statistical outlier filtering, and principal component analysis (PCA) coordinate system correction. Finally, body height, body oblique length, croup height, chest circumference, and croup circumference were automatically computed based on keypoint spatial coordinates. [Results and Discussion] The proposed DSS-YOLO model exhibited parameter and computational costs of 3.48 M and 9.1 G, respectively, with an average accuracy mAP0.5:0.95 reaching 92.5%, and a dDSS of 7.2 pixels. Compared to Hourglass, HRNet, and SimCC, mAP0.5:0.95 increased by 3.6%, 2.8%, and 1.6%, respectively. By relying on keypoint coordinates for automatic calculation of body dimensions and suggesting the use of a mobile least squares curve fitting method to complete the horse's hip point cloud, experiments involving 30 Mongolian horses showed a mean average error (MAE) of 3.77 cm and mean relative error (MRE) of 2.29% in automatic measurements. [Conclusions] The results of this study showed that DSS-YOLO model combined with three-dimensional point cloud processing methods can achieve automatic measurement of Mongolian horse body dimensions with high accuracy. The proposed measurement method can also be extended to different breeds of horses, providing technical support for horse breeding plans and possessing practical application value.
[Objective] Manual disinfection in large-scale sheep farm is laborious, time-consuming, and often results in incomplete coverage and inadequate disinfection. With the rapid development of the application of artificial intelligence and automation technology, the automatic navigation and spraying robot for livestock and poultry breeding, has become a research hotspot. To maintain shed hygiene and ensure sheep health, an automatic navigation and spraying robot was proposed for sheep sheds. [Methods] The automatic navigation and spraying robot was designed with a focus on three aspects: hardware, semantic segmentation model, and control algorithm. In terms of hardware, it consisted of a tracked chassis, cameras, and a collapsible spraying device. For the semantic segmentation model, enhancements were made to the lightweight semantic segmentation model ENet, including the addition of residual structures to prevent network degradation and the incorporation of a squeeze-and-excitation network (SENet) attention mechanism in the initialization module. This helped to capture global features when feature map resolution was high, addressing precision issues. The original 6-layer ENet network was reduced to 5 layers to balance the encoder and decoder. Drawing inspiration from dilated spatial pyramid pooling, a context convolution module (CCM) was introduced to improve scene understanding. A criss-cross attention (CCA) mechanism was adapted to acquire context global features of different scales without cascading, reducing information loss. This led to the development of a double attention enet (DAENet) semantic segmentation model was proposed to achieve real-time and accurate segmentation of sheep shed surfaces. Regarding control algorithms, a method was devised to address the robot's difficulty in controlling its direction at junctions. Lane recognition and lane center point identification algorithms were proposed to identify and mark navigation points during the robot's movement outside the sheep shed by simulating real roads. Two cameras were employed, and a camera switching algorithm was developed to enable seamless switching between them while also controlling the spraying device. Additionally, a novel offset and velocity calculation algorithm was proposed to control the speeds of the robot's left and right tracks, enabling control over the robot's movement, stopping, and turning. [Results and Discussions] The DAENet model achieved a mean intersection over union (mIoU) of 0.945 3 in image segmentation tasks, meeting the required segmentation accuracy. During testing of the camera switching algorithm, it was observed that the time taken for the complete transition from camera to spraying device action does not exceed 15 seconds when road conditions changed. Testing of the center point and offset calculation algorithm revealed that, when processing multiple frames of video streams, the algorithm averages 0.04 to 0.055 per frame, achieving frame rates of 20 to 24 frames per second, meeting real-time operational requirements. In field experiments conducted in sheep farm, the robot successfully completed automatic navigation and spraying tasks in two sheds without colliding with roadside troughs. The deviation from the road and lane centerlines did not exceed 0.3 meters. Operating at a travel speed of 0.2 m/s, the liquid in the medicine tank was adequate to complete the spraying tasks for two sheds. Additionally, the time taken for the complete transition from camera to spraying device action did not exceed 15 when road conditions changed. The robot maintained an average frame rate of 22.4 frames per second during operation, meeting the experimental requirements for accurate and real-time information processing. Observation indicated that the spraying coverage rate of the robot exceeds 90%, meeting the experimental coverage requirements. [Conclusions] The proposed automatic navigation and spraying robot, based on the DAENet semantic segmentation model and center point recognition algorithm, combined with hardware design and control algorithms, achieves comprehensive disinfection within sheep sheds while ensuring safety and real-time operation.
[Significance] Building the agricultural new quality productivity is of great significance. It is the advanced quality productivity which realizes the transformation, upgrading, and deep integration of substantive, penetrating, operational, and media factors, and has outstanding characteristics such as intelligence, greenness, integration, and organization. As a new technology revolution in the field of agriculture, smart agricultural technology transforms agricultural production mode by integrating agricultural biotechnology, agricultural information technology, and smart agricultural machinery and equipment, with information and knowledge as important core elements. The inherent characteristics of "high-tech, high-efficiency, high-quality, and sustainable" in agricultural new quality productivity are fully reflected in the practice of smart agricultural technology innovation. And it has become an important core and engine for promoting the agricultural new quality productivity. [Progress] Through literature review and theoretical analysis, this article conducts a systematic study on the practical foundation, internal logic, and problem challenges of smart agricultural technology innovation leading the development of agricultural new quality productivity. The conclusions show that: (1) At present, the global innovation capability of smart agriculture technology is constantly enhancing, and significant technology breakthroughs have been made in fields such as smart breeding, agricultural information perception, agricultural big data and artificial intelligence, smart agricultural machinery and equipment, providing practical foundation support for leading the development of agricultural new quality productivity. Among them, the smart breeding of 'Phenotype+Genotype+Environmental type' has entered the fast lane, the technology system for sensing agricultural sky, air, and land information is gradually maturing, the research and exploration on agricultural big data and intelligent decision-making technology continue to advance, and the creation of smart agricultural machinery and equipment for different fields has achieved fruitful results; (2) Smart agricultural technology innovation provides basic resources for the development of agricultural new quality productivity through empowering agricultural factor innovation, provides sustainable driving force for the development of agricultural new quality productivity through empowering agricultural technology innovation, provides practical paradigms for the development of agricultural new quality productivity through empowering agricultural scenario innovation, provides intellectual support for the development of agricultural new quality productivity through empowering agricultural entity innovation, and provides important guidelines for the development of agricultural new quality productivity through empowering agricultural value innovation; (3) Compared to the development requirements of agricultural new quality productivity in China and the advanced level of international smart agriculture technology, China's smart agriculture technology innovation is generally in the initial stage of multi-point breakthroughs, system integration, and commercial application. It still faces major challenges such as an incomplete policy system for technology innovation, key technologies with bottlenecks, blockages and breakpoints, difficulties in the transformation and implementation of technology achievements, and incomplete support systems for technology innovation. [Conclusions and Prospects] Regarding the issue of technology innovation in smart agriculture, this article proposes the 'Four Highs' path of smart agriculture technology innovation to fill the gaps in smart agriculture technology innovation and accelerate the formation of agricultural new quality productivity in China. The "Four Highs" path specifically includes the construction of high-energy smart agricultural technology innovation platforms, the breakthroughs in high-precision and cutting-edge smart agricultural technology products, the creation of high-level smart agricultural application scenarios, and the cultivation of high-level smart agricultural innovation talents. Finally, this article proposes four strategic suggestions such as deepening the understanding of smart agriculture technology innovation and agricultural new quality productivity, optimizing the supply of smart agriculture technology innovation policies, building a national smart agriculture innovation development pilot zone, and improving the smart agriculture technology innovation ecosystem.
[Objective] The enumeration of wheat leaves is an essential indicator for evaluating the vegetative state of wheat and predicting its yield potential. Currently, the process of wheat leaf counting in field settings is predominantly manual, characterized by being both time-consuming and labor-intensive. Despite advancements, the efficiency and accuracy of existing automated detection and counting methodologies have yet to satisfy the stringent demands of practical agricultural applications. This study aims to develop a method for the rapid quantification of wheat leaves to refine the precision of wheat leaf tip detection. [Methods] To enhance the accuracy of wheat leaf detection, firstly, an image dataset of wheat leaves across various developmental stages—seedling, tillering, and overwintering—under two distinct lighting conditions and using visible light images sourced from both mobile devices and field camera equipmen, was constructed. Considering the robust feature extraction and multi-scale feature fusion capabilities of YOLOv8 network, the foundational architecture of the proposed model was based on the YOLOv8, to which a coordinate attention mechanism has been integrated. To expedite the model's convergence, the loss functions were optimized. Furthermore, a dedicated small object detection layer was introduced to refine the recognition of wheat leaf tips, which were typically difficult for conventional models to discern due to their small size and resemblance to background elements. This deep learning network was named as YOLOv8-CSD, tailored for the recognition of small targets such as wheat leaf tips, ascertains the leaf count by detecting the number of leaf tips present within the image. A comparative analysis was conducted on the YOLOv8-CSD model in comparison with the original YOLOv8 and six other prominent network architectures, including Faster R-CNN, Mask R-CNN, YOLOv7, and SSD, within a uniform training framework, to evaluate the model's effectiveness. In parallel, the performance of both the original and YOLOv8-CSD models was assessed under challenging conditions, such as the presence of weeds, occlusions, and fluctuating lighting, to emulate complex real-world scenarios. Ultimately, the YOLOv8-CSD model was deployed for wheat leaf number detection in intricate field conditions to confirm its practical applicability and generalization potential. [Results and Discussions] The research presented a methodology that achieved a recognition precision of 91.6% and an mAP0.5 of 85.1% for wheat leaf tips, indicative of its robust detection capabilities. This method exceled in adaptability within complex field environments, featuring an autonomous adjustment mechanism for different lighting conditions, which significantly enhanced the model's robustness. The minimal rate of missed detections in wheat seedlings' leaf counting underscored the method's suitability for wheat leaf tip recognition in intricate field scenarios, consequently elevating the precision of wheat leaf number detection. The sophisticated algorithm embedded within this model had demonstrated a heightened capacity to discern and focus on the unique features of wheat leaf tips during the detection process. This capability was essential for overcoming challenges such as small target sizes, similar background textures, and the intricacies of feature extraction. The model's consistent performance across diverse conditions, including scenarios with weeds, occlusions, and fluctuating lighting, further substantiated its robustness and its readiness for real-world application. [Conclusions] This research offers a valuable reference for accurately detecting wheat leaf numbers in intricate field conditions, as well as robust technical support for the comprehensive and high-quality assessment of wheat growth.
[Objective] The dynamic prediction of carbon emission from cold chain distribution is an important basis for the accurate assessment of carbon emission and its green credit grade. Facing the fact that the carbon emission of vehicles is affected by multiple factors, such as road condition information, driving characteristics, refrigeration parameters, etc., a dynamic prediction model of carbon emission was proposed from refrigerated vehicles that integrates multi-source information. [Methods] The backbone feature extraction network, neck feature fusion network and loss function of YOLOv8s was firstly improved. The full-dimensional dynamic convolution was introduced into the backbone feature extraction network, and the multidimensional attention mechanism was introduced to capture the contextual key information to improve the model feature extraction capability. A progressive feature pyramid network was introduced into the feature extraction network, which reduced the loss of key information by fusing features layer by layer and improved the feature fusion efficiency. The road condition information recognition model based on improved YOLOv8s was constructed to characterize the road condition information in terms of the number of road vehicles and the percentage of pixel area. Pearson's correlation coefficient was used to compare and analyze the correlation between carbon emissions of refrigerated vehicles and different influencing factors, and to verify the necessity and criticality of the selection of input parameters of the carbon emission prediction model. Then the iTransformer temporal prediction model was improved, and the external attention mechanism was introduced to enhance the feature extraction ability of iTransformer model and reduce the computational complexity. The dynamic prediction model of carbon emission of refrigerated vehicles based on the improved iTransformer was constructed by taking the road condition information, driving characteristics (speed, acceleration), cargo weight, and refrigeration parameters (temperature, power) as inputs. Finally, the model was compared and analyzed with other models to verify the robustness of the road condition information and the prediction accuracy of the vehicle carbon emission dynamic prediction model, respectively. [Results and Discussions] The results of correlation analysis showed that the vehicle driving parameters were the main factor affecting the intensity of vehicle carbon emissions, with a correlation of 0.841. The second factor was cargo weight, with a correlation of 0.807, which had a strong positive correlation. Compared with the vehicle refrigeration parameters, the road condition information had a stronger correlation between vehicle carbon emissions, the correlation between refrigeration parameters and the vehicle carbon emissions impact factor were above 0.67. In order to further ensure the accuracy of the vehicle carbon emissions prediction model, The paper was selected as the input parameters for the carbon emissions prediction model. The improved YOLOv8s road information recognition model achieved 98.1%, 95.5%, and 98.4% in precision, recall, and average recognition accuracy, which were 1.2%, 3.7%, and 0.2% higher than YOLOv8s, respectively, with the number of parameters and the amount of computation being reduced by 12.5% and 31.4%, and the speed of detection being increased by 5.4%. This was due to the cross-dimensional feature learning through full-dimensional dynamic convolution, which fully captured the key information and improved the feature extraction capability of the model, and through the progressive feature pyramid network after fusing the information between different classes through gradual step-by-step fusion, which fully retained the important feature information and improved the recognition accuracy of the model. The predictive performance of the improved iTransformer carbon emission prediction model was better than other time series prediction models, and its prediction curve was closest to the real carbon emission curve with the best fitting effect. The introduction of the external attention mechanism significantly improved the prediction accuracy, and its MSE, MAE, RMSE and R2 were 0.026 1 %VOL, 0.079 1 %VOL, 0.161 5 %VOL and 0.940 0, respectively, which were 0.4%, 15.3%, 8.7% and 1.3% lower, respectively, when compared with iTransformer. As the degree of road congestion increased, the prediction accuracy of the constructed carbon emission prediction model increased. [Conclusions] The carbon emission prediction model for cold chain distribution under multi-source information fusion proposed in this study can realize accurate prediction of carbon emission from refrigerated vehicles, provide theoretical basis for rationally formulating carbon emission reduction strategies and promoting the development of low-carbon cold chain distribution.
