基于无人机图像以及不同机器学习和深度学习模型的小麦倒伏率检测

doi:10.12133/j.smartag.2021.3.2.202104-SA003

摘要/Abstract

摘要：

小麦在生长过程中发生倒伏会严重影响其产量，因此实时且准确地对小麦倒伏状况监测有很重要的意义。传统的方法采用手工方式生成数据集，不仅效率低、易出错，而且生成的数据集不准确。针对这一问题，本研究提出了一种基于图像处理的自动数据集生成方法。首先利用无人机在15、46和91 m三个高度采集图像数据；采集完数据后，根据无倒伏、倒伏面积<50%和倒伏面积>50%的标准对每一块地的小麦倒伏情况进行人工评估；采用三种机器学习（支持向量机、随机森林、K近邻）和三种深度学习（ResNet101、GoogLeNet、VGG16）算法对小麦倒伏检测情况进行分类。结果显示，ResNet101的分类结果优于随机森林，并且在91 m高度采集的数据分类精度并不低于在15 m高度采集的数据。本研究证明了针对在91 m高度采集的无人机图像，采用ResNet101对小麦倒伏率检测是一种有效的替代人工检测的方法，其检测精度达到了75%。

关键词: 小麦倒伏率, 机器学习, 深度学习, 数据采集高度, 无人机, ResNet101

Abstract:

Wheat lodging is a negative factor affecting yield production. Obtaining timely and accurate wheat lodging information is critical. Using unmanned aerial systems (UASs) images for wheat lodging detection is a relatively new approach, in which researchers usually apply a manual method for dataset generation consisting of plot images. Considering the manual method being inefficient, inaccurate, and subjective, this study developed a new image processing-based approach for automatically generating individual field plot datasets. Images from wheat field trials at three flight heights (15, 46, and 91 m) were collected and analyzed using machine learning (support vector machine, random forest, and K nearest neighbors) and deep learning (ResNet101, GoogLeNet, and VGG16) algorithms to test their performances on detecting levels of wheat lodging percentages: non- (0%), light (<50%), and severe (>50%) lodging. The results indicated that the images collected at 91 m (2.5 cm/pixel) flight height could yield a similar, even slightly higher, detection accuracy over the images collected at 46 m (1.2 cm/pixel) and 15 m (0.4 cm/pixel) UAS mission heights. Comparison of random forest and ResNet101 model results showed that ResNet101 resulted in more satisfactory performance (75% accuracy) with higher accuracy over random forest (71% accuracy). Thus, ResNet101 is a suitable model for wheat lodging ratio detection. This study recommends that UASs images collected at the height of about 91 m (2.5 cm/pixel resolution) coupled with ResNet101 model is a useful and efficient approach for wheat lodging ratio detection.

Key words: wheat lodging ratio, machine learning, deep learning, mission height, UAS, ResNet101

中图分类号:

TP751

. 基于无人机图像以及不同机器学习和深度学习模型的小麦倒伏率检测[J]. 智慧农业(中英文), 2021, 3(2): 23-34.

FLORES Paulo, ZHANG Zhao. Wheat Lodging Ratio Detection Based on UAS Imagery Coupled with Different Machine Learning and Deep Learning Algorithms[J]. Smart Agriculture, 2021, 3(2): 23-34.

