[1]朱荣胜,闫学慧,陈庆山.基于图像识别和卷积神经网络的大豆优良籽粒筛选研究[J].大豆科学,2020,39(02):189-197.[doi:10.11861/j.issn.1000-9841.2020.02.0189]
　ZHU Rong-sheng,YAN Xue-hui,CHEN Qing-shan.Study on the Optimization of Soybean Seed Selection based on Image Recognition and Convolution Neural Network[J].Soybean Science,2020,39(02):189-197.[doi:10.11861/j.issn.1000-9841.2020.02.0189]
点击复制

基于图像识别和卷积神经网络的大豆优良籽粒筛选研究

《大豆科学》[ISSN:1000-9841/CN:23-1227/S] 卷: 第39卷 期数: 2020年02期 页码: 189-197 栏目: 出版日期: 2020-03-20
Title:
Study on the Optimization of Soybean Seed Selection based on Image Recognition and Convolution Neural Network
作者:
朱荣胜1; 闫学慧2; 陈庆山3
（1.东北农业大学 理学院，黑龙江 哈尔滨 150030； 2.东北农业大学 工程学院，黑龙江 哈尔滨 150030； 3.东北农业大学 农学院，黑龙江 哈尔滨 150030）
Author(s):
ZHU Rong-sheng1;  YAN Xue-hui2;  CHEN Qing-shan3
(1.College of Science,Northeast Agricultural University,Harbin 150030,China; 2.College of Engineering,Northeast Agricultural University,Harbin 150030,China; 3.College of Agricultural,Northeast Agricultural University,Harbin 150030,China)
关键词:
大豆籽粒; 品质; 图像处理; 分类识别; 卷积神经网络
Keywords:
Soybean seed;  Quality;  Image processing;  Classification;  Convolution neural network
DOI:
10.11861/j.issn.1000-9841.2020.02.0189
文献标志码:
A
摘要:
为实现通过籽粒图像识别方法对大豆籽粒的品质进行快速、准确检测，以大豆正常品质籽粒及非正常品质籽粒的分类为例，提出一种基于卷积神经网络的大豆优良籽粒图像筛选分类识别方法。建立大豆籽粒品质数据集，设计卷积神经网络，提取大豆籽粒图像特征。为提高分类准确率和实时性，从设计选择卷积神经网络结构、减小过拟合、加快训练收敛速度、增强网络的鲁棒性等方面对卷积神经网络进行优化，最终选择含有4个卷积层、4个池化层、2个全连接层的6层卷积神经网络，采用L2正则化和小批量训练学习方法对网络进行优化训练测试。将结果与传统机器学习分类方法进行比较，试验结果表明：优化的卷积神经网络对大豆籽粒品质分类的准确率达到98.8%，平均检测一幅大豆单籽粒图像的时间为2.96 ms，可为大豆籽粒品质划分提供重要参考。
Abstract:
In order to quickly and accurately detect the quality of soybean seeds by the method of seed images recognition, a method of image screening and recognition based on convolution neural network is proposed, taking the classification of normal and abnormal quality seeds as an example. The data set of soybean seed quality was established, and convolution neural network was designed to extract the image features of soybean seed. In order to improve the classification accuracy and real-time performance, the convolution neural network was optimized from the aspects of design and selection of convolution neural network structure, reduction of over fitting, acceleration of training convergence speed, and enhancement of network robustness. Finally, the 6-layer convolution neural network with 4 convolution layers, 4 pooling layers and 2 fully connected layers were selected, L2 regularization and mini batch training methods were used for the network′s optimization training and test. Comparing the results with the traditional machine learning classification methods, the experimental results show that the accuracy of the optimized convolution neural network is 98.8%, and the average detection time of a single soybean seed image is 2.96 ms, which can provide an important reference for soybean seeds quality classification.