[Objective] Rice plays a crucial role in daily diet. The rice industry involves numerous links, from paddy planting to the consumer's table, and the integrity of the quality control data chain directly affects the credibility of rice quality control and traceability information. The process of rice traceability also faces security issues, such as the leakage of privacy information, which need immediate solutions. Additionally, the previous practice of uploading all information onto the blockchain leads to high storage costs and low system efficiency. To address these problems, this study proposed a differential privacy-enhanced blockchain-based quality control model for rice, providing new ideas and solutions to optimize the traditional quality regulation and traceability system. [Methods] By exploring technologies of blockchain, interplanetary file system (IPFS), and incorporating differential privacy techniques, a blockchain-based quality control model for rice with differential privacy enhancement was constructed. Firstly, the data transmission process was designed to cover the whole industry chain of rice, including cultivation, acquisition, processing, warehousing, and sales. Each module stored the relevant data and a unique number from the previous link, forming a reliable information chain and ensuring the continuity of the data chain for quality control. Secondly, to address the issue of large data volume and low efficiency of blockchain storage, the key quality control data of each link in the rice industry chain was stored in the IPFS. Subsequently, the hash value of the stored data was returned and recorded on the blockchain. Lastly, to enhance the traceability of the quality control model information, the sensitive information in the key quality control data related to the cultivation process was presented to users after undergoing differential privacy processing. Individual data was obfuscated to increase the credibility of the quality control information while also protecting the privacy of farmers' cultivation practices. Based on this model, a differential privacy-enhanced blockchain-based quality control system for rice was designed. [Results and Discussions] The architecture of the differential privacy-enhanced blockchain-based quality control system for rice consisted of the physical layer, transport layer, storage layer, service layer, and application layer. The physical layer included sensor devices and network infrastructure, ensuring data collection from all links of the industry chain. The transport layer handled data transmission and communication, securely uploading collected data to the cloud. The storage layer utilized a combination of traditional databases, IPFS, and blockchain to efficiently store and manage key data on and off the blockchain. The traditional database was used for the management and querying of structured data. IPFS stored the key quality control data in the whole industry chain, while blockchain was employed to store the hash values returned by IPFS. This integrated storage method improved system efficiency, ensured the continuity, reliability, and traceability of quality control data, and provided consumers with reliable information. The service layer was primarily responsible for handling business logic and providing functional services. The implementation of functions in the application layer relied heavily on the design of a series of interfaces within the service layer. Positioned at the top of the system architecture, the application layer was responsible for providing user-centric functionality and interfaces. This encompassed a range of applications such as web applications and mobile applications, aiming to present data and facilitate interactive features to fulfill the requirements of both consumers and businesses. Based on the conducted tests, the average time required for storing data in a single link of the whole industry chain within the system was 1.125 s. The average time consumed for information traceability query was recorded as 0.691 s. Compared to conventional rice quality regulation and traceability systems, the proposed system demonstrated a reduction of 6.64% in the storage time of single-link data and a decrease of 16.44% in the time required to perform information traceability query. [Conclusions] This study proposes a differential privacy-enhanced blockchain-based quality control model for rice. The model ensures the continuity of the quality control data chain by integrating the various links of the whole industry chain of rice. By combining blockchain with IPFS storage, the model addresses the challenges of large data volume and low efficiency of blockchain storage in traditional systems. Furthermore, the model incorporates differential privacy protection to enhance traceability while safeguarding the privacy of individual farmers. This study can provide reference for the design and improvement of rice quality regulation and traceability systems.
[Objective] In the retail market for agricultural products, consumers are increasingly concerned about the safety and health aspects of those products. Traceability of blockchain has emerged as a crucial solution to address these concerns. Essentially, a blockchain functions as a dynamic, distributed, and shared database. When implemented in the agricultural supply chain, it not only improves product transparency to attract more consumers but also raises concerns about consumer privacy disclosure. The level of consumer apprehension regarding privacy will directly influence their choice to purchase agricultural products traced through blockchain-traced. Moreover, retailers' choices to sell blockchain-traced produce are influenced by consumer privacy concerns. By analyzing the impact of blockchain technology on the competitive strategies, pricing, and decision-making among agricultural retailers, they can develop market competition strategies that suit their market conditions to bolster their competitiveness and optimize the agricultural supply chain to maximize overall benefits. [Methods] Based on Nash equilibrium and Stackelberg game theory, a market competition model was developed to analyze the interactions between existing and new agricultural product retailers. The competitive strategies adopted by agricultural product retailers were analyzed under four different options of whether two agricultural retailers sell blockchain agricultural products. It delved into product utility, optimal pricing, demand, and profitability for each retailer under these different scenarios. How consumer privacy concerns impact pricing and profits of two agricultural product retailers and the optimal response strategy choice of another retailer when the competitor made the decision choice first were also analyzed. This analysis aimed to guide agricultural product retailers in making strategic choices that would safeguard their profits and market positions. To address the cooperative game problem of agricultural product retailers in market competition, ensure that retailers could better cooperate in the game, blockchain smart contract technology was used. By encoding the process and outcomes of the Stackelberg game into smart contracts, retailers could input their specific variables and receive tailored strategy recommendations. Uploading game results onto the blockchain network ensured transparency and encouraged cooperative behavior among retailers. By using the characteristics of blockchain, the game results were uploaded to the blockchain network to regulate the cooperative behavior, to ensure the maximization of the overall interests of the supply chain. [Results and Discussions] The research highlighted the significant improvement in agricultural product quality transparency through blockchain traceability technology. However, concerns regarding consumer privacy arising from this traceability could directly impact the pricing, profitability and retailers' decisions to provide blockchain-traceable items. Furthermore, an analysis of the strategic balance between two agricultural product retailers revealed that in situations of low and high product information transparency, both retailers were inclined to simultaneously offer sell traceable products. In such a scenario, blockchain traceability technology enhanced the utility and profitability of retail agricultural products, leading consumers to prefer purchase these traceable products from retailers. In cases where privacy concerns and agricultural product information transparency were both moderate, the initial retailer was more likely to opt for blockchain-based traceable products. This was because consumers had higher trust in the initial retailer, enabling them to bear a higher cost associated with privacy concerns. Conversely, new retailers failed to gain a competitive advantage and eventually exit the market. When consumer privacy concerns exceeded a certain threshold, both competing agricultural retailers discovered that offering blockchain-based traceable products led to a decline in their profits. [Conclusions] When it comes to agricultural product quality and safety, incorporating blockchain technology in traceability significantly improves the transparency of quality-related information for agricultural products. However, it is important to recognize that the application of blockchain for agricultural product traceability is not universally suitable for all agricultural retailers. Retailers must evaluate their unique circumstances and make the most suitable decisions to enhance the effectiveness of agricultural products, drive sales demand, and increase profits. Within the competitive landscape of the agricultural product retail market, nurturing a positive collaborative relationship is essential to maximize mutual benefits and optimize the overall profitability of the agricultural product supply chain.
[Significance] With the rapid development of robotics technology and the persistently rise of labor costs, the application of robots in facility agriculture is becoming increasingly widespread. These robots can enhance operational efficiency, reduce labor costs, and minimize human errors. However, the complexity and diversity of facility environments, including varying crop layouts and lighting conditions, impose higher demands on robot navigation. Therefore, achieving stable, accurate, and rapid navigation for robots has become a key issue. Advanced sensor technologies and algorithms have been proposed to enhance robots' adaptability and decision-making capabilities in dynamic environments. This not only elevates the automation level of agricultural production but also contributes to more intelligent agricultural management. [Progress] This paper reviews the key technologies of automatic navigation for facility agricultural robots. It details beacon localization, inertial positioning, simultaneous localization and mapping (SLAM) techniques, and sensor fusion methods used in autonomous localization and mapping. Depending on the type of sensors employed, SLAM technology could be subdivided into vision-based, laser-based and fusion systems. Fusion localization is further categorized into data-level, feature-level, and decision-level based on the types and stages of the fused information. The application of SLAM technology and fusion localization in facility agriculture has been increasingly common. Global path planning plays a crucial role in enhancing the operational efficiency and safety of facility aricultural robots. This paper discusses global path planning, classifying it into point-to-point local path planning and global traversal path planning. Furthermore, based on the number of optimization objectives, it was divided into single-objective path planning and multi-objective path planning. In regard to automatic obstacle avoidance technology for robots, the paper discusses sevelral commonly used obstacle avoidance control algorithms commonly used in facility agriculture, including artificial potential field, dynamic window approach and deep learning method. Among them, deep learning methods are often employed for perception and decision-making in obstacle avoidance scenarios. [Conclusions and Prospects] Currently, the challenges for facility agricultural robot navigation include complex scenarios with significant occlusions, cost constraints, low operational efficiency and the lack of standardized platforms and public datasets. These issues not only affect the practical application effectiveness of robots but also constrain the further advancement of the industry. To address these challenges, future research can focus on developing multi-sensor fusion technologies, applying and optimizing advanced algorithms, investigating and implementing multi-robot collaborative operations and establishing standardized and shared data platforms.
[Significance] Fruit-picking robot stands as a crucial solution for achieving intelligent fruit harvesting. Significant progress has been made in developing foundational methods for picking robots, such as fruit recognition, orchard navigation, path planning for picking, and robotic arm control, the practical implementation of a seamless picking system that integrates sensing, movement, and picking capabilities still encounters substantial technical hurdles. In contrast to current picking systems, the next generation of fruit-picking robots aims to replicate the autonomous skills exhibited by human fruit pickers. This involves effectively performing ongoing tasks of perception, movement, and picking without human intervention. To tackle this challenge, this review delves into the latest research methodologies and real-world applications in this field, critically assesses the strengths and limitations of existing methods and categorizes the essential components of continuous operation into three sub-modules: local target recognition, global mapping, and operation planning. [Progress] Initially, the review explores methods for recognizing nearby fruit and obstacle targets. These methods encompass four main approaches: low-level feature fusion, high-level feature learning, RGB-D information fusion, and multi-view information fusion, respectively. Each of these approaches incorporates advanced algorithms and sensor technologies for cluttered orchard environments. For example, low-level feature fusion utilizes basic attributes such as color, shapes and texture to distinguish fruits from backgrounds, while high-level feature learning employs more complex models like convolutional neural networks to interpret the contextual relationships within the data. RGB-D information fusion brings depth perception into the mix, allowing robots to gauge the distance to each fruit accurately. Multi-view information fusion tackles the issue of occlusions by combining data from multiple cameras and sensors around the robot, providing a more comprehensive view of the environment and enabling more reliable sensing. Subsequently, the review shifts focus to orchard mapping and scene comprehension on a broader scale. It points out that current mapping methods, while effective, still struggle with dynamic changes in the orchard, such as variations of fruits and light conditions. Improved adaptation techniques, possibly through machine learning models that can learn and adjust to different environmental conditions, are suggested as a way forward. Building upon the foundation of local and global perception, the review investigates strategies for planning and controlling autonomous behaviors. This includes not only the latest advancements in devising movement paths for robot mobility but also adaptive strategies that allow robots to react to unexpected obstacles or changes within the whole environment. Enhanced strategies for effective fruit picking using the Eye-in-Hand system involve the development of more dexterous robotic hands and improved algorithms for precisely predicting the optimal picking point of each fruit. The review also identifies a crucial need for further advancements in the dynamic behavior and autonomy of these technologies, emphasizing the importance of continuous learning and adaptive control systems to improve operational efficiency in diverse orchard environments. [Conclusions and Prospects] The review underscores the critical importance of coordinating perception, movement, and picking modules to facilitate the transition from a basic functional prototype to a practical machine. Moreover, it emphasizes the necessity of enhancing the robustness and stability of core algorithms governing perception, planning, and control, while ensuring their seamless coordination which is a central challenge that emerges. Additionally, the review raises unresolved questions regarding the application of picking robots and outlines future trends, include deeper integration of stereo vision and deep learning, enhanced global vision sampling, and the establishment of standardized evaluation criteria for overall operational performance. The paper can provide references for the eventual development of robust, autonomous, and commercially viable picking robots in the future.
[Objective] Sunlight-induced chlorophyll fluorescence (SIF) data obtained from satellites suffer from issues such as low spatial and temporal resolution, and discrete footprint because of the limitations imposed by satellite orbits. To address these problems, obtaining higher resolution SIF data, most reconstruction studies are based on low-resolution satellite SIF. Moreover, the spatial resolution of most SIF reconstruction products is still not enough to be directly used for the study of crop photosynthetic rate at the regional scale. Although some SIF products boast elevated resolutions, but these derive not from the original satellite SIF data reconstruct but instead evolve from secondary reconstructions based on preexisting SIF reconstruction products. Satellite OCO-2 (The Orbiting Carbon Obsevatory-2) equipped with a high-resolution spectrometer, OCO-2 SIF has higher spatial resolution (1.29×2.25 km) compared to other original SIF products, making it suitable in advancing the realm of high-resolution SIF data reconstruction, particularly within the context of regional-scale crop studies. [Methods] This research primarily exploration SIF reconstruct at the regional scale, mainly focused on the partial soybean planting regions nestled within the United States. The selection of MODIS raw data hinged on a meticulous consideration of environmental conditions, the distinctive physiological attributes of soybeans, and an exhaustive evaluation of factors intricately linked to OCO-2 SIF within these soybean planting regions. The primary tasks of this research encompassed reconstructing high resolution soybean SIF while concurrently executing a rigorous assessment of the reconstructed SIF's quality. During the dataset construction process, amalgamated SIF data from multiple soybean planting regions traversed by the OCO-2 satellite's footprint to retain as many of the available original SIF samples as possible. This approach provided the subsequent SIF reconstruction model with a rich source of SIF data. SIF data obtained beneath the satellite's trajectory were matched with various MODIS datasets, including enhanced vegetation index (EVI), fraction of photosynthetically active radiation (FPAR), and land surface temperature (LST), resulting in the creation of a multisource remote sensing dataset ultimately used for model training. Because of the multisource remote sensing dataset encompassed the most relevant explanatory variables within each SIF footprint coverage area concerning soybean physiological structure and environmental conditions. Through the activation functions in the BP neural network, it enhanced the understanding of the complex nonlinear relationships between the original SIF data and these MODIS products. Leveraging these inherent nonlinear relationships, compared and analyzed the effects of different combinations of explanatory variables on SIF reconstruction, mainly analyzing the three indicators of goodness of fit R2, root mean square error RMSE, and mean absolute error MAE, and then selecting the best SIF reconstruction model, generate a regional scale, spatially continuous, and high temporal resolution (500 m, 8 d) soybean SIF reconstruction dataset (BPSIF). [Results and Discussions] The research findings confirmed the strong performance of the SIF reconstruction model in predicting soybean SIF. After simultaneously incorporating EVI, FPAR, and LST as explanatory variables to model, achieved a goodness of fit with an R2 value of 0.84, this statistical metric validated the model's capability in predicting SIF data, it also reflected that the reconstructed 8 d time resolution of SIF data's reliability of applying to small-scale agricultural crop photosynthesis research with 500 m×500 m spatial scale. Based on this optimal model, generated the reconstructed SIF product (BPSIF). The Pearson correlation coefficient between the original OCO-2 SIF data and MODIS GPP stood were at a modest 0.53. In stark contrast, the correlation coefficient between BPSIF and MODIS Gross Primary Productivity (GPP) rosed significantly to 0.80. The increased correlation suggests that BPSIF could more accurately reflect the dynamic changes in GPP during the soybean growing season, making it more reliable compared to the original SIF data. Selected soybean planting areas in the United States with relatively single crop cultivation as the research area, based on high spatial resolution (1.29 km×2.25 km) OCO-2 SIF data, greatly reduced vegetation heterogeneity under a single SIF footprint. [Conclusions] The BPSIF proposed has significantly enhancing the regional and temporal continuity of OCO-2 SIF while preserving the time and spatial attributes contained in the original SIF dataset. Within the study area, BPSIF exhibits a significantly improved correlation with MODIS GPP compared to the original OCO-2 SIF. The proposed OCO-2 SIF data reconstruction method in this study holds the potential to provide a more reliable SIF dataset. This dataset has the potential to drive further understanding of soybean SIF at finer spatial and temporal scales, as well as find its relationship with soybean GPP.
