基于机器视觉和YOLOv5的油菜角果表型参数分析

doi:10.3969/j.issn.1004-1524.20230786

浙江农业学报 ›› 2024, Vol. 36 ›› Issue (6): 1379-1388.DOI: 10.3969/j.issn.1004-1524.20230786

基于机器视觉和YOLOv5的油菜角果表型参数分析

陈威¹(), 朱怡航², 顾清², 林宝刚³, 张小斌²^,^*()

1.浙江农林大学数学与计算机科学学院,浙江杭州 311300
2.浙江省农业科学院数字农业研究所,浙江杭州 310021
3.浙江省农业科学院作物与核技术利用研究所,浙江杭州 310021

收稿日期:2023-06-21 出版日期:2024-06-25 发布日期:2024-07-02
作者简介:陈威(1999—),男,湖北蕲春人,硕士,主要从事农业信息技术研究。E-mail:jiushishu@stu.zafu.edu.cn
通讯作者: *张小斌,E-mail:riceipm1@zju.edu.cn
基金资助:
浙江省农业科学院生物育种融通计划“作物表型高通量获取与智能应用平台建设”(2022SWYZ);浙江省农业旱粮新品种选育重大科技专项 “油菜新品种选育”(2021C02064-2-3)

Analysis of phenotypic parameters of rapeseed silique based on machine vision and YOLOv5

CHEN Wei¹(), ZHU Yihang², GU Qing², LIN Baogang³, ZHANG Xiaobin²^,^*()

1. College of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, China
2. Institute of Digital Agriculture, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China
3. Institute of Crop and Nuclear Technology Utilization, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China

Received:2023-06-21 Online:2024-06-25 Published:2024-07-02

摘要/Abstract

摘要：

为了更加高效和准确地获取油菜角果表型参数,在图像处理技术和深度学习算法的基础上,以迎春一号油菜角果为实验材料,综合考虑油菜育种对角果外观表型参数的需求,提出了一种基于机器视觉的油菜角果表型分析方法:利用图像处理技术实现了油菜角果的柄喙长度、果身长度、果身宽度、弦长、弧长、面积等外观表型性状的提取,使用YOLOv5对单角果籽粒进行无损计数。对角果实物及标定物进行测量验证,结果表明,图像分析出的角果表型指标与人工实际测量值无显著性差异(P>0.05),决定系数(R²)均大于0.96,均方根误差(root mean square error, RMSE)均小于3 mm,平均绝对值误差(mean absolute error, MAE)均小于2.80 mm,平均绝对百分比误差(mean absolute percentage error, MAPE)均不超过4%。标定物直径最大RMSE为0.3 mm, MAE均小于0.28 mm, MAPE均小于2.00%,面积指标最大RMSE为12.09 mm², MAE均小于11.56 mm²,MAPE均小于5%。YOLOv5识别出的籽粒数与实际值无显著性差异(P>0.05), R²为0.987,RMSE为0.68粒,MAE为0.27粒,MAPE为1%。该研究的油菜角果表型分析方法操作简单、成本较低,能有效地减少人工测量的误差,提高获取表型信息的可靠性和油菜育种工作的效率,为油菜表型信息的定量化分析提供了一定的参考。

关键词: 油菜角果, 表型, 图像处理, 籽粒计数

Abstract:

To obtain the phenotypes of rapeseed siliques more efficiently and accurately, a machine vision-based method for analyzing the phenotypes of rapeseed siliques was proposed using image processing technology and deep learning algorithms. Siliques of rapeseed cultivar Yingchun No. 1 were used as materials. The need to acquire morphological phenotypes of siliques is considered for rapeseed breeding. Image processing technology was used to extract the morphological phenotypes of the rapeseed siliques, including pedicel length, silique length, silique width, chord length, arc length, and area. YOLOv5 was used to perform nondestructive silique grain counting. Measurements and verifications were made on the siliques and calibration objects. There was no significant difference (P>0.05) between the phenotypic indexes of siliques assessed by image analysis and the actual measured values. The R²was more than 0.96, the root mean square error (RMSE) was less than 3 mm, the mean absolute error (MAE) was less than 2.80 mm, and the mean absolute percentage error (MAPE) was less than 4%. The calibration object diameter had a maximum RMSE of 0.3 mm, the MAE was less than 0.28 mm, and the MAPE was less than 2.00%. The area index had a maximum RMSE of 12.09 mm², the MAE was less than 11.56 mm², and the MAPE was less than 5%. There was no significant difference between the number of grains identified by YOLOv5 and the actual value (P>0.05), R²was 0.99, RMSE was 0.68, MAE was 0.27, and MAPE was 1%. The method for analyzing the phenotypes of rapeseed siliques proposed in this study is easy to operate and labor-saving. It can effectively reduce the manual measurement error, improve the reliability of obtaining phenotypic information, and increase the efficiency of rapeseed breeding work. It also provides particular guidance for the quantitative analysis of rapeseed phenotypic information.

Key words: rapeseed silique, phenotype, image processing, grain counting

中图分类号:

TP391.4

陈威, 朱怡航, 顾清, 林宝刚, 张小斌. 基于机器视觉和YOLOv5的油菜角果表型参数分析[J]. 浙江农业学报, 2024, 36(6): 1379-1388.

CHEN Wei, ZHU Yihang, GU Qing, LIN Baogang, ZHANG Xiaobin. Analysis of phenotypic parameters of rapeseed silique based on machine vision and YOLOv5[J]. Acta Agriculturae Zhejiangensis, 2024, 36(6): 1379-1388.

图/表 14

图1 外观表型图像获取装置(A)及参照物(B)三维模型示意图

Fig.1 Diagram of the 3D model for the appearance phenotype image acquisition device (A) and the reference (B)

图2 亚克力板(A)及籽粒数图像获取装置(B)三维模型示意图

Fig.2 Diagram of the 3D model for the acrylic plate (A) and rapeseed image acquisition device (B)

图3 外观表型原图(A)及预处理效果图(B)

Fig.3 Original diagram of the appearance phenotype (A) and diagram of the pretreatment effect (B)

图4 干角果(A)及青角果(B)籽粒数图像处理效果

Fig.4 Effect of image processing on dry silique (A) and green silique (B) rapeseed counts

图5 角果果身长处理流程图

Fig.5 Flow chart for processing the length of silique

图6 角果柄长和喙长处理流程图

Fig.6 Flow chart for processing the shank length and beak length of silique

图7 角果表型图像分析算法示意图

Fig.7 Diagram of the Silique phenotypic image analysis algorithm

图8 角果弦长处理流程图

Fig.8 Flow chart for processing the chord length of silique

图9 角果弧长处理流程图

Fig.9 Flow chart for processing the arc length of silique

表1 油菜角果图像测量与实际结果比较

Table 1 Comparison of rapeseed siliques image measured results and actual results

参数 Item	图像测量值 Image measured value/mm	实测值 Actual value/mm	R²	RMSE/mm	MAE/mm	MAPE/%
柄长Shank length	22.01±5.24 d	22.07±5.25 d	0.97	0.94	0.82	4.00
喙长Beak length	15.82±3.69 e	15.54±3.65 e	0.98	0.64	0.52	4.00
果长Silique length	63.16±16.04 c	63.58±15.44 c	0.99	1.88	1.59	3.00
弦长Chord length	93.92±17.19 b	96.27±17.26 b	0.99	2.49	2.35	2.00
凸边弧长Convex arc length	100.46±18.59 a	103.26±18.73 a	0.99	2.99	2.80	3.00