参考文献/References:

［1］吴冬梅,严菊敏,何会超,等.不同贮藏方式对菜用大豆外观和品质的影响［J］.大豆科学,2012,31(1):155-158. (Wu D M, Yan J M, He H C, et al. Effects of different storage method on appearance and quality of vegetable soybean［J］. Soybean Science, 2012, 31(1): 155-158.［2］徐江, 谭敏, 张春庆. 电晕场与介电分选提高水稻种子活力［J］. 农业工程学报, 2013, 29(23): 233-240. (Xu J, Tan M, Zhang C Q, et al. Improving paddy seed vigor by corona discharge field processing and dielectric separation［J］. Transactions of the Chinese Society of Agricultural Engineering, 2013, 29(23): 233-240.)［3］杨冬风. 基于软X-射线造影和机器智能的玉米种子活力检测方法研究［J］. 作物杂志, 2013(3):136-140. (Yang D F. Research on detection method of maize vigor based on soft X-ray and computer intelligence［J］. Crops, 2013(3):136-140. ［4］展慧, 李小昱, 周竹，等. 基于近红外光谱和机器视觉融合技术的板栗缺陷检测［J］. 农业工程学报, 2011, 27(2):345-349. (Zhan H, Li X Y, Zhou Z, et al. Detection of chestnut defect based on data fusion of near-infrared spectroscopy and machine vision［J］. Transactions of the CSAE, 2011, 27(2): 345-349. ) ［5］柴玉华, 毕文佳, 谭克竹, 等. 基于高光谱图像技术的大豆品种无损鉴别［J］. 东北农业大学学报, 2016, 47(3): 86-93. (Chai Y H, Bi W J, Tan K Z, et al. Nondestructive identification of soybean seed varieties based on hyperspectral image technology［J］. Journal of Northeast Agricultural University, 2016, 47(3): 86-93.) ［6］Kotabagi S, Desai B L, Linganagouda K, et al. Classification of soybean seed using color image analysis［C］//Afita: The Fifth International Conference of the Asian Federation for Information Technology in Agriculture, 2006.［7］王润涛, 张长利, 房俊龙, 等. 基于机器视觉的大豆籽粒精选技术［J］. 农业工程学报, 2011, 27(8): 355-359. (Wang R T, Zhang C L, Fang J L, et al. Soybean seeds selection based on computer vision［J］. Transactions of the CSAE, 2011, 27(8): 355-359.) ［8］Liu D,Ning X, Li Z, et al. Discriminating and elimination of damaged soybean seeds based on image characteristics［J］. Journal of Stored Products Research, 2015, 60: 67-74.［9］郝建平, 杨锦忠, 杜天庆, 等. 基于图像处理的玉米品种的种子形态分析及其分类研究［J］. 中国农业科学, 2008, 41(4): 994-102. (Hao J P, Yang J Z, Du T Q, et al. A Study on basic morphologic information and classification of maize cultivars based on seed image process［J］. Scientia Agricultura Sinica, 2008,41(4):994-100.)［10］Vilar W T S, Aranha R M, Medeiros E P, et al. Classification of individual castor seed using digital imaging and, multivariate analysis［J］. Journal of the Brazilian Chemical Society, 2014, 26(1):102-109.［11］Kara M,Sayinci B, Elkoca E, et al. Seed size and shape analysis of registered common bean (Phaseolus vulgaris L.) cultivars in Turkey using digital photography［J］.Tarim Bilimleri Dergisi, 2013,19(3):219-234.