[Objective] Acurately determining the suitable sowing date for winter wheat is of great significance for improving wheat yield and ensuring national food security. Traditional visual interpretation method is not only time-consuming and labor-intensive, but also covers a relatively small area. Remote sensing monitoring, belongs to post-event monitoring, exhibits a time lag. The aim of this research is to use the temperature threshold method and accumulated thermal time requirements for wheat leaves appearance method to analyze the suitable sowing date for winter wheat in county-level towns under the influence of long-term sequence of climate warming. [Methods] The research area were various townships in Qihe county, Shandong province. Based on European centre for medium-range weather forecasts (ECMWF) reanalysis data from 1997 to 2022, 16 meteorological data grid points in Qihe county were selected. Firstly, the bilinear interpolation method was used to interpolate the temperature data of grid points into the approximate center points of each township in Qihe county, and the daily average temperatures for each township were obtained. Then, temperature threshold method was used to determine the final dates of stable passage through 18, 16, 14 and 0 ℃. Key sowing date indicators such as suitable sowing temperature for different wheat varieties, growing degree days (GDD)≥0 ℃ from different sowing dates to before overwintering, and daily average temperature over the years were used for statistical analysis of the suitable sowing date for winter wheat. Secondly, the accumulated thermal time requirements for wheat leaves appearance method was used to calculate the appropriate date of GDD for strong seedlings before winter by moving forward from the stable date of dropping to 0 ℃. Accumulating the daily average temperatures above 0 ℃ to the date when the GDD above 0 ℃ was required for the formation of strong seedlings of wheat, a range of ±3 days around this calculated date was considered the theoretical suitable sowing date. Finally, combined with actual production practices, the appropriate sowing date of winter wheat in various townships of Qihe county was determined under the trend of climate warming. [Results and Discussions] The results showed that, from November 1997 to early December 2022, winter and annual average temperatures in Qihe county had all shown an upward trend, and there was indeed a clear trend of climate warming in various townships of Qihe county. Judging from the daily average temperature over the years, the temperature fluctuation range in November was the largest in a year, with a maximum standard deviation was 2.61 ℃. This suggested a higher likelihood of extreme weather conditions in November. Therefore, it was necessary to take corresponding measures to prevent and reduce disasters in advance to avoid affecting the growth and development of wheat. In extreme weather conditions, it was limited to determine the sowing date only by temperature or GDD. In cold winter years, it was too one-sided to consider only from the perspective of GDD. It was necessary to expand the range of GDD required for winter wheat before overwintering based on temperature changes to ensure the normal growth and development of winter wheat. The suitable sowing date for semi winter wheat obtained by temperature threshold method was from October 4th to October 16th, and the suitable sowing date for winter wheat was from September 27th to October 4th. Taking into account the GDD required for the formation of strong seedlings before winter, the suitable sowing date for winter wheat was from October 3rd to October 13th, and the suitable sowing date for semi winter wheat was from October 15th to October 24th, which was consisted with the suitable sowing date for winter wheat determined by the accumulated thermal time requirements for wheat leaves appearance method. Considering the winter wheat varieties planted in Qihe county, the optimal sowing date for winter wheat in Qihe county was from October 3rd to October 16th, and the optimal sowing date was from October 5th to October 13th. With the gradual warming of the climate, the suitable sowing date for wheat in various townships of Qihe county in 2022 was later than that in 2002. However, the sowing date for winter wheat was still influenced by factors such as soil moisture, topography, and seeding quality. The suitable sowing date for a specific year still needed to be adjusted to local conditions and flexibly sown based on the specific situation of that year. [Conclusions] The experimental results proved the feasibility of the temperature threshold method and accumulated thermal time requirements for wheat leaves appearance method in determining the suitable sowing date for winter wheat. The temperature trend can be used to identify cold or warm winters, and the sowing date can be adjusted in a timely manner to enhance wheat yield and reduce the impact of excessively high or low temperatures on winter wheat. The research results can not only provide decision-making reference for winter wheat yield assessment in Qihe county, but also provide an important theoretical basis for scientifically arrangement of agricultural production.
[Objective] Nitrogen in soil is an absolutely crucial element for plant growth. Insufficient nitrogen supply can severely affect crop yield and quality, while excessive use of nitrogen fertilizers can lead to significant environmental issues such as water eutrophication and groundwater pollution. Therefore, large-scale, rapid detection of soil nitrogen content and precise fertilization are of great importance for smart agriculture. In this study, the hyperspectral data from the GF-5 satellite was emploied, and the various machine learning algorithms introduced to establish a prediction model for soil total nitrogen (TN) content and a distribution map of soil TN content was generated in the study area, aiming to provide scientific evidence for intelligent monitoring in smart agriculture. [Method] The study area was the Jian Sanjiang Reclamation Area in Fujin city, Heilongjiang province. Fieldwork involved the careful collection of 171 soil samples, obtaining soil spectral data, chemical analysis data of soil TN content, and the GF-5 hyperspectral data. Among these samples, 140 were randomly selected as the modeling sample set for calibration, and the remaining 31 samples were used as the test sample set. Three machine learning algorithms were introduced: Partial least squares regression (PLSR), backpropagation neural network (BPNN), and support vector machine (SVM) driven by a polynomial kernel function (Poly). Three distinct soil TN inversion models were constructed using these algorithms. To optimize model performance, ten-fold cross-validation was employed to determine the optimal parameters for each model. Additionally, multiple scatter correction (MSC) was applied to obtain band characteristic values, thus enhancing the model's prediction capability. Model performance was evaluated using three indicators: Coefficient of determination (R²), root mean square error (RMSE), and relative prediction deviation (RPD), to assess the prediction accuracy of different models. [Results and Discussions] MSC-Poly-SVM model exhibited the best prediction performance on the test sample set, with an R² of 0.863, an RMSE of 0.203, and an RPD of 2.147. This model was used to perform soil TN content inversion mapping using GF-5 satellite hyperspectral data. In accordance with the stringent requirements of land quality geochemical evaluation, the GF-5 hyperspectral land organic nitrogen parameter distribution map was drawn based on the "Determination of Land Quality Geochemical Evaluation". The results revealed that 86.1% of the land in the Jian Sanjiang study area had a total nitrogen content of more than 2.0 g/kg, primarily concentrated in first and second-grade plots, while third and fourth-grade plots accounted for only 11.83% of the total area. The study area exhibited sufficient soil nitrogen reserves, with high TN background values mainly concentrated along the riverbanks in the central part, distributed in a northeast-east direction. Specifically, in terms of soil spectral preprocessing, the median filtering method performed best in terms of smoothness and maintaining spectral characteristics. The spectra extracted from GF-5 imagery were generally quite similar to ground-measured spectral data, despite some noise, which had a minimal overall impact. [Conclusions] This study demonstrates the clear feasibility of using GF-5 satellite hyperspectral remote sensing data and machine learning algorithm for large-scale quantitative detection and visualization analysis of soil TN content. The soil TN content distribution map generated based on GF-5 hyperspectral remote sensing data is detailed and consistent with results from other methods, providing technical support for future large-scale quantitative detection of soil nutrient status and rational fertilization.
[Objective] In order to intelligently monitor the distribution of agricultural land cover types, high-spectral cameras are usually mounted on drones to collect high-spectral data, followed by classification of the high-spectral data to automatically draw crop distribution maps. Different crops have similar shapes, and the same crop has significant differences in different growth stages, so the network model for agricultural land cover classification requires a high degree of accuracy. However, network models with high classification accuracy are often complex and cannot be deployed on hardware systems. In view of this problem, a lightweight high-low frequency enhanced Reluformer network (ReluformerN) was proposed in this research. [Methods] Firstly, an adaptive octave convolution was proposed, which utilized the softmax function to automatically adjust the spectral dimensions of high-frequency features and low-frequency features, effectively alleviating the influence of manually setting the spectral dimensions and benefiting the subsequent extraction of spatial and spectral domain features of hyperspectral images. Secondly, a Reluformer was proposed to extract global features, taking advantage of the fact that low-frequency information could capture global features. Reluformer replaced the softmax function with a function of quadratic computational complexity, and through theoretical and graphical analysised, Relu function, LeakRelu function, and Gelu function were compared, it was found that the ReLU function and the softmax function both had non-negativity, which could be used for feature relevance analysis. Meanwhile, the ReLU function has a linearization feature, which is more suitable for self-relevance analysis. Therefore, the ReLU self-attention mechanism was proposed, which used the ReLU function to perform feature self-attention analysis. In order to extract deep global features, multi-scale feature fusion was used, and the ReLU self-attention mechanism was used as the core to construct the multi-head ReLU self-attention mechanism. Similar to the transformer architecture, the Reluformer structure was built by combining multi-head ReLU self-attention mechanism, feedforward layers, and normalization layers. With Reluformer as the core, the Reluformer network (ReluformerN) was proposed. This network considered frequency from the perspective of high-frequency information, taking into account the local features of image high-frequency information, and used deep separable convolution to design a lightweight network for fine-grained feature extraction of high-frequency information. It proposed Reluformer to extract global features for low-frequency information, which represented the global features of the image. ReluformerN was experimented on three public high-spectral data sets (Indian Pines, WHU-Hi-LongKou and Salinas) for crop variety fine classification, and was compared with five popular classification networks (2D-CNN, HybirdSN, ViT, CTN and LSGA-VIT). [Results and Discussion] ReluformerN performed best in overall accuracy (OA), average accuracy (AA), and other accuracy evaluation indicators. In the evaluation indicators of model parameters, model computation (FLOPs), and model complexity, ReluformerN had the smallest number of parameters and was less than 0.3 M, and the lowest computation. In the visualization comparison, the classification effect diagram of the model using ReluformerN had clearer image edges and more complete morphological structures, with fewer classification errors. The validity of the adaptive octave convolution was verified by comparing it with the traditional eightfold convolution. The classification accuracy of the adaptive octave convolution was 0.1% higher than that of the traditional octave convolution. When the artificial parameters were set to different values, the maximum and minimum classification accuracies of the traditional octave convolution were about 0.3% apart, while those of the adaptive octave convolution were only 0.05%. This showed that the adaptive octave convolution not only had the highest classification accuracy, but was also less sensitive to the artificial parameter setting, effectively overcoming the influence of the artificial parameter setting on the classification result. To validated the Reluformer module, it was compared with transformer, LeakRelufromer, and Linformer in terms of accuracy evaluation metrics such as OA and AA. The Reluformer achieved the highest classification accuracy and the lowest model parameter count among these models. This indicated that Reluformer not only effectively extracted global features but also reduced computational complexity. Finally, the effectiveness of the high-frequency and low-frequency branch networks was verified. The effectiveness of the high-frequency and low-frequency feature extraction branches was verified, and the characteristics of the feature distribution after high-frequency feature extraction, after high-low frequency feature extraction, and after the classifier were displayed using a 2D t-sne, compared with the original feature distribution. It was found that after high-frequency feature extraction, similar features were generally clustered together, but the spacing between different features was small, and there were also some features with overlapping situations. After low-frequency feature extraction, it was obvious that similar features were clustered more tightly. After high-low frequency feature fusion, and after the classifier, it was obvious that similar features were clustered, and different types of features were clearly separated, indicating that high-low frequency feature extraction enhanced the classification effect. [Conclusion] This network achieves a good balance between crop variety classification accuracy and model complexity, and is expected to be deployed on hardware systems with limited resources in the future to achieve real-time classification functions.