图/表 8

参考文献 32

1	BALFOURIER F, BOUCHET S, ROBERT S, et al. Worldwide phylogeography and history of wheat genetic diversity[J]. Science Advances, 2019, 5(5): ID 0536.
2	SHEWRY P R, HEY S J. The contribution of wheat to human diet and health[J]. Food Energy Security, 2015, 4(3): 178-202.
3	STATISTA. Global wheat production from2011/2012 to 2020/2021[EB/OL]. (2021-03-05)[2021-03-30]. .
4	JAHAN N, FLORES P, LIU Z, et al. Detecting and distinguishing wheat diseases using image processing and machine learning algorithms[C]// 2020 ASABE Annual International Virtual Meeting. St. Joseph, MI: ASABE, U.
	S., 2020.
5	MONDAL S, RUTKOSKI J E, VELU G, et al. Harnessing diversity in wheat to enhance grain yield, climate resilience, disease and insect pest resistance and nutrition through conventional and modern breeding approaches[J]. Frontiers in Plant Science, 2016, 7: ID 991.
6	WEBBER H, EWERT F, OLESEN J E, et al. Diverging importance of drought stress for maize and winter wheat in Europe[J]. Nature Communications, 2018, 9(1): 1-10.
7	ZHANG Z, FLORES P, IGATHINATHANE C, et al. Wheat lodging detection from UAS imagery using machine learning algorithms[J]. Remote Sensing, 2020, 12(11): ID 1838.
8	CHAUHAN S, DARVISHZADEH R, BOSCHETTI M, et al. Remote sensing-based crop lodging assessment: Current status and perspectives[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2019,151: 124-140.
9	WU W, MA B L. A new method for assessing plant lodging and the impact of management options on lodging in canola crop production[J]. Scientific Reports, 2016, 6(1): 1-17.
10	SETTER T L, LAURELES E V, MAZAREDO A M. Lodging reduces yield of rice by self-shading and reductions in canopy photosynthesis[J]. Field Crops Research, 1997, 49: 95-106.
11	PINTHUS M J. Lodging in wheat, barley, and oats: the phenomenon, its causes, and preventive measures[J]. Advanced Agronomy, 1974, 25: 209-263.
12	ISLAM M S, PENG S, VISPERAS R M, et al. Lodging-related morphological traits of hybrid rice in a tropical irrigated ecosystem[J]. Field Crops Research, 2007, 101(2): 240-248.
13	ZHANG Z, HEINEMANN P H, LIU J, et al. The development of mechanical apple harvesting technology: A review[J]. Transactions of the ASABE, 2016, 59(5): 1165-1180.
14	ZHANG Z, POTHULA A, LU R. Improvements and evaluation of an infield bin filler for apple bruising and distributions[J]. Transactions of the ASABE, 2019, 62(2): 271-280.
15	ZHANG Z, IGATHINATHANE C, LI J, et al. Technology progress in mechanical harvest of fresh market apples[J]. Computers and Electronics in Agriculture, 2020, 175: ID 105606.
16	LU R, POTHULA A, MIZUSHIMA A, et al. System for sorting fruit: 9919345[P]. 2018-03-20.
17	ZHANG Z, LU Y, LU R. Development and evaluation of an apple infield grading and sorting system[J]. Postharvest Biology and Technology, 2021, 180: ID 111588.
18	YAO L, HU D, ZHAO C, et al. Wireless positioning and path tracking for a mobile platform in greenhouse[J]. International Journal of Agricultural and Biological Engineering, 2021, 14(1): 216-223.
19	FLORES P, ZHANG Z, IGATHINATHANE C. et al. Distinguishing seedling volunteer corn from soybean through greenhouse color, color-infrared, and fused images using machine and deep learning[J]. Industrial Crops and Products, 2021, 161: ID 113223.
20	FISCHER R A, STAPPER M. Lodging effects on high-yielding crops of irrigated semidwarf wheat[J]. Field Crops Research, 1987, 17: 245-258.
21	PIÑERA-CHAVEZ F J, BERRY P M, FOULKES M J, et al. Avoiding lodging in irrigated spring wheat. I. Stem and root structural requirements[J]. Field Crops Research, 2016, 196: 325-336.
22	YANG M D, HUANG K S, KUO Y H, et al. Spatial and spectral hybrid image classification for rice lodging assessment through UAV imagery[J]. Remote Sensing, 2017, 9(6): ID 583.
23	YANG H, CHEN E, LI Z, et al. Wheat lodging monitoring using polarimetric index from RADARSAT-2 data[J]. International Journal of Applied Earth Observation and Geoinformation, 2015, 34: 157-166.
24	ZHAO L, YANG J, LI P, et al. Characterizing lodging damage in wheat and canola using Radarsat-2 polarimetric SAR data[J]. Remote Sensing Letter, 2017, 8(7): 667-675.
25	VARGAS J Q, KHOT L R, PETERS R T, et al. Low orbiting satellite and small UAS-based high-resolution imagery data to quantify crop lodging: A case study in irrigated spearmint[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 17(5): 755-759.
26	CHAUHAN S, DARVISHZADEH R, LU Y, et al. Understanding wheat lodging using multi-temporal Sentinel-1 and Sentinel-2 data[J]. Remote Sensing of Environment, 2020, 243: ID 111804.
27	CHU T, STAREK M J, BREWER M J, et al. Assessing lodging severity over an experimental maize (Zeamays L.) field using UAS images[J]. Remote Sensing, 2017, 9(9): ID 923.
28	RAJAPAKSA S, ERAMIAN M, DUDDU H, et al. Classification of crop lodging with gray level co-occurrence matrix[C]// 2018 IEEE Winter Conference on Applications of Computer Vision. Piscataway, New York, USA: IEEE, 2018: 251-258.
29	LI X, LI X, LIU W, et al. A UAV-based framework for crop lodging assessment[J]. European Journal of Agronomy, 2021, 123: ID 126201.
30	MARDANISAMANI S, MALEKI F, HOSSEINZADEH K, et al. Crop lodging prediction from UAV-acquired images of wheat and canola using a DCNN augmented with handcrafted texture features[C]// 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Piscataway, New York, USA: IEEE, 2019.
31	ABALLA A, CEN H, WAN L, et al. Nutrient status diagnosis of infield oilseed rape via deep learning-enabled dynamic model[J]. IEEE Transactions on Industrial Informatics, 2020, 17(6): 4379-4389.
32	FLORES P, ZHANG Z, JITHIN M, et al. Distinguishing volunteer corn from soybean at seedling stage using images and machine learning[J]. Smart Agriculture, 2020, 2(3): 61-74.

