Shrimp Diseases Detection Method Based on Improved YOLOv8 and Multiple Features

doi:10.12133/j.smartag.SA201311014

Abstract

Abstract:

Objective Shrimp farming holds a pivotal position within the aquaculture industry. However, in recent years, there has been a steady rise in the incidence and mortality of shrimp diseases, leading to significant repercussions in shrimp farming. These diseases are characterized by their rapid onset, high contagion, challenging control, and high mortality rates. As shrimp factory farming continues to expand, manual detection methods are no longer sufficient to meet current demands. Consequently, there is an urgent need for an automated method to detect shrimp diseases. The primary objective of this study is to develop a cost-effective inspection method based on computer vision that strikes a balance between cost and detection effectiveness. Methods The improved YOLOv8 (You Only Look Once) network and multiple features were employed to detect shrimp diseases. To eliminate the interference of surface foam, the improved YOLOv8 network was applied to detect and extract surface shrimps as the foreground of the image. This target detection method accurately identifies objects of interest in the image, determining their category and location, and its extraction effect surpasses that of threshold segmentation. Considering the cost constraints of platform computing power in practical production conditions, the network was enhanced by reducing parameters and calculations, thereby improving detection speed and deployment performance. Additionally, the Farnberck optical flow method and gray level co-occurrence matrix (GLCM) were employed to extract 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. Finally, a classifier was employed to classify the multiple feature parameters in video clips, enabling 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%. 150 normal fragments and 150 diseased fragments were randomly selected as test samples. The classifier exhibited a detection accuracy of 89% on this dataset of 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.

Key words: shrimp diseases, computer vision, YOLOv8, Farnberck optical flow, gray level co-occurrence matrix, support vector machine

XU Ruifeng, WANG Yaohua, DING Wenyong, YU Junqi, YAN Maocang, CHEN Chen. Shrimp Diseases Detection Method Based on Improved YOLOv8 and Multiple Features[J]. Smart Agriculture, doi: 10.12133/j.smartag.SA201311014.