[Objective] The number, location, and crown spread of nursery stock are important foundations data for their scientific management. Traditional approach of conducting nursery stock inventories through on-site individual plant surveys is labor-intensive and time-consuming. Low-cost and convenient unmanned aerial vehicles (UAVs) for on-site collection of nursery stock data are beginning to be utilized, and the statistical analysis of nursery stock information through technical means such as image processing achieved. During the data collection process, as the flight altitude of the UAV increases, the number of trees in a single image also increases. Although the anchor box can cover more information about the trees, the cost of annotation is enormous in the case of a large number of densely populated tree images. To tackle the challenges of tree adhesion and scale variance in images captured by UAVs over nursery stock, and to reduce the annotation costs, using point-labeled data as supervisory signals, an improved dense detection and counting model was proposed to accurately obtain the location, size, and quantity of the targets. [Method] To enhance the diversity of nursery stock samples, the spruce dataset, the Yosemite, and the KCL-London publicly available tree datasets were selected to construct a dense nursery stock dataset. A total of 1 520 nursery stock images were acquired and divided into training and testing sets at a ratio of 7:3. To enhance the model's adaptability to tree data of different scales and variations in lighting, data augmentation methods such as adjusting the contrast and resizing the images were applied to the images in the training set. After enhancement, the training set consists of 3 192 images, and the testing set contains 456 images. Considering the large number of trees contained in each image, to reduce the cost of annotation, the method of selecting the center point of the trees was used for labeling. The LSC-CNN model was selected as the base model. This model can detect the quantity, location, and size of trees through point-supervised training, thereby obtaining more information about the trees. The LSC-CNN model was made improved to address issues of missed detections and false positives that occurred during the testing process. Firstly, to address the issue of missed detections caused by severe adhesion of densely packed trees, the last convolutional layer of the feature extraction network was replaced with dilated convolution. This change enlarges the receptive field of the convolutional kernel on the input while preserving the detailed features of the trees. So the model is better able to capture a broader range of contextual information, thereby enhancing the model's understanding of the overall scene. Secondly, the convolutional block attention module (CBAM) attention mechanism was introduced at the beginning of each scale branch. This allowed the model to focus on the key features of trees at different scales and spatial locations, thereby improving the model's sensitivity to multi-scale information. Finally, the model was trained using label smooth cross-entropy loss function and grid winner-takes-all strategy, emphasizing regions with highest losses to boost tree feature recognition. [Results and Discussions] The mean counting accuracy (MCA), mean absolute error (MAE), and root mean square error (RMSE) were adopted as evaluation metrics. Ablation studies and comparative experiments were designed to demonstrate the performance of the improved LSC-CNN model. The ablation experiment proved that the improved LSC-CNN model could effectively resolve the issues of missed detections and false positives in the LSC-CNN model, which were caused by the density and large-scale variations present in the nursery stock dataset. IntegrateNet, PSGCNet, CANet, CSRNet, CLTR and LSC-CNN models were chosen as comparative models. The improved LSC-CNN model achieved MCA, MAE, and RMSE of 91.23%, 14.24, and 22.22, respectively, got an increase in MCA by 6.67%, 2.33%, 6.81%, 5.31%, 2.09% and 2.34%, respectively; a reduction in MAE by 21.19, 11.54, 18.92, 13.28, 11.30 and 10.26, respectively; and a decrease in RMSE by 28.22, 28.63, 26.63, 14.18, 24.38 and 12.15, respectively, compared to the IntegrateNet, PSGCNet, CANet, CSRNet, CLTR and LSC-CNN models. These results indicate that the improved LSC-CNN model achieves high counting accuracy and exhibits strong generalization ability. [Conclusions] The improved LSC-CNN model integrated the advantages of point supervision learning from density estimation methods and the generation of target bounding boxes from detection methods.These improvements demonstrate the enhanced performance of the improved LSC-CNN model in terms of accuracy, precision, and reliability in detecting and counting trees. This study could hold practical reference value for the statistical work of other types of nursery stock.
[Objective] The number of effective tillers per plant is one of the important agronomic traits affecting rice yield. In order to solve the problems of high cost and low accuracy of effective tiller detection caused by dense tillers, mutual occlusion and ineffective tillers in rice, a method for dividing effective tillers and ineffective tillers in rice was proposed. Combined with the deep learning model, a high-throughput and low-cost mobile phone App for effective tiller detection in rice was developed to solve the practical problems of effective tiller investigation in rice under field conditions. [Methods] The investigations of rice tillering showed that the number of effective tillers of rice was often higher than that of ineffective tillers. Based on the difference in growth height between effective and ineffective tillers of rice, a new method for distinguishing effective tillers from ineffective tillers was proposed. A fixed height position of rice plants was selected to divide effective tillers from ineffective tillers, and rice was harvested at this position. After harvesting, cross-sectional images of rice tillering stems were taken using a mobile phone, and the stems were detected and counted by the YOLOv8 model. Only the cross-section of the stem was identified during detection, while the cross-section of the panicle was not identified. The number of effective tillers of rice was determined by the number of detected stems. In order to meet the needs of field work, a mobile phone App for effective tiller detection of rice was developed for real-time detection. GhostNet was used to lighten the YOLOv8 model. Ghost Bottle-Neck was integrated into C2f to replace the original BottleNeck to form C2f-Ghost module, and then the ordinary convolution in the network was replaced by Ghost convolution to reduce the complexity of the model. Based on the lightweight Ghost-YOLOv8 model, a mobile App for effective tiller detection of rice was designed and constructed using the Android Studio development platform and intranet penetration counting. [Results and Discussions] The results of field experiments showed that there were differences in the growth height of effective tillers and ineffective tillers of rice. The range of 52 % to 55 % of the total plant height of rice plants was selected for harvesting, and the number of stems was counted as the number of effective tillers per plant. The range was used as the division standard of effective tillers and ineffective tillers of rice. The accuracy and recall rate of effective tillers counting exceeded 99%, indicating that the standard was accurate and comprehensive in guiding effective tillers counting. Using the GhostNet lightweight YOLOv8 model, the parameter quantity of the lightweight Ghost-YOLOv8 model was reduced by 43%, the FPS was increased by 3.9, the accuracy rate was 0.988, the recall rate was 0.980, and the mAP was 0.994. The model still maintains excellent performance while light weighting. Based on the lightweight Ghost-YOLOv8 model, a mobile phone App for detecting effective tillers of rice was developed. The App was tested on 100 cross-sectional images of rice stems collected under the classification criteria established in this study. Compared with the results of manual counting of effective tillers per plant, the accuracy of the App's prediction results was 99.61%, the recall rate was 98.76%, and the coefficient of determination was 0.985 9, indicating the reliability of the App and the established standards in detecting effective tillers of rice. [Conclusions] Through the lightweight Ghost-YOLOv8 model, the number of stems in the cross-sectional images of stems collected under the standard was detected to obtain the effective tiller number of rice. An Android-side rice effective tillering detection App was developed, which can meet the field investigation of rice effective tillering, help breeders to collect data efficiently, and provide a basis for field prediction of rice yield. Further research could supplement the cross-sectional image dataset of multiple rice stems to enable simultaneous measurement of effective tillers across multiple rice plants and improve work efficiency. Further optimization and enhancement of the App's functionality is necessary to provide more tiller-related traits, such as tiller angle.
[Objective] In the agricultural production, accurately classifying dried daylily grades is a critical task with significant economic implications. However, current target detection models face challenges such as inadequate accuracy and excessive parameters when applied to dried daylily grading, limiting their practical application and widespread use in real-world settings. To address these issues, an innovative lightweight YOLOv10-AD network model was proposed. The model aims to enhance detection accuracy by optimizing the network structure and loss functions while reducing parameters and computational costs, making it more suitable for deployment in resource-constrained agricultural production environments. [Methods] The dried daylilies selected from the Qingyang region of Gansu province as the research subject. A large number of images of dried daylilies, categorized into three grades superior, medium, and inferior, were collected using mobile phones under varying lighting conditions and backgrounds. The images were carefully annotated and augmented to build a comprehensive dataset for dried daylily grade classification. YOLOv10 was chosen as the base network, and a newly designed backbone network called AKVanillaNet was introduced. AKVanillaNet combines AKConv (adaptive kernel convolution) with VanillaNet's deep learning and shallow inference mechanisms. The second convolutional layer in VanillaNet was replaced with AKConv, and AKConv was merged with standard convolution layers at the end of the training phase to optimize the model for capturing the unique shape characteristics of dried daylilies. This innovative design not only improved detection accuracy but also significantly reduced the number of parameters and computational costs. Additionally, the DysnakeConv module was integrated into the C2f structure, replacing the Bottleneck layer with a Bottleneck-DS layer to form the new C2f-DysnakeConv module. This module enhanced the model's sensitivity to the shapes and boundaries of targets, allowing the neural network to better capture the shape information of irregular objects like dried daylilies, further improving the model's feature extraction capability. The Powerful-IOU (PIOU) loss function was also employed, which introduced a target-size-adaptive penalty factor and a gradient adjustment function. This design guided the anchor box regression along a more direct path, helping the model better fit the data and improve overall performance. [Results and Discussions] The testing results on the dried daylily grade classification dataset demonstrated that the YOLOv10-AD model achieved a mean average precision (mAP) of 85.7%. The model's parameters, computational volume, and size were 2.45 M, 6.2 GFLOPs, and 5.0 M, respectively, with a frame rate of 156 FPS. Compared to the benchmark model, YOLOv10-AD improved mAP by 5.7% and FPS by 25.8%, while reducing the number of parameters, computational volume, and model size by 9.3%, 24.4%, and 9.1%, respectively. These results indicated that YOLOv10-AD not only improved detection accuracy but also reduced the model's complexity, making it easier to deploy in real-world production environments. Furthermore, YOLOv10-AD outperformed larger models in the same series, such as YOLOv10s and YOLOv10m. Specifically, the weight, parameters, and computational volume of YOLOv10-AD were only 31.6%, 30.5%, and 25.3% of those in YOLOv10s, and 15.7%, 14.8%, and 9.8% of YOLOv10m. Despite using fewer resources, YOLOv10-AD achieved a mAP increase of 2.4% over YOLOv10s and 1.9% over YOLOv10m. These findings confirm that YOLOv10-AD maintains high detection accuracy while requiring significantly fewer resources, making it more suitable for agricultural production environments where computational capacity may be limited. The study also examined the performance of YOLOv10-AD under different lighting conditions. The results showed that YOLOv10-AD achieved an average accuracy of 92.3% in brighter environments and 78.6% in darker environments. In comparison, the YOLOv10n model achieved 88.9% and 71.0% in the same conditions, representing improvements of 3.4% and 7.6%, respectively. These findings demonstrate that YOLOv10-AD has a distinct advantage in maintaining high accuracy and confidence in grading dried daylilies across varying lighting conditions. [Conclusions] The YOLOv10-AD network model proposed significantly reduces the number of parameters and computational costs without compromising detection accuracy. This model presents a valuable technical reference for intelligent classification of dried daylily grades in agricultural production environments, particularly where resources are constrained.
[Objective] The picking of famous and high-quality tea is a crucial link in the tea industry. Identifying and locating the tender buds of famous and high-quality tea for picking is an important component of the modern tea picking robot. Traditional neural network methods suffer from issues such as large model size, long training times, and difficulties in dealing with complex scenes. In this study, based on the actual scenario of the Xiqing Tea Garden in Hunan Province, proposes a novel deep learning algorithm was proposed to solve the precise segmentation challenge of famous and high-quality tea picking points. [Methods] The primary technical innovation resided in the amalgamation of a lightweight network architecture, MobilenetV2, with an attention mechanism known as efficient channel attention network (ECANet), alongside optimization modules including atrous spatial pyramid pooling (ASPP). Initially, MobilenetV2 was employed as the feature extractor, substituting traditional convolution operations with depth wise separable convolutions. This led to a notable reduction in the model's parameter count and expedited the model training process. Subsequently, the innovative fusion of ECANet and ASPP modules constituted the ECA_ASPP module, with the intention of bolstering the model's capacity for fusing multi-scale features, especially pertinent to the intricate recognition of tea shoots. This fusion strategy facilitated the model's capability to capture more nuanced features of delicate shoots, thereby augmenting segmentation accuracy. The specific implementation steps entailed the feeding of image inputs through the improved network, whereupon MobilenetV2 was utilized to extract both shallow and deep features. Deep features were then fused via the ECA_ASPP module for the purpose of multi-scale feature integration, reinforcing the model's resilience to intricate backgrounds and variations in tea shoot morphology. Conversely, shallow features proceeded directly to the decoding stage, undergoing channel reduction processing before being integrated with upsampled deep features. This divide-and-conquer strategy effectively harnessed the benefits of features at differing levels of abstraction and, furthermore, heightened the model's recognition performance through meticulous feature fusion. Ultimately, through a sequence of convolutional operations and upsampling procedures, a prediction map congruent in resolution with the original image was generated, enabling the precise demarcation of tea shoot harvesting points. [Results and Discussions] The experimental outcomes indicated that the enhanced DeepLabV3+ model had achieved an average Intersection over Union (IoU) of 93.71% and an average pixel accuracy of 97.25% on the dataset of tea shoots. Compared to the original model based on Xception, there was a substantial decrease in the parameter count from 54.714 million to a mere 5.818 million, effectively accomplishing a significant lightweight redesign of the model. Further comparisons with other prevalent semantic segmentation networks revealed that the improved model exhibited remarkable advantages concerning pivotal metrics such as the number of parameters, training duration, and average IoU, highlighting its efficacy and precision in the domain of tea shoot recognition. This considerable decreased in parameter numbers not only facilitated a more resource-economical deployment but also led to abbreviated training periods, rendering the model highly suitable for real-time implementations amidst tea garden ecosystems. The elevated mean IoU and pixel accuracy attested to the model's capacity for precise demarcation and identification of tea shoots, even amidst intricate and varied datasets, demonstrating resilience and adaptability in pragmatic contexts. [Conclusions] This study effectively implements an efficient and accurate tea shoot recognition method through targeted model improvements and optimizations, furnishing crucial technical support for the practical application of intelligent tea picking robots. The introduction of lightweight DeepLabV3+ not only substantially enhances recognition speed and segmentation accuracy, but also mitigates hardware requirements, thereby promoting the practical application of intelligent picking technology in the tea industry.