表2 标定物图像测量直径结果与实际结果比较

Table 2 Comparison of the diameter measured results and actual results

硬币 Coins	图像测量值 Image measured value/mm	实测值 Actual value/mm	RMSE/mm	MAE/mm	MAPE/%
欧元1 cent Euro 1 cent	16.28±0.10 e	16.25±0.08 e	0.10	0.08	1.00
欧元10 cent Euro 10 cent	19.68±0.07 d	19.75±0.09 d	0.09	0.09	0.00
欧元50 cent Euro 50 cent	24.23±0.07 b	24.25±0.11 b	0.07	0.04	0.00
人民币1元 RMB 1 yuan	25.02±0.10 a	25.00±0.11 a	0.09	0.07	0.20
澳元2 dollars Australian 2 dollars	20.34±0.10 c	20.62±0.09 c	0.30	0.28	2.00

表3 标定物图像测量面积结果与实际结果比较

Table 3 Comparison of the area measured result of the calibration object image with the actual results

硬币 Coins	图像测量值 Image measured value/mm²	实测值 Actual value/mm²	RMSE/mm²	MAE/mm²	MAPE/%
欧元1 cent Euro 1 cent	216.44±1.13 e	207.39±2.86 e	9.44	9.44	5.00
欧元10 cent Euro 10 cent	313.56±1.01 d	306.35±1.94 d	7.62	7.56	2.00
欧元50 cent Euro 50 cent	471.78±2.11 b	461.86±1.55 b	9.98	9.78	2.00
人民币1元 RMB 1 yuan	502.56±3.78 a	490.87±1.56 a	12.09	11.56	2.00
澳元2 dollars Australian 2 dollars	336.00±1.80 c	333.94±1.96 c	2.62	2.44	1.00

图10 干角果原图(A)及YOLOv5检测结果(B)

Fig.10 Original diagram of the dry silique (A) and detection results of YOLOv5 (B)

图11 青角果原图(A)及YOLOv5检测结果(B)

Fig.11 Original diagram of the green silique(A) and detection results of YOLOv5 (B)