Feature number	Feature	Description
1	R	Red channel
2	G	Green channel
3	B	Blue Channel
4	H	Hue Channel
5	S	Saturation channel
6	I	Intensity channel
7	L	Luminosity channel
8	a	a channel
9	b	b channel
10	NDI	Normalized difference index
11	ExG	Excess green index
12	ExR	Excess red index
13	GLCM	Texture feature

Model	Flight hight
Model	15 m （50 ft.）	46 m （150 ft.）	91 m （300 ft.）
RF	71.5	70.8	71.4
ResNet101	75.2	75.1	75.3

[1]	李志军, 杨圣慧, 史德帅, 刘星星, 郑永军. 基于轻量化改进YOLOv5的苹果树产量测定方法[J]. 智慧农业（中英文）, 2021, 3(2): 100-114.
[2]	邱文杰, 叶进, 胡亮青, 杨娟, 李其利, 莫贱友, 易万茂. 面向植物病害识别的卷积神经网络精简结构Distilled-MobileNet模型[J]. 智慧农业（中英文）, 2021, 3(1): 109-117.
[3]	李亮德, 王秀娟, 康孟珍, 华净, 樊梦涵. 基于语义融合与模型蒸馏的农业实体识别[J]. 智慧农业（中英文）, 2021, 3(1): 118-128.
[4]	黄凯, 舒磊, 李凯亮, 杨星, 朱艳, 汪小旵, 苏勤. 太阳能杀虫灯物联网节点的防盗防破坏设计及展望[J]. 智慧农业（中英文）, 2021, 3(1): 129-143.
[5]	束美艳, 陈向阳, 王喜庆, 马韫韬. 基于高光谱数据的玉米叶面积指数和生物量评估[J]. 智慧农业（中英文）, 2021, 3(1): 29-39.
[6]	张建, 谢田晋, 杨万能, 周广生. 近地遥感技术在大田作物株高测量中的研究现状与展望[J]. 智慧农业（中英文）, 2021, 3(1): 1-15.
[7]	王昌陵, 何雄奎, BONDSJane, 齐鹏, 杨苡, 高万林. 不同飞行参数下八旋翼植保无人机下洗气流场对雾滴沉积分布特性的影响[J]. 智慧农业（中英文）, 2020, 2(4): 124-136.
[8]	邵国敏, 王亚杰, 韩文霆. 基于无人机多光谱遥感的夏玉米叶面积指数估算方法[J]. 智慧农业（中英文）, 2020, 2(3): 118-128.
[9]	郭威, 吴华瑞, 朱华吉. 设施温室影像采集与环境监测机器人系统设计及应用[J]. 智慧农业（中英文）, 2020, 2(3): 48-60.
[10]	李华龙, 李淼, 詹凯, 刘先旺, 杨选将, 胡泽林, 郭盼盼. 蛋鸡设施养殖环境质量评价预测模型构建方法及性能测试[J]. 智慧农业（中英文）, 2020, 2(3): 37-47.
[11]	魏靖, 王玉亭, 袁会珠, 张梦蕾, 王振营. 基于深度学习与特征可视化方法的草地贪夜蛾及其近缘种成虫识别[J]. 智慧农业（中英文）, 2020, 2(3): 75-85.
[12]	FLORESPaulo, MATHEWJithin, JAHANNusrat, STENGERJohn. 利用图像和机器学习检测大豆作物幼苗期玉米杂苗[J]. 智慧农业（中英文）, 2020, 2(3): 61-74.
[13]	何雄奎. 中国精准施药技术和装备研究现状及发展建议[J]. 智慧农业（中英文）, 2020, 2(1): 133-146.
[14]	周成全, 叶宏宝, 俞国红, 胡俊, 徐志福. 基于机器视觉与深度学习的西兰花表型快速提取方法研究[J]. 智慧农业（中英文）, 2020, 2(1): 121-132.
[15]	万亮, 岑海燕, 朱姜蓬, 张佳菲, 杜晓月, 何勇. 基于纹理特征与植被指数融合的水稻含水量无人机遥感监测[J]. 智慧农业（中英文）, 2020, 2(1): 58-67.