[Objective] As one of China's most important agricultural products, apples hold a significant position in cultivation area and yield. However, during the growth process, apples are prone to various diseases that not only affect the quality of the fruit but also significantly reduce the yield, impacting farmers' economic benefits and the stability of market supply. To reduce the incidence of apple diseases and increase fruit yield, developing efficient and fast apple leaf disease detection technology is of great significance. An improved YOLOv8 algorithm was proposed to identify the leaf diseases that occurred during the growth of apples. [Methods] YOLOv8n model was selected to detect various leaf diseases such as brown rot, rust, apple scab, and sooty blotch that apples might encounter during growth. SPD-Conv was introduced to replace the original convolutional layers to retain fine-grained information and reduce model parameters and computational costs, thereby improving the accuracy of disease detection. The multi-scale dilated attention (MSDA) attention mechanism was added at appropriate positions in the Neck layer to enhance the model's feature representation capability, which allowed the model to learn the receptive field dynamically and adaptively focus on the most representative regions and features in the image, thereby enhancing the ability to extract disease-related features. Finally, inspired by the RepVGG architecture, the original detection head was optimized to achieve a separation of detection and inference architecture, which not only accelerated the model's inference speed but also enhanced feature learning capability. Additionally, a dataset of apple leaf diseases containing the aforementioned diseases was constructed, and experiments were conducted. [Results and Discussions] Compared to the original model, the improved model showed significant improvements in various performance metrics. The mAP50 and mAP50:95 achieved 88.2% and 37.0% respectively, which were 2.7% and 1.3% higher than the original model. In terms of precision and recall, the improved model increased to 83.1% and 80.2%, respectively, representing an improvement of 0.9% and 1.1% over the original model. Additionally, the size of the improved model was only 7.8 MB, and the computational cost was reduced by 0.1 G FLOPs. The impact of the MSDA placement on model performance was analyzed by adding it at different positions in the Neck layer, and relevant experiments were designed to verify this. The experimental results showed that adding MSDA at the small target layer in the Neck layer achieved the best effect, not only improving model performance but also maintaining low computational cost and model size, providing important references for the optimization of the MSDA mechanism. To further verify the effectiveness of the improved model, various mainstream models such as YOLOv7-tiny, YOLOv9-c, RetinaNet, and Faster-RCNN were compared with the propoed model. The experimental results showed that the improved model outperformed these models by 1.4%, 1.3%, 7.8%, and 11.6% in mAP50, 2.8%, 0.2%, 3.4%, and 5.6% in mAP50:95. Moreover, the improved model showed significant advantages in terms of floating-point operations, model size, and parameter count, with a parameter count of only 3.7 MB, making it more suitable for deployment on hardware-constrained devices such as drones. In addition, to assess the model's generalization ability, a stratified sampling method was used, selecting 20% of the images from the dataset as the test set. The results showed that the improved model could maintain a high detection accuracy in complex and variable scenes, with mAP50 and mAP50:95 increasing by 1.7% and 1.2%, respectively, compared to the original model. Considering the differences in the number of samples for each disease in the dataset, a class balance experiment was also designed. Synthetic samples were generated using oversampling techniques to increase the number of minority-class samples. The experimental results showed that the class-balanced dataset significantly improved the model's detection performance, with overall accuracy increasing from 83.1% to 85.8%, recall from 80.2% to 83.6%, mAP50 from 88.2% to 88.9%, and mAP50:95 from 37.0% to 39.4%. The class-balanced dataset significantly enhanced the model's performance in detecting minority diseases, thereby improving the overall performance of the model. [Conclusions] The improved model demonstrated significant advantages in apple leaf disease detection. By introducing SPD-Conv and MSDA attention mechanisms, the model achieved noticeable improvements in both precision and recall while effectively reducing computational costs, leading to more efficient detection capabilities. The improved model could provide continuous health monitoring throughout the apple growth process and offer robust data support for farmers' scientific decision-making before fruit harvesting.
[Objective] Traditional object detection algorithms applied in the agricultural field, such as those used for crop growth monitoring and harvesting, often suffer from insufficient accuracy. This is particularly problematic for small crops like mushrooms, where recognition and detection are more challenging. The introduction of small object detection technology promises to address these issues, potentially enhancing the precision, efficiency, and economic benefits of agricultural production management. However, achieving high accuracy in small object detection has remained a significant challenge, especially when dealing with varying image sizes and target scales. Although the YOLO series models excel in speed and large object detection, they still have shortcomings in small object detection. To address the issue of maintaining high accuracy amid changes in image size and target scale, a novel detection model, Multi-Strategy Handling YOLOv8 (MSH-YOLOv8), was proposed. [Methods] The proposed MSH-YOLOv8 model builds upon YOLOv8 by incorporating several key enhancements aimed at improving sensitivity to small-scale targets and overall detection performance. Firstly, an additional detection head was added to increase the model's sensitivity to small objects. To address computational redundancy and improve feature extraction, the Swin Transformer detection structure was introduced into the input module of the head network, creating what was termed the "Swin Head (SH)". Moreover, the model integrated the C2f_Deformable convolutionv4 (C2f_DCNv4) structure, which included deformable convolutions, and the Swin Transformer encoder structure, termed "Swinstage", to reconstruct the YOLOv8 backbone network. This optimization enhanced feature propagation and extraction capabilities, increasing the network's ability to handle targets with significant scale variations. Additionally, the normalization-based attention module (NAM) was employed to improve performance without compromising detection speed or computational complexity. To further enhance training efficacy and convergence speed, the original loss function CIoU was replaced with wise-intersection over union (WIoU) Loss. Furthermore, experiments were conducted using mushrooms as the research subject on the open Fungi dataset. Approximately 200 images with resolution sizes around 600×800 were selected as the main research material, along with 50 images each with resolution sizes around 200×400 and 1 000×1 200 to ensure representativeness and generalization of image sizes. During the data augmentation phase, a generative adversarial network (GAN) was utilized for resolution reconstruction of low-resolution images, thereby preserving semantic quality as much as possible. In the post-processing phase, dynamic resolution training, multi-scale testing, soft non-maximum suppression (Soft-NMS), and weighted boxes fusion (WBF) were applied to enhance the model's small object detection capabilities under varying scales. [Results and Discussions] The improved MSH-YOLOv8 achieved an average precision at 50% (AP50) intersection over union of 98.49% and an AP@50-95 of 75.29%, with the small object detection metric APs reaching 39.73%. Compared to mainstream models like YOLOv8, these metrics showed improvements of 2.34%, 4.06% and 8.55%, respectively. When compared to the advanced TPH-YOLOv5 model, the improvements were 2.14%, 2.76% and 6.89%, respectively. The ensemble model, MSH-YOLOv8-ensemble, showed even more significant improvements, with AP50 and APs reaching 99.14% and 40.59%, respectively, an increase of 4.06% and 8.55% over YOLOv8. These results indicate the robustness and enhanced performance of the MSH-YOLOv8 model, particularly in detecting small objects under varying conditions. Further application of this methodology on the Alibaba Cloud Tianchi databases "Tomato Detection" and "Apple Detection" yielded MSH-YOLOv8-t and MSH-YOLOv8-a models (collectively referred to as MSH-YOLOv8). Visual comparison of detection results demonstrated that MSH-YOLOv8 significantly improved the recognition of dense and blurry small-scale tomatoes and apples. This indicated that the MSH-YOLOv8 method possesses strong cross-dataset generalization capability and effectively recognizes small-scale targets. In addition to quantitative improvements, qualitative assessments showed that the MSH-YOLOv8 model could handle complex scenarios involving occlusions, varying lighting conditions, and different growth stages of the crops. This demonstrates the practical applicability of the model in real-world agricultural settings, where such challenges are common. [Conclusions] The MSH-YOLOv8 improvement method proposed in this study effectively enhances the detection accuracy of small mushroom targets under varying image sizes and target scales. This approach leverages multiple strategies to optimize both the architecture and the training process, resulting in a robust model capable of high-precision small object detection. The methodology's application to other datasets, such as those for tomato and apple detection, further underscores its generalizability and potential for broader use in agricultural monitoring and management tasks.
[Objective] The phenomenon of hoof slipping occurs during the walking process of cows, which indicates the deterioration of the farming environment and a decline in the cows' locomotor function. Slippery grounds can lead to injuries in cows, resulting in unnecessary economic losses for farmers. To achieve automatically recognizing and detecting slippery hoof postures during walking, the study focuses on the localization and analysis of key body points of cows based on deep learning methods. Motion curves of the key body points were analyzed, and features were extracted. The effectiveness of the extracted features was verified using a decision tree classification algorithm, with the aim of achieving automatic detection of slippery hoof postures in cows. [Method] An improved localization method for the key body points of cows, specifically the head and four hooves, was proposed based on the DeepLabCut model. Ten networks, including ResNet series, MobileNet-V2 series, and EfficientNet series, were selected to respectively replace the backbone network structure of DeepLabCut for model training. The root mean square error(RMSE), model size, FPS, and other indicators were chosen, and after comprehensive consideration, the optimal backbone network structure was selected as the pre-improved network. A network structure that fused the convolutional block attention module (CBAM) attention mechanism with ResNet-50 was proposed. A lightweight attention module, CBAM, was introduced to improve the ResNet-50 network structure. To enhance the model's generalization ability and robustness, the CBAM attention mechanism was embedded into the first convolution layer and the last convolution layer of the ResNet-50 network structure. Videos of cows with slippery hooves walking in profile were predicted for key body points using the improved DeepLabCut model, and the obtained key point coordinates were used to plot the motion curves of the cows' key body points. Based on the motion curves of the cows' key body points, the feature parameter Feature1 for detecting slippery hooves was extracted, which represented the local peak values of the derivative of the motion curves of the cows' four hooves. The feature parameter Feature2 for predicting slippery hoof distances was extracted, specifically the minimum local peak points of the derivative curve of the hooves, along with the local minimum points to the left and right of these peaks. The effectiveness of the extracted Feature1 feature parameters was verified using a decision tree classification model. Slippery hoof feature parameters Feature1 for each hoof were extracted, and the standard deviation of Feature1 was calculated for each hoof. Ultimately, a set of four standard deviations for each cow was extracted as input parameters for the classification model. The classification performance was evaluated using four common objective metrics, including accuracy, precision, recall, and F1-Score. The prediction accuracy for slippery hoof distances was assessed using RMSE as the evaluation metric. [Results and Discussion] After all ten models reached convergence, the loss values ranked from smallest to largest were found in the EfficientNet series, ResNet series, and MobileNet-V2 series, respectively. Among them, ResNet-50 exhibited the best localization accuracy in both the training set and validation set, with RMSE values of only 2.69 pixels and 3.31 pixels, respectively. The MobileNet series had the fastest inference speed, reaching 48 f/s, while the inference speeds of the ResNet series and MobileNet series were comparable, with ResNet series performing slightly better than MobileNet series. Considering the above factors, ResNet-50 was ultimately selected as the backbone network for further improvements on DeepLabCut. Compared to the original ResNet-50 network, the ResNet-50 network improved by integrating the CBAM module showed a significant enhancement in localization accuracy. The accuracy of the improved network increased by 3.7% in the training set and by 9.7% in the validation set. The RMSE between the predicted body key points and manually labeled points was only 2.99 pixels, with localization results for the right hind hoof, right front hoof, left hind hoof, left front hoof, and head improved by 12.1%, 44.9%, 0.04%, 48.2%, and 39.7%, respectively. To validate the advancement of the improved model, a comparison was made with the mainstream key point localization model, YOLOv8s-pose, which showed that the RMSE was reduced by 1.06 pixels compared to YOLOv8s-pose. This indicated that the ResNet-50 network integrated with the CBAM attention mechanism possessed superior localization accuracy. In the verification of the cow slippery hoof detection classification model, a 10-fold cross-validation was conducted to evaluate the performance of the cow slippery hoof classification model, resulting in average values of accuracy, precision, recall, and F1-Score at 90.42%, 0.943, 0.949, and 0.941, respectively. The error in the calculated slippery hoof distance of the cows, using the slippery hoof distance feature parameter Feature2, compared to the manually calibrated slippery hoof distance was found to be 1.363 pixels. [Conclusion] The ResNet-50 network model improved by integrating the CBAM module showed a high accuracy in the localization of key body points of cows. The cow slippery hoof judgment model and the cow slippery hoof distance prediction model, based on the extracted feature parameters for slippery hoof judgment and slippery hoof distance detection, both exhibited small errors when compared to manual detection results. This indicated that the proposed enhanced deeplabcut model obtained good accuracy and could provide technical support for the automatic detection of slippery hooves in cows.
[Significance] Rice smart unmanned farm is the core component of smart agriculture, and it is a key path to realize the modernization of rice production and promote the high-quality development of agriculture. Leveraging advanced information technologies such as the Internet of Things (IoT) and artificial intelligence (AI), these farms enable deep integration of data-driven decision making and intelligent machines. This integration creates an unmanned production system that covers the entire process from planting and managing rice crops to harvesting, greatly improving the efficiency and precision of rice cultivation. [Progress] This paper systematically sorted out the key technologies of rice smart unmanned farms in the three main links of pre-production, production and post-production, and the key technologies of pre-production mainly include the construction of high-standard farmland, unmanned nursery, land leveling, and soil nutrient testing. The construction of high-standard farmland is the foundation of the physical environment of the smart unmanned farms of rice, which provides perfect operating environment for the operation of modernized smart farm machinery through the reasonable layout of the field roads, good drainage and irrigation systems, and the scientific planting structure. Agricultural machine operation provides a perfect operating environment. The technical level of unmanned nursery directly determines the quality of rice cultivation and harvesting in the later stage, and a variety of rice seeding machines and nursery plate setting machines have been put into use. Land leveling technology can improve the growing environment of rice and increase the land utilization rate, and the current land leveling technology through digital sensing and path planning technology, which improves the operational efficiency and reduces the production cost at the same time. Soil nutrient detection technology is mainly detected by electrochemical analysis and spectral analysis, but both methods have their advantages and disadvantages, how to integrate the two methods to achieve an all-round detection of soil nutrient content is the main direction of future research. The key technologies in production mainly include rice dry direct seeding, automated transplanting, precise variable fertilization, intelligent irrigation, field weed management, and disease diagnosis. Among them, the rice dry direct seeding technology requires the planter to have high precision and stability to ensure reasonable seeding depth and density. Automated rice transplanting technology mainly includes three ways: root washing seedling machine transplanting, blanket seedling machine transplanting, and potting blanket seedling machine transplanting; at present, the incidence of problems in the automated transplanting process should be further reduced, and the quality and efficiency of rice machine transplanting should be improved. Precision variable fertilization technology is mainly composed of three key technologies: information perception, prescription decision-making and precise operation, but there are still fewer cases of unmanned farms combining the three technologies, and in the future, the main research should be on the method of constructing the whole process operation system of variable fertilization. The smart irrigation system is based on the water demand of the whole life cycle of rice to realize adaptive irrigation control, and the current smart irrigation technology can automatically adjust the irrigation strategy through real-time monitoring of soil, climate and crop growth conditions to further improve irrigation efficiency and agricultural production benefits. The field weed management and disease diagnosis technology mainly recognizes rice weeds as well as diseases through deep learning and other methods, and combines them with precision application technology for prevention and intervention. Post-production key technologies mainly include rice yield estimation, unmanned harvesting, rice storage and processing quality testing. Rice yield estimation technology is mainly used to predict yield by combining multi-source data and algorithms, but there are still problems such as the difficulty of integrating multi-source data, which requires further research. In terms of unmanned aircraft harvesting technology, China's rice combine harvester market has tended to stabilize, and the safety of the harvester's autopilot should be further improved in the future. Rice storage and processing quality detection technology mainly utilizes spectral technology and machine vision technology to detect spectra and images, and future research can combine deep learning and multimodal fusion technology to improve the machine vision system's ability and adaptability to recognize the appearance characteristics of rice. [Conclusions and Prospects] This paper reviews the researches of the construction of intelligent unmanned rice farms at home and abroad in recent years, summarizes the main difficulties faced by the key technologies of unmanned farms in practical applications, analyzes the challenges encountered in the construction of smart unmanned farms, summarizes the roles and responsibilities of the government, enterprises, scientific research institutions, cooperatives and other subjects in promoting the construction of intelligent unmanned rice farms, and puts forward relevant suggestions. It provides certain support and development ideas for the construction of intelligent unmanned rice farms in China.