参考文献 32

[1]	LI H T, FENG H, GUO C C, et al. High-throughput phenotyping accelerates the dissection of the dynamic genetic architecture of plant growth and yield improvement in rapeseed[J]. Plant Biotechnology Journal, 2020, 18(11): 2345-2353.
[2]	刘倩, 王鑫, 赵云. 甘蓝型油菜角果和种子发育相关的Bna-novel-miR432功能分析[J]. 四川大学学报(自然科学版), 2022, 59(5): 169-178.
	LIU Q, WANG X, ZHAO Y. Functional analysis of Bna-novel-miR432 related to silique and seed development in Brassica napus[J]. Journal of Sichuan University(Natural Science Edition), 2022, 59(5): 169-178. (in Chinese with English abstract)
[3]	郭燕枝, 杨雅伦, 孙君茂. 我国油菜产业发展的现状及对策[J]. 农业经济, 2016(7): 44-46.
	GUO Y Z, YANG Y L, SUN J M. Present situation and countermeasures of rapeseed industry development in China[J]. Agricultural Economy, 2016(7): 44-46. (in Chinese)
[4]	孙程明, 陈松, 彭琦, 等. 甘蓝型油菜角果长度性状的全基因组关联分析[J]. 作物学报, 2019, 45(9): 1303-1310.
	SUN C M, CHEN S, PENG Q, et al. Genome-wide association study of silique length in rapeseed(Brassica napus L.)[J]. Acta Agronomica Sinica, 2019, 45(9): 1303-1310. (in Chinese with English abstract)
[5]	BENNETT E J, ROBERTS J A, WAGSTAFF C. The role of the pod in seed development: strategies for manipulating yield[J]. The New Phytologist, 2011, 190(4): 838-853.
[6]	FIORANI F, SCHURR U. Future scenarios for plant phenotyping[J]. Annual Review of Plant Biology, 2013, 64: 267-291.
[7]	ZHANG Y, ZHANG N Q. Imaging technologies for plant high-throughput phenotyping: a review[J]. Frontiers of Agricultural Science and Engineering, 2018, 5(4): 406-419.
[8]	李岚涛, 任涛, 汪善勤, 等. 基于角果期高光谱的冬油菜产量预测模型研究[J]. 农业机械学报, 2017, 48(3): 221-229.
	LI L T, REN T, WANG S Q, et al. Prediction models of winter oilseed rape yield based on hyperspectral data at pod-filling stage[J]. Transactions of the Chinese Society for Agricultural Machinery, 2017, 48(3): 221-229. (in Chinese with English abstract)
[9]	RAJKOVIĆ D, MARJANOVIĆ JEROMELA A, PEZO L, et al. Yield and quality prediction of winter rapeseed-artificial neural network and random forest models[J]. Agronomy, 2021, 12(1): 58.
[10]	ZHAO H C, AN F Y, DU D Z. New idioplasmic resource B. napus L. with multi-loculus founded by interspecific hybridization[C]// Proceedings of the 12th International Rapeseed Congress, Wuhan. 2007, 3: 294-295.
[11]	HAUSMANN J. Challenges for integrated pest management of Dasineura brassicae in oilseed rape[J]. Arthropod-Plant Interactions, 2021, 15(5): 645-656.
[12]	ZHENG X R, KOOPMANN B, ULBER B, et al. A global survey on diseases and pests in oilseed rape-current challenges and innovative strategies of control[J]. Frontiers in Agronomy, 2020, 2: 590908.
[13]	张小斌, 谢宝良, 朱怡航, 等. 基于图像处理技术的菜用大豆豆荚高通量表型采集与分析[J]. 核农学报, 2022, 36(3): 602-612.
	ZHANG X B, XIE B L, ZHU Y H, et al. High-throughput phenotype collection and analysis of vegetable soybean pod based on image processing technology[J]. Journal of Nuclear Agricultural Sciences, 2022, 36(3): 602-612. (in Chinese with English abstract)
[14]	LI Z B, GUO R H, LI M, et al. A review of computer vision technologies for plant phenotyping[J]. Computers and Electronics in Agriculture, 2020, 176: 105672.
[15]	TENG X W, ZHOU G S, WU Y X, et al. Three-dimensional reconstruction method of rapeseed plants in the whole growth period using RGB-D camera[J]. Sensors, 2021, 21(14): 4628.
[16]	IRAJI M S. Comparison between soft computing methods for tomato quality grading using machine vision[J]. Journal of Food Measurement and Characterization, 2019, 13(1): 1-15.
[17]	SU Q H, KONDO N, LI M Z, et al. Potato quality grading based on machine vision and 3D shape analysis[J]. Computers and Electronics in Agriculture, 2018, 152: 261-268.
[18]	WAN P, TOUDESHKI A, TAN H Q, et al. A methodology for fresh tomato maturity detection using computer vision[J]. Computers and Electronics in Agriculture, 2018, 146: 43-50.