Figures/Tables 12

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Table 1

Table 2

Table 3

Fig. 7

Table 4

Fig. 8

References 31

1	YU Y B, CHOI J H, KANG J C, et al. Shrimp bacterial and parasitic disease listed in the OIE: A review[J]. Microbial pathogenesis, 2022, 166: ID 105545.
2	LEE D, YU Y B, CHOI J H, et al. Viral shrimp diseases listed by the OIE: A review[J]. Viruses, 2022, 14(3): ID 585.
3	蔡欣欣. 池塘养殖凡纳滨对虾急性肝胰腺坏死病(AHPND)的预警数学模型构建[D]. 上海: 上海海洋大学, 2021.
	CAI X X. Early warning mathematical model of acute hepatopancreatic necrosis disease (AHPND) in pond cultured shrimp Litopenaeus vannamei [D]. Shanghai: Shanghai Ocean University, 2021.
4	KHIEM N M, TAKAHASHI Y, YASUMA H, et al. Use of GIS and machine learning to predict disease in shrimp farmed on the east coast of the Mekong Delta, Vietnam[J]. Fisheries science, 2022, 88(1): 1-13.
5	TUYEN T T, AL-ANSARI N, NGUYEN D D, et al. Prediction of white spot disease susceptibility in shrimps using decision trees based machine learning models[J]. Applied water science, 2023, 14(1): 1-15.
6	孙世玉, 江敏, 金若晨, 等. 凡纳滨对虾池塘养殖过程中水质与虾虹彩病毒病发生的相关性[J]. 上海海洋大学学报, 2020, 29(5): 641-649.
	SUN S Y, JIANG M, JIN R C, et al. Correlation between water quality and shrimp hemocyte iridescent virus disease occurrence of Litopenaeus vannamei in ponds[J]. Journal of Shanghai ocean university, 2020, 29(5): 641-649.
7	RAMACHANDRAN L, MOHAN V. A novel neural network model for shrimp segmentation to detect white spot syndrome[J]. Journal of intelligent & fuzzy systems: Applications in engineering and technology, 2022, 43(1): 1453-1466.
8	WANG Q P, QIAN C, NIE P C, et al. Rapid detection of Penaeus vannamei diseases via an improved LeNet[J]. Aquacultural engineering, 2023, 100: ID 102296.
9	刘子豪, 张素兰, 贾小军, 等. 基于自适应卷积神经网络的染病虾识别方法[J]. 农业机械学报, 2022, 53(5): 246-256.
	LIU Z H, ZHANG S L, JIA X J, et al. Diseased shrimp identification method based on adaptive convolutional neural networks[J]. Transactions of the Chinese society for agricultural machinery, 2022, 53(5): 246-256.
10	JIANG B, SONG H B, WANG H, et al. Dairy cow lameness detection using a back curvature feature[J]. Computers and electronics in agriculture, 2022, 194: ID 106729.
11	CHEN J X, ZHOU J, LIU L S, et al. Sow farrowing early warning and supervision for embedded board implementations[J]. Sensors, 2023, 23(2): ID 727.
12	张然. 蛋鸡啄羽现象识别及羽毛覆盖度计算研究[D]. 保定: 河北农业大学, 2023.
	ZHANG R. Study on identification of feather pecking phenomenon and calculation of feather coverage in laying hens[D]. Baoding: Hebei Agricultural University, 2023.
13	XU Y L, NIE J, CEN H L, et al. Spatio-temporal-based identification of aggressive behavior in group sheep[J]. Animals: An open access journal from MDPI, 2023, 13(16): ID 2636.
14	位威. 高密度水产养殖鱼视觉监测技术研究[D]. 福州: 福建工程学院, 2022.
	WEI W. Research on visual monitoring technology of high-density aquaculture fish[D]. Fuzhou: Fujian University of Technology, 2022.
15	周超, 徐大明, 吝凯, 等. 基于近红外机器视觉的鱼类摄食强度评估方法研究[J]. 智慧农业, 2019, 1(1): 76-84.
	ZHOU C, XU D M, LIN K, et al. Evaluation of fish feeding intensity in aquaculture based on near-infrared machine vision[J]. Smart agriculture, 2019, 1(1): 76-84.
16	刘朝阳. 室外养殖池塘智能投饵管控系统设计[D]. 天津: 天津科技大学, 2022.
	LIU C Y. Design of intelligent feeding contorl system for outdoor aquaculture pond[D]. Tianjin: Tianjin University of Science & Technology, 2022.
17	刘世晶, 涂雪滢, 钱程, 等. 基于帧间光流特征和改进RNN的草鱼摄食状态分类[J]. 水生生物学报, 2022, 46(6): 914-921.
	LIU S J, TU X Y, QIAN C, et al. Feeding state classification of grass carp based on optical flow and improved RNN[J]. Acta hydrobiologica sinica, 2022, 46(6): 914-921.
18	汪峰. 基于计算机视觉的虾苗活力分析研究[D]. 上海: 上海海洋大学, 2022.
	WANG F. Research on the vitality analysis of shrimp fry based on computer vision[D]. Shanghai: Shanghai Ocean University, 2022.
19	ZHAO J, GU Z B, SHI M M, et al. Spatial behavioral characteristics and statistics-based kinetic energy modeling in special behaviors detection of a shoal of fish in a recirculating aquaculture system[J]. Computers and electronics in agriculture, 2016, 127: 271-280.
20	HAN F F, ZHU J C, LIU B, et al. Fish shoals behavior detection based on convolutional neural network and spatiotemporal information[J]. IEEE access, 2020, 8: 126907-126926.
21	LI X H, LIU M Q, ZHANG S L, et al. Fish trajectory extraction based on object detection[C]// 2020 39th Chinese Control Conference (CCC). Piscataway, New Jersey, USA: IEEE, 2020: 6584-6588.
22	LI W R, LI F, LI Z B. CMFTNet: Multiple fish tracking based on counterpoised JointNet[J]. Computers and electronics in agriculture, 2022, 198: ID 107018.
23	XU W K, LIU C, WANG G X, et al. Behavioral response of fish under ammonia nitrogen stress based on machine vision[J]. Engineering applications of artificial intelligence, 2024, 128: ID 107442.
24	PÉREZ-ESCUDERO A, VICENTE-PAGE J, HINZ R C, et al. idTracker: Tracking individuals in a group by automatic identification of unmarked animals[J]. Nature methods, 2014, 11(7): 743-748.
25	FARNEBÄCK G. Two-frame motion estimation based on polynomial expansion[M]// BIGUN J, GUSTAVSSON T. Eds. Image analysis. Berlin, Heidelberg: Springer Berlin Heidelberg, 2003: 363-370.
26	CHEN J R, KAO S H, HE H, et al. Run, don't walk: Chasing higher FLOPS for faster neural networks[C]// 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway, New Jersey, USA: IEEE, 2023: 12021-12031.
27	OUYANG D L, HE S, ZHANG G Z, et al. Efficient multi-scale attention module with cross-spatial learning[C]// ICASSP 2023—2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Piscataway, New Jersey, USA: IEEE, 2023: 1-5.
28	陈凯琳. 南美白对虾栖息、生长及游泳特性试验研究[D]. 舟山: 浙江海洋大学, 2020.
	CHEN K L. Experimental study on Penaeus vannamei's habitat, growth and swimming characteristics[D]. Zhoushan: Zhejiang Ocean University, 2020.
29	LIAO X X, CAI Z M, CHEN J, et al. Physics-based optical flow estimation under varying illumination conditions[J]. Signal processing: Image communication, 2023, 117: ID 117007.
30	CHEN H, WANG Y, GUO J, et al. VanillaNet: The power of minimalism in deep learning[J]. arXiv: 2305.12972v2, 2023.
31	CHENG G, YUAN X, YAO X W, et al. Towards large-scale small object detection: Survey and benchmarks[C]// IEEE Transactions on Pattern Analysis and Machine Intelligence. Piscataway, New Jersey, USA: IEEE, 2023: 13467-13488.