[Significance] The rapid development of artificial intelligence and automation has greatly expanded the scope of agricultural automation, with applications such as precision farming using unmanned machinery, robotic grazing in outdoor environments, and automated harvesting by orchard-picking robots. Collaborative operations among multiple agricultural robots enhance production efficiency and reduce labor costs, driving the development of smart agriculture. Multi-robot simultaneous localization and mapping (SLAM) plays a pivotal role by ensuring accurate mapping and localization, which are essential for the effective management of unmanned farms. Compared to single-robot SLAM, multi-robot systems offer several advantages, including higher localization accuracy, larger sensing ranges, faster response times, and improved real-time performance. These capabilities are particularly valuable for completing complex tasks efficiently. However, deploying multi-robot SLAM in agricultural settings presents significant challenges. Dynamic environmental factors, such as crop growth, changing weather patterns, and livestock movement, increase system uncertainty. Additionally, agricultural terrains vary from open fields to irregular greenhouses, requiring robots to adjust their localization and path-planning strategies based on environmental conditions. Communication constraints, such as unstable signals or limited transmission range, further complicate coordination between robots. These combined challenges make it difficult to implement multi-robot SLAM effectively in agricultural environments. To unlock the full potential of multi-robot SLAM in agriculture, it is essential to develop optimized solutions that address the specific technical demands of these scenarios. [Progress] Existing review studies on multi-robot SLAM mainly focus on a general technological perspective, summarizing trends in the development of multi-robot SLAM, the advantages and limitations of algorithms, universally applicable conditions, and core issues of key technologies. However, there is a lack of analysis specifically addressing multi-robot SLAM under the characteristics of complex agricultural scenarios. This study focuses on the main features and applications of multi-robot SLAM in complex agricultural scenarios. The study analyzes the advantages and limitations of multi-robot SLAM, as well as its applicability and application scenarios in agriculture, focusing on four key components: multi-sensor data fusion, collaborative localization, collaborative map building, and loopback detection. From the perspective of collaborative operations in multi-robot SLAM, the study outlines the classification of SLAM frameworks, including three main collaborative types: centralized, distributed, and hybrid. Based on this, the study summarizes the advantages and limitations of mainstream multi-robot SLAM frameworks, along with typical scenarios in robotic agricultural operations where they are applicable. Additionally, it discusses key issues faced by multi-robot SLAM in complex agricultural scenarios, such as low accuracy in mapping and localization during multi-sensor fusion, restricted communication environments during multi-robot collaborative operations, and low accuracy in relative pose estimation between robots. [Conclusions and Prospects] To enhance the applicability and efficiency of multi-robot SLAM in complex agricultural scenarios, future research needs to focus on solving these critical technological issues. Firstly, the development of enhanced data fusion algorithms will facilitate improved integration of sensor information, leading to greater accuracy and robustness of the system. Secondly, the combination of deep learning and reinforcement learning techniques is expected to empower robots to better interpret environmental patterns, adapt to dynamic changes, and make more effective real-time decisions. Thirdly, large language models will enhance human-robot interaction by enabling natural language commands, improving collaborative operations. Finally, the integration of digital twin technology will support more intelligent path planning and decision-making processes, especially in unmanned farms and livestock management systems. The convergence of digital twin technology with SLAM is projected to yield innovative solutions for intelligent perception and is likely to play a transformative role in the realm of agricultural automation. This synergy is anticipated to revolutionize the approach to agricultural tasks, enhancing their efficiency and reducing the reliance on labor.
[Significance] In view of the lack of monitoring means of quality influence factors in the production process of characteristic agricultural products with in central and western regions of China, the weak ability of intelligent control, the unclear coupling relationship of quality control elements and the low degree of systematic application, the existing technologies described such as intelligent monitoring of facility environment, growth and nutrition intelligent control model, architecture of intelligent management and control platform and so on. Through the application of the Internet of Things, big data and the new generation of artificial intelligence technology, it provides technical support for the construction and application of intelligent process quality control system for the whole growth period of characteristic agricultural products. [Progress] The methods of environmental regulation and nutrition regulation are analyzed, including single parameters and combined control methods, such as light, temperature, humidity, CO2 concentration, fertilizer and water, etc. The multi-parameter coupling control method has the advantage of more comprehensive scene analysis. Based on the existing technology, a multi-factor coupling method of integrating growth state, agronomy, environment, input and agricultural work is put forward. This paper probes into the system architecture of the whole process service of quality control, the visual identification system of the growth process of agricultural products and the knowledge-driven agricultural technical service system, and introduces the technology of the team in the disease knowledge Q & A scene through multi-modal knowledge graph and large model technology. [Conclusions and Prospects] Based on the present situation of the production of characteristic facility agricultural products and the overall quality of farmers in the central and western regions of China, it is appropriate to transfer the whole technical system such as facility tomato, facility cucumber and so on. According to the varieties of characteristic agricultural products, cultivation models, quality control objectives to adapt to light, temperature, humidity and other parameters, as well as fertilizer, water, medicine and other input plans, a multi-factor coupling model suitable for a specific planting area is generated and long-term production verification and model correction are carried out. And popularize it in a wider area, making full use of the advantages of intelligent equipment and data elements will promote the realization of light simplification of production equipment, scene of intelligent technology, diversification of service models, on-line quality control, large-scale production of digital intelligence, and value of data elements, further cultivate facilities to produce new quality productivity.
[Objective] In the era of digital agriculture, real-time monitoring and predictive modeling of crop growth are paramount, especially in autonomous farming systems. Traditional crop growth models, often constrained by their reliance on static, rule-based methods, fail to capture the dynamic and multifactorial nature of vegetable crop growth. This research tried to address these challenges by leveraging the advanced reasoning capabilities of pre-trained large language models (LLMs) to simulate and predict vegetable crop growth with accuracy and reliability. Modeling the growth of vegetable crops within these platforms has historically been hindered by the complex interactions among biotic and abiotic factors. [Methods] The methodology was structured in several distinct phases. Initially, a comprehensive dataset was curated to include extensive information on vegetable crop growth cycles, environmental conditions, and management practices. This dataset incorporates continuous data streams such as soil moisture, nutrient levels, climate variables, pest occurrence, and historical growth records. By combining these data sources, the study ensured that the model was well-equipped to understand and infer the complex interdependencies inherent in crop growth processes. Then, advanced techniques was emploied for pre-training and fine-tuning LLMs to adapt them to the domain-specific requirements of vegetable crop modeling. A staged intelligent agent ensemble was designed to work within the digital twin platform, consisting of a central managerial agent and multiple stage-specific agents. The managerial agent was responsible for identifying transitions between distinct growth stages of the crops, while the stage-specific agents were tailored to handle the unique characteristics of each growth phase. This modular architecture enhanced the model's adaptability and precision, ensuring that each phase of growth received specialized attention and analysis. [Results and Discussions] The experimental validation of this method was conducted in a controlled agricultural setting at the Xiaotangshan Modern Agricultural Demonstration Park in Beijing. Cabbage (Zhonggan 21) was selected as the test crop due to its significance in agricultural production and the availability of comprehensive historical growth data. Over five years, the dataset collected included 4 300 detailed records, documenting parameters such as plant height, leaf count, soil conditions, irrigation schedules, fertilization practices, and pest management interventions. This dataset was used to train the LLM-based system and evaluate its performance using ten-fold cross-validation. The results of the experiments demonstrating the efficacy of the proposed system in addressing the complexities of vegetable crop growth modeling. The LLM-based model achieved 98% accuracy in predicting crop growth degrees and a 99.7% accuracy in identifying growth stages. These metrics significantly outperform traditional machine learning approaches, including long short-term memory (LSTM), XGBoost, and LightGBM models. The superior performance of the LLM-based system highlights its ability to reason over heterogeneous data inputs and make precise predictions, setting a new benchmark for crop modeling technologies. Beyond accuracy, the LLM-powered system also excels in its ability to simulate growth trajectories over extended periods, enabling farmers and agricultural managers to anticipate potential challenges and make proactive decisions. For example, by integrating real-time sensor data with historical patterns, the system can predict how changes in irrigation or fertilization practices will impact crop health and yield. This predictive capability is invaluable for optimizing resource allocation and mitigating risks associated with climate variability and pest outbreaks. [Conclusions] The study emphasizes the importance of high-quality data in achieving reliable and generalizable models. The comprehensive dataset used in this research not only captures the nuances of cabbage growth but also provides a blueprint for extending the model to other crops. In conclusion, this research demonstrates the transformative potential of combining large language models with digital twin technology for vegetable crop growth modeling. By addressing the limitations of traditional modeling approaches and harnessing the advanced reasoning capabilities of LLMs, the proposed system sets a new standard for precision agriculture. Several avenues also are proposed for future work, including expanding the dataset, refining the model architecture, and developing multi-crop and multi-region capabilities.
[Objective] Crop line extraction is critical for improving the efficiency of autonomous agricultural machines in the field. However, traditional detection methods struggle to maintain high accuracy and efficiency under challenging conditions, such as strong light exposure and weed interference. The aims are to develop an effective crop line extraction method by combining YOLOv8-G, Affinity Propagation, and the Least Squares method to enhance detection accuracy and performance in complex field environments. [Methods] The proposed method employs machine vision techniques to address common field challenges. YOLOv8-G, an improved object detection algorithm that combines YOLOv8 and GhostNetV2 for lightweight, high-speed performance, was used to detect the central points of crops. These points were then clustered using the Affinity Propagation algorithm, followed by the application of the Least Squares method to extract the crop lines. Comparative tests were conducted to evaluate multiple backbone networks within the YOLOv8 framework, and ablation studies were performed to validate the enhancements made in YOLOv8-G. [Results and Discussions] The performance of the proposed method was compared with classical object detection and clustering algorithms. The YOLOv8-G algorithm achieved average precision (AP) values of 98.22%, 98.15%, and 97.32% for corn detection at 7, 14, and 21 days after emergence, respectively. Additionally, the crop line extraction accuracy across all stages was 96.52%. These results demonstrate the model's ability to maintain high detection accuracy despite challenging conditions in the field. [Conclusions] The proposed crop line extraction method effectively addresses field challenges such as lighting and weed interference, enabling rapid and accurate crop identification. This approach supports the automatic navigation of agricultural machinery, offering significant improvements in the precision and efficiency of field operations.
[Objective] Spraying calcium can effectively prevent the occurrence of dry burning heart disease in Chinese cabbage. Accurately targeting spraying calcium can more effectively improve the utilization rate of calcium. Since the sprayer needs to move rapidly in the field, this can lead to over-application or under-application of the pesticide. This study aims to develop a targeted spray control system based on deep learning technology, explore the relationship between the advance speed, spray volume, and coverage of the sprayer, thereby addressing the uneven application issues caused by different nebulizer speeds by studying the real scenario of calcium administration to Chinese cabbage hearts. [Methods] The targeted spraying control system incorporates advanced sensors and computing equipment that were capable of obtaining real-time data regarding the location of crops and the surrounding environmental conditions. This data allowed for dynamic adjustments to be made to the spraying system, ensuring that pesticides were delivered with high precision. To further enhance the system's real-time performance and accuracy, the YOLOv8 object detection model was improved. A Ghost-Backbone lightweight network structure was introduced, integrating remote sensing technologies along with the sprayer's forward speed and the frequency of spray responses. This innovative combination resulted in the creation of a YOLOv8-Ghost-Backbone lightweight model specifically tailored for agricultural applications. The model operated on the Jetson Xavier NX controller, which was a high-performance, low-power computing platform designed for edge computing. The system was allowed to process complex tasks in real time directly in the field. The targeted spraying system was composed of two essential components: A pressure regulation unit and a targeted control unit. The pressure regulation unit was responsible for adjusting the pressure within the spraying system to ensure that the output remains stable under various operational conditions. Meanwhile, the targeted control unit played a crucial role in precisely controlling the direction, volume, and coverage of the spray to ensure that the pesticide was applied effectively to the intended areas of the plants. To rigorously evaluate the performance of the system, a series of intermittent spray tests were conducted. During these tests, the forward speed of the sprayer was gradually increased, allowing to assess how well the system responded to changes in speed. Throughout the testing phase, the response frequency of the electromagnetic valve was measured to calculate the corresponding spray volume for each nozzle. [Results and Conclusions] The experimental results indicated that the overall performance of the targeted spraying system was outstanding, particularly under conditions of high-speed operation. By meticulously recording the response times of the three primary components of the system, the valuable data were gathered. The average time required for image processing was determined to be 29.50 ms, while the transmission of decision signals took an average of 6.40 ms. The actual spraying process itself required 88.83 ms to complete. A thorough analysis of these times revealed that the total response time of the spraying system lagged by approximately 124.73 ms when compared to the electrical signal inputs. Despite the inherent delays, the system was able to maintain a high level of spraying accuracy by compensating for the response lag of the electromagnetic valve. Specifically, when tested at a speed of 7.2 km/h, the difference between the actual spray volume delivered and the required spray volume, after accounting for compensation, was found to be a mere 0.01 L/min. This minimal difference indicates that the system met the standard operational requirements for effective pesticide application, thereby demonstrating its precision and reliability in practical settings. [Conclusions] In conclusion, this study developed and validated a deep learning-based targeted spraying control system that exhibited excellent performance regarding both spraying accuracy and response speed. The system serves as a significant technical reference for future endeavors in agricultural automation. Moreover, the research provides insights into how to maintain consistent spraying effectiveness and optimize pesticide utilization efficiency by dynamically adjusting the spraying system as the operating speed varies. The findings of this research will offer valuable experiences and guidance for the implementation of agricultural robots in the precise application of pesticides, with a particular emphasis on parameter selection and system optimization.