[19]	李就好, 林乐坚, 田凯, 等. 改进Faster R-CNN的田间苦瓜叶部病害检测[J]. 农业工程学报, 2020, 36(12): 179-185.
	LI J H, LIN L J, TIAN K, et al. Detection of leaf diseases of balsam pear in the field based on improved Faster R-CNN[J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(12): 179-185. (in Chinese with English abstract)
[20]	GOBALAKRISHNAN N, PRADEEP K, RAMAN C J, et al. A systematic review on image processing and machine learning techniques for detecting plant diseases[C]// 2020 International Conference on Communication and Signal Processing (ICCSP). July 28-30, 2020, Chennai, India. IEEE, 2020: 465-468.
[21]	SHARIF M, KHAN M A, IQBAL Z, et al. Detection and classification of citrus diseases in agriculture based on optimized weighted segmentation and feature selection[J]. Computers and Electronics in Agriculture, 2018, 150: 220-234.
[22]	宁远霖, 杨颖, 李振波, 等. 基于改进YOLOv5的复杂跨域场景下的猪个体识别与计数[J]. 农业工程学报, 2022, 38(17): 168-175.
	NING Y L, YANG Y, LI Z B, et al. Detecting and counting pig number using improved YOLOv5 in complex scenes[J]. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38(17): 168-175. (in Chinese with English abstract)
[23]	SONG H S, LIANG H X, LI H Y, et al. Vision-based vehicle detection and counting system using deep learning in highway scenes[J]. European Transport Research Review, 2019, 11(1): 1-16.
[24]	姚业浩, 李毅念, 陈玉仑, 等. 基于油菜角果长度图像识别的每角粒数测试方法[J]. 农业工程学报, 2021, 37(23): 153-160.
	YAO Y H, LI Y N, CHEN Y L, et al. Testing method for the seed number per silique of oilrape based on recognizing the silique length images[J]. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(23): 153-160. (in Chinese with English abstract)
[25]	姚业浩, 李毅念, 邹玮, 等. 油菜籽粒千粒重图像测定方法[J]. 中国油料作物学报, 2022, 44(1): 201-210.
	YAO Y H, LI Y N, ZOU W, et al. Determination method on thousand-seed weight of rapeseed based on image processing[J]. Chinese Journal of Oil Crop Sciences, 2022, 44(1): 201-210. (in Chinese with English abstract)
[26]	刘仁峰, 黄诗瑶, 聂勇鹏, 等. 油菜角果数量及关键表型参数的自动化检测方法研究[J]. 中国油料作物学报, 2020, 42(1): 71-77.
	LIU R F, HUANG S Y, NIE Y P, et al. Automated detection research for number and key phenotypic parameters of rapeseed silique[J]. Chinese Journal of Oil Crop Sciences, 2020, 42(1): 71-77. (in Chinese with English abstract)
[27]	FETTER K C, EBERHARDT S, BARCLAY R S, et al. StomataCounter: a neural network for automatic stomata identification and counting[J]. The New Phytologist, 2019, 223(3): 1671-1681.
[28]	周成全, 叶宏宝, 俞国红, 等. 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.
	ZHOU C Q, YE H B, YU G H, et al. A fast extraction method of broccoli phenotype based on machine vision and deep learning[J]. Smart Agriculture, 2020, 2(1): 121-132. (in Chinese with English abstract)
[29]	王婉心, 贾立锋. 骨架提取中的毛刺去除方法[J]. 广东工业大学学报, 2014, 31(4): 90-94.
	WANG W X, JIA L F. The method of removing burrs in skeleton extraction[J]. Journal of Guangdong University of Technology, 2014, 31(4): 90-94. (in Chinese with English abstract)
[30]	翁杨, 曾睿, 吴陈铭, 等. 基于深度学习的农业植物表型研究综述[J]. 中国科学: 生命科学, 2019, 49(6): 698-716.
	WENG Y, ZENG R, WU C M, et al. A survey on deep-learning-based plant phenotype research in agriculture[J]. Scientia Sinica(Vitae), 2019, 49(6): 698-716. (in Chinese with English abstract)
[31]	LI L, ZHANG Q, HUANG D F. A review of imaging techniques for plant phenotyping[J]. Sensors, 2014, 14(11): 20078-20111.
[32]	朱怡航, 张小斌, 沈颖越, 等. 基于图像识别技术的金针菇表型高通量采集与分析[J]. 菌物学报, 2021, 40(3): 626-640.
	ZHU Y H, ZHANG X B, SHEN Y Y, et al. High-throughput phenotyping collection and analysis of Flammulina filiformis based on image recognition technology[J]. Mycosystema, 2021, 40(3): 626-640. (in Chinese with English abstract)