［12］Wiesnerová D, Ivo W. Computer image analysis of seed shape and seed color for flax cultivar description［J］. Computers and Electronics in Agriculture, 2008, 61(2):126-135.［13］展慧, 李小昱, 王为, 等. 基于机器视觉的板栗分级检测方法［J］. 农业工程学报, 2010, 26(4):327-331. (Zhan H, Li X Y, Wang W, et al. Determination of chestnuts grading based on machine vision［J］. Transactions of the CSAE, 2010, 26(4): 327-331.)［14］Pourreza A, Pourreza H, Abbaspour-Fard M H, et al. Identification of nine Iranian wheat seed varieties by textural analysis with image processing［J］. Computers and Electronics in Agriculture, 2012, 83:102-108.［15］Szegedy C, Liu W, Jia Y, et al. Going deeper with convolutions［J/OL］. Computer Science, 2014:1-12［2020-1-20］. https://arxiv.org/pdf/1409.4842v1.pdf.［16］He K , Zhang X , Ren S , et al. Deep residual learning for image recognition［J］. IEEE Computer Science, 2015:1-47［2020-1-20］.https://www.researchgate.net/publication/307964703_Deep_Residual_Learning.［17］Russakovsky O , Deng J , Su H , et al. ImageNet large scale visual recognition challenge［J］. International Journal of Computer Vision, 2014, 115(3): 211-252.［18］Jeon W S, Rhee S Y, Jeon W S, et al. Plant leaf recognition using a convolution neural network［J］. International Journal of Fuzzy Logic and Intelligent Systems, 2017, 17(1): 26-34.［19］张帅, 淮永建. 基于分层卷积深度学习系统的植物叶片识别研究［J］. 北京林业大学学报, 2016, 38(9):108-115.（Zhang S, Huai Y J. Leaf image recognition based on layered convolutions neural network deep learning［J］.Journal of Beijing Forestry University,2016, 38(9) :108-115.)［20］Mohanty S P, Hughes D P, Salathé M . Using deep learning for image-based plant disease detection［J］. Frontiers in Plant Science,2016,7:1419.［21］孙俊，谭文军，毛罕平，等. 基于改进卷积神经网络的多种植物叶片病害识别［J］. 农业工程学报，2017，33(19):209-215. (Sun J, Tan W J, Mao H P, et al. Recognition of multiple plant leaf diseases based on improved convolutional neural network［J］. Transactions of the Chinese Society of Agricultural Engineering, 2017, 33(19): 209-215.) ［22］Zhu L, Li Z B, Li C, et al.High performance vegetable classification from images based on Alex Net deep learning model［J］. International Journal of Agricultural and Biological Engineering, 2018; 11(4): 217-223.［23］赵志衡，宋欢，朱江波，等. 基于卷积神经网络的花生籽粒完整性识别算法及应用［J］. 农业工程学报，2018，34(21):195-201.（Zhao Z H,Song H,Zhu J B,et al. Identification algorithm and application of peanut kernel integrity based on convolution neural network［J］. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(21):195-201.）