Objective Rice diseases significantly impact both the yield and quality of rice production. Automatic recognition of rice diseases using computer vision is crucial for ensuring high yields, quality, and efficiency. However, rice disease image recognition faces challenges such as limited availability of datasets, insufficient sample sizes, and imbalanced sample distributions across different disease categories. To address these challenges, a data augmentation method for rice leaf disease images was proposed based on an improved CycleGAN model in this reseach which aimed to expand disease image datasets by generating disease features, thereby alleviating the burden of collecting real disease data and providing more comprehensive and diverse data to support automatic rice disease recognition. Methods The proposed approach built upon the CycleGAN framework, with a key modification being the integration of a convolutional block attention module (CBAM) into the generator's residual module. This enhancement strengthened the network's ability to extract both local key features and global contextual information pertaining to rice disease-affected areas. The model increased its sensitivity to small-scale disease targets and subtle variations between healthy and diseased domains. This design effectively mitigated the potential loss of critical feature information during the image generation process, ensuring higher fidelity in the resulting images. Additionally, skip connections were introduced between the residual modules and the CBAM. These connections facilitate improved information flow between different layers of the network, addressing common issues such as gradient vanishing during the training of deep networks. Furthermore, a perception similarity loss function, designed to align with the human visual system, was incorporated into the overall loss function. This addition enabled the deep learning model to more accurately measure perceptual differences between the generated images and real images, thereby guiding the network towards producing higher-quality samples. This adjustment also helped to reduce visual artifacts and excessive smoothing, while concurrently improving the stability of the model during the training process. To comprehensively evaluate the quality of the rice disease images generated by the proposed model and to assess its impact on disease recognition performance, both subjective and objective evaluation metrics were utilized. These included user perception evaluation (UPE), structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), and the performance of disease recognition within object detection frameworks. Comparative experiments were conducted across multiple GAN models, enabling a thorough assessment of the proposed model's performance in generating rice disease images. Additionally, different attention mechanisms, including efficient channel attention (ECA), coordinate attention (CA), and CBAM, were individually embedded into the generator's residual module. These variations allowed for a detailed comparison of the effects of different attention mechanisms on network performance and the visual quality of the generated images. Ablation studies were further performed to validate the effectiveness of the CBAM residual module and the perception similarity loss function in the network's overall architecture. Based on the generated rice disease samples, transfer learning experiments were conducted using various object detection models. By comparing the performance of these models before and after transfer learning, the effectiveness of the generated disease image data in enhancing the performance of object detection models was empirically verified. Results and Discussions The rice disease images generated by the improved CycleGAN model surpassed those produced by other GAN variants in terms of image detail clarity and the prominence of disease-specific features. In terms of objective quality metrics, the proposed model exhibited a 3.15% improvement in SSIM and an 8.19% enhancement in PSNR compared to the original CycleGAN model, underscoring its significant advantage in structural similarity and signal-to-noise ratio. The comparative experiments involving different attention mechanisms and ablation studies revealed that embedding the CBAM into the generator effectively increased the network's focus on critical disease-related features, resulting in more realistic and clearly defined disease-affected regions in the generated images. Furthermore, the introduction of the perception similarity loss function substantially enhanced the network's ability to perceive and represent disease-related information, thereby improving the visual fidelity and realism of the generated images. Additionally, transfer learning applied to object detection models such as YOLOv5s, YOLOv7-tiny, and YOLOv8s led to significant improvements in disease detection performance on the augmented dataset. Notably, the detection accuracy of the YOLOv5s model increased from 79.7% to 93.8%, representing a considerable enhancement in both generalization ability and robustness. This improvement also effectively reduced the rates of false positives and false negatives, resulting in more stable and reliable performance in rice disease detection tasks. Conclusions The rice leaf disease image generation method based on the improved CycleGAN model, as proposed in this study, effectively transforms images of healthy leaves into those depicting disease symptoms. By addressing the challenge of insufficient disease samples, this method significantly improves the disease recognition capabilities of object detection models. Therefore, it holds considerable application potential in the domain of leaf disease image augmentation and offers a promising new direction for expanding datasets of disease images for other crops.
[Objective] Crop leaf area is an important indicator reflecting light absorption efficiency and growth conditions. This paper established a diverse Chinese yam image dataset and proposesd a deep learning-based method for Chinese yam leaf image segmentation. This method can be used for real-time measurement of Chinese yam leaf area, addressing the inefficiency of traditional measurement techniques. This will provide more reliable data support for genetic breeding, growth and development research of Chinese yam, and promote the development and progress of the Chinese yam industry. [Methods] A lightweight segmentation network based on improved ENet was proposed. Firstly, based on ENet, the third stage was pruned to reduce redundant calculations in the model. This improved the computational efficiency and running speed, and provided a good basis for real-time applications. Secondly, PConv was used instead of the conventional convolution in the downsampling bottleneck structure and conventional bottleneck structure, the improved bottleneck structure was named P-Bottleneck. PConv applied conventional convolution to only a portion of the input channels and left the rest of the channels unchanged, which reduced memory accesses and redundant computations for more efficient spatial feature extraction. PConv was used to reduce the amount of model computation while increase the number of floating-point operations per second on the hardware device, resulting in lower latency. Additionally, the transposed convolution in the upsampling module was improved to bilinear interpolation to enhance model accuracy and reduce the number of parameters. Bilinear interpolation could process images smoother, making the processed images more realistic and clear. Finally, coordinate attention (CA) module was added to the encoder to introduce the attention mechanism, and the model was named CBPA-ENet. The CA mechanism not only focused on the channel information, but also keenly captured the orientation and position-sensitive information. The position information was embedded into the channel attention to globally encode the spatial information, capturing the channel information along one spatial direction while retaining the position information along the other spatial direction. The network could effectively enhance the attention to important regions in the image, and thus improve the quality and interpretability of segmentation results. [Results and Discussions] Trimming the third part resulted in a 28% decrease in FLOPs, a 41% decrease in parameters, and a 9 f/s increase in FPS. Improving the upsampling method to bilinear interpolation not only reduces the floating-point operation and parameters, but also slightly improves the segmentation accuracy of the model, increasing FPS by 4 f/s. Using P-Bottleneck instead of downsampling bottleneck structure and conventional bottleneck structure can reduce mIoU by only 0.04%, reduce FLOPs by 22%, reduce parameters by 16%, and increase FPS by 8 f/s. Adding CA mechanism to the encoder could only increase a small amount of FLOPs and parameters, improving the accuracy of the segmentation network. To verify the effectiveness of the improved segmentation algorithm, classic semantic segmentation networks of UNet, DeepLabV3+, PSPNet, and real-time semantic segmentation network LinkNet, DABNet were selected to train and validate. These six algorithms got quite high segmentation accuracy, among which UNet had the best mIoU and the mPA, but the model size was too large. The improved algorithm only accounts for 1% of the FLOPs and 0.41% of the parameters of UNet, and the mIoU and mPA were basically the same. Other classic semantic segmentation algorithms, such as DeepLabV3+, had similar accuracy to improved algorithms, but their large model size and slow inference speed were not conducive to embedded development. Although the real-time semantic segmentation algorithm LinkNet had a slightly higher mIoU, its FLOPs and parameters count were still far greater than the improved algorithm. Although the PSPNet model was relatively small, it was also much higher than the improved algorithm, and the mIoU and mPA were lower than the algorithm. The experimental results showed that the improved model achieved a mIoU of 98.61%. Compared with the original model, the number of parameters and FLOPs significantly decreased. Among them, the number of model parameters decreased by 51%, the FLOPs decreased by 49%, and the network operation speed increased by 38%. [Conclusions] The improved algorithm can accurately and quickly segment Chinese yam leaves, providing not only a more accurate means for determining Chinese yam phenotype data, but also a new method and approach for embedded research of Chinese yam. Using the model, the morphological feature data of Chinese yam leaves can be obtained more efficiently, providing a reliable foundation for further research and analysis.
[Objective] Grape picking is a key link in increasing production. However, in this process, a large amount of manpower and material resources are required, which makes the picking process complex and slow. To enhance harvesting efficiency and achieve automated grape harvesting, an improved YOLOv8n object detection model named 3C-YOLOv8n was proposed, which integrates the RealSense D415 depth camera for grape recognition and localization. [Methods] The propoesed 3C-YOLOv8n incorporated a convolutional block attention module (CBAM) between the first C2f module and the third Conv module in the backbone network. Additionally, a channel attention (CA) module was added at the end of the backbone structure, resulting in a new 2C-C2f backbone network architecture. This design enabled the model to sequentially infer attention maps across two independent dimensions (channel and spatial), optimize features by considering relationships between channels and positional information. The network structure was both flexible and lightweight. Furthermore, the Content-aware ReAssembly of Features up sampling operator was implemented to support instance-specific kernels (such as deconvolution) for feature reconstruction with neighboring pixels, replacing the nearest neighbor interpolation operator in the YOLOv8n neck network. This enhancement increased the receptive field and guided the reconstruction process based on input features while maintaining low parameter and computational complexity, thereby forming the 3C-YOLOv8n model. The pyrealsense2 library was utilized to obtain pixel position information from the target area using the Intel RealSense D415 camera. During this process, the depth camera was used to capture images, and target detection algorithms were employed to pinpoint the location of grapes. The camera's depth sensor facilitated the acquisition of the three-dimensional point cloud of grapes, allowing for the calculation of the distance from the pixel point to the camera and the subsequent determination of the three-dimensional coordinates of the center of the target's bounding box in the camera coordinate system, thus achieving grape recognition and localization. [Results and Discussions] Comparative and ablation experiments were conducted. it was observed that the 3C-YOLOv8n model achieved a mean average precision (mAP) of 94.3% at an intersection ratio of 0.5 (IOU=0.5), surpassing the YOLOv8n model by 1%. The accuracy (P) and recall (R) rates were recorded at 91.6% and 86.4%, respectively, reflecting increases of 0.1% and 0.7%. The F1-Score also improved by 0.4%, demonstrating that the improved network model met the experimental accuracy and recall requirements. In terms of loss, the 3C-YOLOv8n algorithm exhibited superior performance, with a rapid decrease in loss values and minimal fluctuations, ultimately leading to a minimized loss value. This indicated that the improved algorithm quickly reached a convergence state, enhancing both model accuracy and convergence speed. The ablation experiments revealed that the original YOLOv8n model yielded a mAP of 93.3%. The integration of the CBAM and CA attention mechanisms into the YOLOv8n backbone resulted in mAP values of 93.5% each. The addition of the Content-aware ReAssembly of Features up sampling operator to the neck network of YOLOv8n produced a 0.5% increase in mAP, culminating in a value of 93.8%. The combination of the three improvement strategies yielded mAP increases of 0.3, 0.7, and 0.8%, respectively, compared to the YOLOv8n model. Overall, the 3C-YOLOv8n model demonstrated the best detection performance, achieving the highest mAP of 94.3%. The ablation results confirmed the positive impact of the proposed improvement strategies on the experimental outcomes. Compared to other mainstream YOLO series algorithms, all evaluation metrics showed enhancements, with the lowest missed detection and false detection rates among all tested algorithms, underscoring its practical advantages in detection tasks. [Conclusions] By effectively addressing the inefficiencies of manual labor, 3C-YOLOv8n network model not only enhances the precision of grape recognition and localization but also significantly optimizes overall harvesting efficiency. Its superior performance in evaluation metrics such as precision, recall, mAP, and F1-Score, alongside the lowest recorded loss values among YOLO series algorithms, indicates a remarkable advancement in model convergence and operational effectiveness. Furthermore, the model's high accuracy in grape target recognition not only lays the groundwork for automated harvesting systems but also enables the implementation of complementary intelligent operations.
[Objective] Plug tray seedling cultivation is a contemporary method known for its high germination rates, uniform seedling growth, shortened transplant recovery period, diminished pest and disease incidence, and enhanced labor efficiency. Despite these advantages, challenges such as missing or underdeveloped seedlings can arise due to seedling quality and environmental factors. To ensure uniformity and consistency of the seedlings, sorting is frequently necessary, and the adoption of automated seedling sorting technology can significantly reduce labor costs. Nevertheless, the overgrowth of seedlings within the plugs can effect the accuracy of detection algorithms. A method for grading and locating strawberry seedlings based on a lightweight YOLOv8s model was presented in this research to effectively mitigate the interference caused by overgrown seedlings. [Methods] The YOLOv8s model was selected as the baseline for detecting different categories of seedlings in the strawberry plug tray cultivation process, namely weak seedlings, normal seedlings, and plug holes. To improve the detection efficiency and reduce the model's computational cost, the layer-adaptive magnitude-based pruning(LAMP) score-based channel pruning algorithm was applied to compress the base YOLOv8s model. The pruning procedure involved using the dependency graph to derive the group matrices, followed by normalizing the group importance scores using the LAMP Score, and ultimately pruning the channels according to these processed scores. This pruning strategy effectively reduced the number of model parameters and the overall size of the model, thereby significantly enhancing its inference speed while maintaining the capability to accurately detect both seedlings and plug holes. Furthermore, a two-stage seedling-hole matching algorithm was introduced based on the pruned YOLOv8s model. In the first stage, seedling and plug hole bounding boxes were matched according to their the degree of overlap (Dp), resulting in an initial set of high-quality matches. This step helped minimize the number of potential matching holes for seedlings exhibiting overgrowth. Subsequently, before the second stage of matching, the remaining unmatched seedlings were ranked according to their potential matching hole scores (S), with higher scores indicating fewer potential matching holes. The seedlings were then prioritized during the second round of matching based on these scores, thus ensuring an accurate pairing of each seedling with its corresponding plug hole, even in cases where adjacent seedling leaves encroached into neighboring plug holes. [Results and Discussions] The pruning process inevitably resulted in the loss of some parameters that were originally beneficial for feature representation and model generalization. This led to a noticeable decline in model performance. However, through meticulous fine-tuning, the model's feature expression capabilities were restored, compensating for the information loss caused by pruning. Experimental results demonstrated that the fine-tuned model not only maintained high detection accuracy but also achieved significant reductions in FLOPs (86.3%) and parameter count (95.4%). The final model size was only 1.2 MB. Compared to the original YOLOv8s model, the pruned version showed improvements in several key performance metrics: precision increased by 0.4%, recall by 1.2%, mAP by 1%, and the F1-Score by 0.1%. The impact of the pruning rate on model performance was found to be non-linear. As the pruning rate increased, model performance dropped significantly after certain crucial channels were removed. However, further pruning led to a reallocation of the remaining channels' weights, which in some cases allowed the model to recover or even exceed its previous performance levels. Consequently, it was necessary to experiment extensively to identify the optimal pruning rate that balanced model accuracy and speed. The experiments indicated that when the pruning rate reached 85.7%, the mAP peaked at 96.4%. Beyond this point, performance began to decline, suggesting that this was the optimal pruning rate for achieving a balance between model efficiency and performance, resulting in a model size of 1.2 MB. To further validate the improved model's effectiveness, comparisons were conducted with different lightweight backbone networks, including MobileNetv3, ShuffleNetv2, EfficientViT, and FasterNet, while retaining the Neck and Head modules of the original YOLOv8s model. Results indicated that the modified model outperformed these alternatives, with mAP improvements of 1.3%, 1.8%, 1.5%, and 1.1%, respectively, and F1-Score increases of 1.5%, 1.8%, 1.1%, and 1%. Moreover, the pruned model showed substantial advantages in terms of floating-point operations, model size, and parameter count compared to these other lightweight networks. To verify the effectiveness of the proposed two-stage seedling-hole matching algorithm, tests were conducted using a variety of complex images from the test set. Results indicated that the proposed method achieved precise grading and localization of strawberry seedlings even under challenging overgrowth conditions. Specifically, the correct matching rate for normal seedlings reached 96.6%, for missing seedlings 84.5%, and for weak seedlings 82.9%, with an average matching accuracy of 88%, meeting the practical requirements of the strawberry plug tray cultivation process. [Conclusions] The pruned YOLOv8s model successfully maintained high detection accuracy while reducing computational costs and improving inference speed. The proposed two-stage seedling-hole matching algorithm effectively minimized the interference caused by overgrown seedlings, accurately locating and classifying seedlings of various growth stages within the plug tray. The research provides a robust and reliable technical solution for automated strawberry seedling sorting in practical plug tray cultivation scenarios, offering valuable insights and technical support for optimizing the efficiency and precision of automated seedling grading systems.