编辑推荐

Metrics

阅读次数

全文

321

HTML			PDF

最新录用	在线预览	正式出版	最新录用	在线预览	正式出版
0	0	7	0	0	314

来源	本网站	其他网站

次数	220	101
比例	69%	31%

摘要

161

最新录用	在线预览	正式出版

0	0	161

	来源	本网站

	次数	161
	比例	100%

基于机器视觉和YOLOv5的油菜角果表型参数分析

Analysis of phenotypic parameters of rapeseed silique based on machine vision and YOLOv5

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 32

相关文章 15

编辑推荐

Metrics

本文评价

[1]	杨新宇, 冯全, 张建华, 杨森. 基于对比学习的植物叶片病害识别[J]. 浙江农业学报, 2024, 36(1): 215-224.
[2]	戚建莉, 张荣, 吴文俊, 姜成英, 赵梦炯, 金高明, 陈炜青. 油橄榄种质资源表型性状多样性分析与评价[J]. 浙江农业学报, 2023, 35(5): 1001-1015.
[3]	陈浩, 张跃伟, 王宁, 王梓, 常青山. 毛白杨凋落叶对节节麦种子萌发和表型可塑性的影响[J]. 浙江农业学报, 2023, 35(3): 615-623.
[4]	郑冉, 吕丹, 武清贵, 邸晓红, 朱通通, 邱冠杰, 罗红兵. 玉米C型胞质不育系S37-2败育的生物学与生理生化机制分析[J]. 浙江农业学报, 2023, 35(2): 259-265.
[5]	丁一, 郑旭霞, 黄海涛, 毛宇骁, 赵芸. 浙江4个主要茶树群体种资源表型性状及遗传多样性分析[J]. 浙江农业学报, 2023, 35(2): 364-372.
[6]	翟艺兰, 张楚磊, 楚爱香, 高俊鸽, 夏晴情, 卢志昌. 二十七种槭属植物表型多样性分析[J]. 浙江农业学报, 2023, 35(11): 2621-2635.
[7]	杨迪, 张乃群, 王雪勇, 张军, 王新军. 基于数量性状的伏牛山野生中华猕猴桃资源综合评价[J]. 浙江农业学报, 2023, 35(10): 2354-2363.
[8]	杨海龙, 王晖, 雷锦超, 蔡金洋. 浙江省早籼稻种质资源的表型多样性分析与评价[J]. 浙江农业学报, 2022, 34(8): 1571-1581.
[9]	沈升法, 项超, 李兵, 罗志高, 吴列洪. 浙江省马铃薯种质资源的表型鉴定与多样性分析[J]. 浙江农业学报, 2022, 34(11): 2319-2328.
[10]	何庆海, 刘本同, 周政德, 方茹, 杨少宗. 基于枝条和叶片表型性状的掌叶覆盆子种质资源多样性研究[J]. 浙江农业学报, 2021, 33(9): 1660-1667.
[11]	王治会, 彭华, 杨普香, 江新凤, 李文金, 岳翠男, 李琛, 李延升. 17份黄金菊茶树自然杂交单株的表型变异与资源价值评价[J]. 浙江农业学报, 2021, 33(2): 298-307.
[12]	洪霞, 赵永彬, 屈为栋, 陈银龙, 邱莉萍, 王娇阳. 基于表型性状与简单重复序列标记的浙江省芋种质资源遗传多样性比较[J]. 浙江农业学报, 2020, 32(9): 1544-1554.
[13]	陈红林, 秦高婵, 楼宝, 钱豪杰, 姚振海. 不同生长时期红螯螯虾表型性状差异分析[J]. 浙江农业学报, 2020, 32(12): 2154-2161.
[14]	鲍烈, 王曼韬, 刘江川, 文波, 明月. 基于卷积神经网络的小麦产量预估方法[J]. 浙江农业学报, 2020, 32(12): 2244-2252.
[15]	谢文钢, 李晓松, 李伟, 唐茜. 四川地方中小叶茶树资源的表型遗传多样性[J]. 浙江农业学报, 2019, 31(9): 1405-1415.