相似文献/References:

[1]杨如萍,张国宏,王立明,等.甘肃省大豆主产区产量性状及品质分析[J].大豆科学,2013,32(01):50.[doi:10.3969/j.issn.1000-9841.2013.01.012]
　YANG Ru-ping,ZHANG Guo-hong,WANG Li-ming,et al.Analysis on Yield and Quality Traits of Soybean in Different Planting Region of Gansu[J].Soybean Science,2013,32(02):50.[doi:10.3969/j.issn.1000-9841.2013.01.012]
[2]雍太文,刘小明,肖秀喜,等.不同种子处理对苗期干旱胁迫条件下大豆农艺性状、产量及品质的影响[J].大豆科学,2013,32(05):620.[doi:10.11861/j.issn.1000-9841.2013.05.0620]
　YONG Tai-wen,LIU Xiao-ming,XIAO Xiu-xi,et al.Effects of Different Seed Treatments on Agronomic Properties,Yield and Quality of Soybean under Drought Stress at Seedling Stage[J].Soybean Science,2013,32(02):620.[doi:10.11861/j.issn.1000-9841.2013.05.0620]
[3]张玉梅,胡润芳,林国强.菜用大豆品质性状研究进展[J].大豆科学,2013,32(05):698.[doi:10.11861/j.issn.1000-9841.2013.05.0698]
　ZHANG Yu-mei,HU Run-fang,LIN Guo-qiang.Research Advance on Quality Traits of Vegetable Soybean[J].Soybean Science,2013,32(02):698.[doi:10.11861/j.issn.1000-9841.2013.05.0698]
[4]季 良,王 仙,彭 琳,等.氮离子束注入和钴60伽玛辐射对大豆产量性状和品质的影响[J].大豆科学,2013,32(02):189.[doi:10.3969/j.issn.1000-9841.2013.02.011]
　JI Liang,WANG Xian,PENG Lin,et al.Effects of Nitrogen Ion Beam Implantation and 60Co-γ Irradiation on Yield Traits and Quality of Soybean[J].Soybean Science,2013,32(02):189.[doi:10.3969/j.issn.1000-9841.2013.02.011]
[5]何龙凉,胡红东,李小琴,等.防城港口岸进境转基因大豆贸易概况及检验检疫分析[J].大豆科学,2013,32(04):539.[doi:10.11861/j.issn.1000-9841.2013.04.0539]
　HE Long-liang,HU Hong-dong,LI Xiao-qin,et al.General Situation of Imported Genetically Modified Soybean in Fangchenggang Port and Its Inspection and Quarantine Analysis[J].Soybean Science,2013,32(02):539.[doi:10.11861/j.issn.1000-9841.2013.04.0539]
[6]成雪峰.黄淮海地区大豆品种主要农艺性状演变分析[J].大豆科学,2011,30(04):585.[doi:10.11861/j.issn.1000-9841.2011.04.0585]
　CHENG Xue-feng.Evolution of Soybean Major Agronomy Characters in Huang-Huai-Hai Region[J].Soybean Science,2011,30(02):585.[doi:10.11861/j.issn.1000-9841.2011.04.0585]
[7]王家军,刘杰,张瑞萍,等.沼渣与化肥配合施用对大豆生长发育、产量和品质的影响[J].大豆科学,2012,31(01):96.[doi:10.3969/j.issn.1000-9841.2012.01.021]
　WANG Jia-jun,LIU Jie,ZHANG Rui-ping,et al.Effect of Matching Application of Biogas Residues and Chemical Fertilizers on Development,Yield and Quality of Soybean[J].Soybean Science,2012,31(02):96.[doi:10.3969/j.issn.1000-9841.2012.01.021]
[8]吴冬梅,严菊敏,何会超,等.不同贮藏方式对菜用大豆外观和品质的影响[J].大豆科学,2012,31(01):155.[doi:10.3969/j.issn.1000-9841.2012.01.036]
　WU Dong-mei,YAN Ju-min,HE Hui-chao,et al.Effects of Different Storage Method on Appearance and Quality of Vegetable Soybean[J].Soybean Science,2012,31(02):155.[doi:10.3969/j.issn.1000-9841.2012.01.036]
[9]张丽华,谭国波,赵洪祥,等.灌溉方式对大豆产量及品质的影响[J].大豆科学,2011,30(05):800.[doi:10.11861/j.issn.1000-9841.2011.05.0800]
　ZHANG Li-hua,TAN Guo-bo,ZHAO Hong-xiang,et al.Effect of Irrigation Method on Yield and Quality of Soybean[J].Soybean Science,2011,30(02):800.[doi:10.11861/j.issn.1000-9841.2011.05.0800]
[10]朱宝国,张春峰,于忠和,等.控释尿素与普通尿素混施对大豆农艺性状及产量和品质的影响[J].大豆科学,2012,31(02):281.[doi:10.3969/j.issn.1000-9841.2012.02.025]
　ZHU Bao-guo,ZHANG Chun-feng,YU Zhong-he,et al.Effect of Controlled Release Urea and Common Urea Blending Application on Agronomic Characters,Yield and Quality of Soybean[J].Soybean Science,2012,31(02):281.[doi:10.3969/j.issn.1000-9841.2012.02.025]

备注/Memo

收稿日期：2019-09-20基金项目：国家重点研发计划子课题(科技部)(2016YED0100201-21-1)。第一作者简介：朱荣胜(1975-)，男，博士，副教授，硕导，主要从事生物信息学研究。E-mail: rshzhu@126.com。 通讯作者：陈庆山(1973-)，男，博士，教授，博导，主要从事大豆遗传育种研究。E-mail: qshchen@126.com。

更新日期/Last Update: 2020-06-10