[Objective] Apple Alternaria leaf spot can easily lead to premature defoliation of apple tree leaves, thereby affecting the quality and yield of apples. Consequently, accurately detecting of the disease has become a critical issue in the precise prevention and control of apple tree diseases. Due to factors such as backlighting, traditional image segmentation-based methods for detecting disease spots struggle to accurately identify the boundaries of diseased areas against complex backgrounds. There is an urgent need to develop new methods for detecting apple Alternaria leaf spot, which can assist in the precise prevention and control of apple tree diseases. [Methods] A novel detection method named Deep Semi-Non-negative Matrix Factorization-based Mahalanobis Distance Anomaly Detection (DSNMFMAD) was proposed, which combines Deep Semi-Non-negative Matrix Factorization (DSNMF) with Mahalanobis distance for robust anomaly detection in complex image backgrounds. The proposed method began by utilizing DSNMF to extract low-rank background components and sparse anomaly features from the apple Alternaria leaf spot images. This enabled effective separation of the background and anomalies, mitigating interference from complex background noise while preserving the non-negativity constraints inherent in the data. Subsequently, Mahalanobis distance was employed, based on the Singular Value Decomposition (SVD) feature subspace, to construct a lesion detector. The detector identified lesions by calculating the anomaly degree of each pixel in the anomalous regions. The apple tree leaf disease dataset used was provided by PaddlePaddle AI-Studio. Each image in the dataset has a resolution of 512×512 pixels, in RGB color format, and was in JPEG format. The dataset was captured in both laboratory and natural environments. Under laboratory conditions, 190 images of apple leaves with spot-induced leaf drop were used, while 237 images were collected under natural conditions. Furthermore, the dataset was augmented with geometric transformations and random changes in brightness, contrast, and hue, resulting in 1 145 images under laboratory conditions and 1 419 images under natural conditions. These images reflect various real-world scenarios, capturing apple leaves at different stages of maturity, in diverse lighting conditions, angles, and noise environments. This diversed dataset ensured that the proposed method could be tested under a wide range of practical conditions, providing a comprehensive evaluation of its effectiveness in detecting apple Alternaria leaf spot. [Results and Discussions] DSNMFMAD demonstrated outstanding performance under both laboratory and natural conditions. A comparative analysis was conducted with several other detection methods, including GRX (Reed-Xiaoli detector), LRX (Local Reed-Xiaoli detector), CRD (Collaborative-Representation-Based Detector), LSMAD (LRaSMD-Based Mahalanobis Distance Detector), and the deep learning model Unet. The results demonstrated that DSNMFMAD exhibited superior performance in the laboratory environment. The results demonstrated that DSNMFMAD attained a recognition accuracy of 99.8% and a detection speed of 0.087 2 s/image. The accuracy of DSNMFMAD was found to exceed that of GRX, LRX, CRD, LSMAD, and Unet by 0.2%, 37.9%, 10.3%, 0.4%, and 24.5%, respectively. Additionally, the DSNMFMAD exhibited a substantially superior detection speed in comparison to LRX, CRD, LSMAD, and Unet, with an improvement of 8.864, 107.185, 0.309, and 1.565 s, respectively. In a natural environment, where a dataset of 1 419 images of apple Alternaria leaf spot was analysed, DSNMFMAD demonstrated an 87.8% recognition accuracy, with an average detection speed of 0.091 0 s per image. In this case, its accuracy outperformed that of GRX, LRX, CRD, LSMAD, and Unet by 2.5%, 32.7%, 5%, 14.8%, and 3.5%, respectively. Furthermore, the detection speed was faster than that of LRX, CRD, LSMAD, and Unet by 2.898, 132.017, 0.224, and 1.825 s, respectively. [Conclusions] The DSNMFMAD proposed in this study was capable of effectively extracting anomalous parts of an image through DSNMF and accurately detecting the location of apple Alternaria leaf spot using a constructed lesion detector. This method achieved higher detection accuracy compared to the benchmark methods, even under complex background conditions, demonstrating excellent performance in lesion detection. This advancement could provide a valuable technical reference for the detection and prevention of apple Alternaria leaf spot.
[Objective] During the feeding process of fish populations in aquaculture, the video image characteristics of floating extruded feed on the water surface undergo continuous variations due to a myriad of environmental factors and fish behaviors. These variations pose significant challenges to the accurate detection of feed particles, which is crucial for effective feeding management. To address these challenges and enhance the detection of floating extruded feed particles on the water surface, ,thereby providing precise decision support for intelligent feeding in intensive aquaculture modes, the YOLOv11-AP2S model, an advanced detection model was proposed. [Methods] The YOLOv11-AP2S model enhanced the YOLOv11 algorithm by incorporating a series of improvements to its backbone network, neck, and head components. Specifically, an attention for fine-grained categorization (AFGC) mechanism was introduced after the 10th layer C2PSA of the backbone network. This mechanism aimed to boost the model's capability to capture fine-grained features, which were essential for accurately identifying feed particles in complex environments with low contrast and overlapping objects. Furthermore, the C3k2 module was replaced with the VoV-GSCSP module, which incorporated more sophisticated feature extraction and fusion mechanisms. This replacement further enhanced the network's ability to extract relevant features and improve detection accuracy. To improve the model's detection of small targets, a P2 layer was introduced. However, adding a P2 layer may increase computational complexity and resource consumption, so the overall performance and resource consumption of the model must be carefully balanced. To maintain the model's real-time performance while improving detection accuracy, a lightweight VoV-GSCSP module was utilized for feature fusion at the P2 layer. This approach enabled the YOLOv11-AP2S model to achieve high detection accuracy without sacrificing detection speed or model lightweights, making it suitable for real-time applications in aquaculture. [Results and Discussions] The ablation experimental results demonstrated the superiority of the YOLOv11-AP2S model over the original YOLOv11 network. Specifically, the YOLOv11-AP2S model achieved a precision ( P) and recall ( R) of 78.70%. The mean average precision (mAP50) at an intersection over union (IoU) threshold of 0.5 was as high as 80.00%, and the F1-Score had also reached 79.00%. These metrics represented significant improvements of 6.7%, 9.0%, 9.4% (for precision, as previously mentioned), and 8.0%, respectively, over the original YOLOv11 network. These improvements showed the effectiveness of the YOLOv11-AP2S model in detecting floating extruded feed particles in complex environments. When compared to other YOLO models, the YOLOv11-AP2S model exhibits clear advantages in detecting floating extruded feed images on a self-made dataset. Notably, under the same number of iterations, the YOLOv11-AP2S model achieved higher mAP50 values and lower losses, demonstrating its superiority in detection performance. This indicated that the YOLOv11-AP2S model strikes a good balance between learning speed and network performance, enabling it to efficiently and accurately detect images of floating extruded feed on the water surface. Furthermore, the YOLOv11-AP2S model's ability to handle complex detection scenarios, such as overlapping and adhesion of feed particles and occlusion by bubbles, was noteworthy. These capabilities were crucial for accurate detection in practical aquaculture environments, where such challenges were common and can significantly impair the performance of traditional detection systems. The improvements in detection accuracy and efficiency made the YOLOv11-AP2S model a valuable tool for intelligent feeding systems in aquaculture, as it could provide more reliable and timely information on fish feeding behavior. Additionally, the introduction of the P2 layer and the use of the lightweight VoV-GSCSP module for feature fusion at this layer contributed to the model's overall performance. These enhancements enabled the model to maintain high detection accuracy while keeping computational costs and resource consumption within manageable limits. This was particularly important for real-time applications in aquaculture, where both accuracy and efficiency were critical for effective feeding management. [Conclusions] The successful application of the YOLOv11-AP2S model in detecting floating extruded feed particles demonstrates its potential to intelligent feeding systems in aquaculture. By providing accurate and timely information on fish feeding behavior, the model can help optimize feeding strategies, reduce feed waste, and improve the overall efficiency and profitability of aquaculture operations. Furthermore, the model's ability to handle complex detection scenarios and maintain high detection accuracy while keeping computational costs within manageable limits makes it a practical and valuable tool for real-time applications in aquaculture. Therefore, the YOLOv11-AP2S model holds promise for wide application in intelligent aquaculture management, contributing to the sustainability and growth of the aquaculture industry.
[Objective] In order to summarize exemplary cases of high-quality development in regional smart agriculture and contribute strategies for the sustainable advancement of the national smart agriculture cause, the spatiotemporal characteristics and key driving factors of smart farms in the Yangtze River Economic Belt were studied. [Methods] Based on data from 11 provinces (municipalities) spanning the years 2014 to 2023, a comprehensive analysis was conducted on the spatio-temporal differentiation characteristics of smart farms in the Yangtze River Economic Belt using methods such as kernel density analysis, spatial auto-correlation analysis, and standard deviation ellipse. Including the overall spatial clustering characteristics, high-value or low-value clustering phenomena, centroid characteristics, and dynamic change trends. Subsequently, the geographic detector was employed to identify the key factors driving the spatio-temporal differentiation of smart farms and to discern the interactions between different factors. The analysis was conducted across seven dimensions: special fiscal support, industry dependence, human capital, urbanization, agricultural mechanization, internet infrastructure, and technological innovation. [Results and Discussions] Firstly, in terms of temporal characteristics, the number of smart farms in the Yangtze River Economic Belt steadily increased over the past decade. The year 2016 marked a significant turning point, after which the growth rate of smart farms had accelerated noticeably. The development of the upper, middle, and lower reaches exhibited both commonalities and disparities. Specifically, the lower sub-regions got a higher overall development level of smart farms, with a fluctuating upward growth rate; the middle sub-regions were at a moderate level, showing a fluctuating upward growth rate and relatively even provincial distribution; the upper sub-regions got a low development level, with a stable and slow growth rate, and an unbalanced provincial distribution. Secondly, in terms of spatial distribution, smart farms in the Yangtze River Economic Belt exhibited a dispersed agglomeration pattern. The results of global auto-correlation indicated that smart farms in the Yangtze River Economic Belt tended to be randomly distributed. The results of local auto-correlation showed that the predominant patterns of agglomeration were H-L and L-H types, with the distribution across provinces being somewhat complex; H-H type agglomeration areas were mainly concentrated in Sichuan, Hubei, and Anhui; L-L type agglomeration areas were primarily in Yunnan and Guizhou. The standard deviation ellipse results revealed that the mean center of smart farms in the Yangtze River Economic Belt had shifted from Anqing city in Anhui province in 2014 to Jingzhou city in Hubei province in 2023, with the spatial distribution showing an overall trend of shifting southwestward and a slow expansion toward the northeast and south. Finally, in terms of key driving factors, technological innovation was the primary critical factor driving the formation of the spatio-temporal distribution pattern of smart farms in the Yangtze River Economic Belt, with a factor explanatory degree of 0.311 1. Moreover, after interacting with other indicators, it continued to play a crucial role in the spatio-temporal distribution of smart farms, which aligned with the practical logic of smart farm development. Urbanization and agricultural mechanization levels were the second and third largest key factors, with factor explanatory degrees of 0.292 2 and 0.251 4, respectively. The key driving factors for the spatio-temporal differentiation of smart farms in the upper, middle, and lower sub-regions exhibited both commonalities and differences. Specifically, the top two key factors driver identification in the upper region were technological innovation (0.841 9) and special fiscal support (0.782 3). In the middle region, they were technological innovation (0.619 0) and human capital (0.600 1), while in the lower region, they were urbanization (0.727 6) and technological innovation (0.425 4). The identification of key driving factors and the detection of their interactive effects further confirmed that the spatio-temporal distribution characteristics of smart farms in the Yangtze River Economic Belt were the result of the comprehensive action of multiple factors. [Conclusions] The development of smart farms in the Yangtze River Economic Belt is showing a positive momentum, with both the total number of smart farms and the number of sub-regions experiencing stable growth. The development speed and level of smart farms in the sub-regions exhibit a differentiated characteristic of "lower reaches > middle reaches > upper reaches". At the same time, the overall distribution of smart farms in the Yangtze River Economic Belt is relatively balanced, with the degree of sub-regional distribution balance being "middle reaches (Hubei province, Hunan province, Jiangxi province are balanced) > lower reaches (dominated by Anhui) > upper reaches (Sichuan stands out)". The coverage of smart farm site selection continues to expand, forming a "northeast-southwest" horizontal diffusion pattern. In addition, the spatio-temporal characteristics of smart farms in the Yangtze River Economic Belt are the result of the comprehensive action of multiple factors, with the explanatory power of factors ranked from high to low as follows: Technological innovation > urbanization > agricultural mechanization > human capital > internet infrastructure > industry dependence > special fiscal support. Moreover, the influence of each factor is further strengthened after interaction. Based on these conclusions, suggestions are proposed to promote the high-quality development of smart farms in the Yangtze River Economic Belt. This study not only provides a theoretical basis and reference for the construction of smart farms in the Yangtze River Economic Belt and other regions, but also helps to grasp the current status and future trends of smart farm development.