日期/MM–DD	对虾情况	V _avg
07–20	正常	7.97
07–28	发病	8.12
08–20	正常	7.43
08–20	正常	7.17
09–26	发病	6.9
09–30	发病	7.52

消融试验	FasterNet	EMA	mAP0.5/%	mAP0.95/%	Parameters/M	Flops/G
方法1	不添加	不添加	94.9	67.3	11.1	28.6
方法2	添加	不添加	93.6	63.1	6.1	16.2
方法3	不添加	添加	93.3	67.7	10.7	28.6
方法4	添加	添加	95.1	65.2	6.1	16.5

网络模型	mAP0.5/%	mAP0.95/%	Parameters/M	Flops/G
YOLOv5	93	64.4	7.0	16.0
YOLOv7	95.5	65.8	37.2	105.2
YOLOv7-Tiny	72.7	29.0	6.1	13.2
YOLOv8-VanillaNet^［30］	92.8	60.5	3.8	10.4
改进YOLOv8	95.1	65.2	6.1	16.5

组别	是否发病	总数	检测数	准确率/%
1	正常	50	44	88
1	发病	50	36	72
2	正常	50	49	98
2	发病	50	35	70
3	正常	50	47	94
3	发病	50	38	76
	合计	300	249	83

[1]	ZUO Haoxuan, HUANG Qicheng, YANG Jiahao, MENG Fanjia, LI Sien, LI Li. In Situ Identification Method of Maize Stalk Width Based on Binocular Vision and Improved YOLOv8 [J]. Smart Agriculture, 2023, 5(3): 86-95.
[2]	KANG Xi, LIU Gang, CHU Mengyuan, LI Qian, WANG Yanchao. Advances and Challenges in Physiological Parameters Monitoring and Diseases Diagnosing of Dairy Cows Based on Computer Vision [J]. Smart Agriculture, 2022, 4(2): 1-18.
[3]	ZHAO Jinling, DU Shizhou, HUANG Linsheng. Monitoring Wheat Powdery Mildew (Blumeria graminis f. sp. tritici) Using Multisource and Multitemporal Satellite Images and Support Vector Machine Classifier [J]. Smart Agriculture, 2022, 4(1): 17-28.
[4]	XIAO Shijie, WANG Qiaohua, LI Chunfang, ZHAO Limei, LIU Xinya, LU Shiyu, ZHANG Shujun. Construction of Milk Purchase Classification Model Based on Shuffled Frog Leaping Algorithm and Support Vector Machine [J]. Smart Agriculture, 2021, 3(4): 77-85.
[5]	LI Dongbo, HUANG Lyuwen, ZHAO Xubo. Detection Method of Apple Mould Core Based on Dielectric Characteristics [J]. Smart Agriculture, 2021, 3(4): 66-76.
[6]	FLORES Paulo, ZHANG Zhao, MATHEW Jithin, JAHAN Nusrat, STENGER John. Distinguishing Volunteer Corn from Soybean at Seedling Stage Using Images and Machine Learning [J]. Smart Agriculture, 2020, 2(3): 61-74.
[7]	LIU Tianlin, ZHU Xicun, BAI Xueyuan, PENG Yufeng, LI Meixuan, TIAN Zhongyu, JIANG Yuanmao, YANG Guijun. Hyperspectral Estimation Model Construction and Accuracy Comparison of Soil Organic Matter Content [J]. Smart Agriculture, 2020, 2(3): 129-138.
[8]	Zhou Chengquan, Ye Hongbao, Yu Guohong, Hu Jun, Xu Zhifu. A fast extraction method of broccoli phenotype based on machine vision and deep learning [J]. Smart Agriculture, 2020, 2(1): 121-132.
[9]	Zhou Chao, Xu Daming, Lin Kai, Chen Lan, Zhang Song, Sun Chuanheng, Yang Xinting. Evaluation of fish feeding intensity in aquaculture based on near-infrared machine vision [J]. Smart Agriculture, 2019, 1(1): 76-84.