[1]黄琼,杨红云,万颖.基于特征数据的水稻种子分类识别方法[J].江苏农业学报,2021,(01):8-15.[doi:doi:10.3969/j.issn.1000-4440.2021.01.002]
　HUANG Qiong,YANG Hong-yun,WAN Ying.Classification and recognition method of rice seeds based on feature data[J].,2021,(01):8-15.[doi:doi:10.3969/j.issn.1000-4440.2021.01.002]
点击复制

基于特征数据的水稻种子分类识别方法()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2021年01期
页码:
8-15
栏目:
遗传育种·生理生化
出版日期:
2021-02-28

文章信息/Info

Title:
Classification and recognition method of rice seeds based on feature data
作者:
黄琼1; 杨红云2; 万颖1
(1.江西农业大学计算机与信息工程学院，江西南昌330045；2.江西农业大学软件学院，江西南昌330045)
Author(s):
HUANG Qiong1; YANG Hong-yun2; WAN Ying1
（1.College of Computer and Information Engineering, Jiangxi Agricultural University, Nanchang 330045, China;2.College of Software, Jiangxi Agricultural University, Nanchang 330045, China)
关键词:
水稻种子; 线性判别; Bayes; 分类识别
Keywords:
rice seeds; linear discrimination; Bayes; classification and recognition
分类号:
S126；S511
DOI:
doi:10.3969/j.issn.1000-4440.2021.01.002
文献标志码:
A
摘要:
针对水稻种子相似度高、识别困难等问题，提出一种线性判别分析（Linear discriminant analysis,LDA）和贝叶斯分类（Bayes）相结合的分类识别方法，以提高水稻种子分类识别速度和识别准确率。通过对4类水稻种子（楚粳7号、马坝油粘、玉杨糯、玉针香）的图像进行裁剪和分割等预处理操作，提取出水稻种子图像的颜色特征、几何特征和纹理特征。利用线性判别分析、主成分分析、因子分析和局部线性嵌入对特征数据进行分析降维，并分别选择Bayes、K-邻近、支持向量机、多层感知机分类器对原始特征数据和降维数据进行分类识别研究。为提高模型泛化能力，通过图像增强技术对稻种原始数据集进行样本扩充，利用图像增强技术模拟多种环境对水稻种子图片数据集进行增强处理，结果显示，基于数据增强后的LDA_Bayes模型运行时间为0.019 s，识别准确率为99.4%。与其他模型比较，该模型具有更强的鲁棒性和适用性，能高效地分类识别不同环境下的水稻种子，可为水稻种子分类识别提供一种新方法。
Abstract:
Aiming at the problems of high similarity and difficult identification of rice seeds, a classification and identification method which combined linear discriminant analysis (LDA) and Bayesian classification (Bayes) was proposed to improve the identification speed and accuracy of rice seed classification. By performing pre-processing operations such as cropping and segmentation on the images of four types of rice seeds (Chujing No.7, Maba Younian, Yuyangnuo and Yuzhenxiang), the color characteristics, geometric and texture features of rice seed images were extracted. Linear discriminant analysis, principal component analysis, factor analysis and locally linear embedding were used to analyze and reduce the dimensionality of the feature data, and Bayes, K-nearest neighbors, support vector machine, multilayer perceptron classifiers were selected respectively to conduct classification and identification research on the original feature data and dimensionality reduced data. To improve the generalization ability of the model, image enhancement technology was used to extend the original data set of rice seed image samples and to simulate multiple environments to enhance the image data set of rice seeds. The results showed that the running time of LDA_Bayes model based on data enhancement was 0.019 s and the recognition accuracy rate was 99.4%. Compared with other models, LDA_Bayes model shows stronger robustness and applicability, it can classify and recognize rice seeds in different environments efficiently and provide a new method for rice seed classification and identification.

参考文献/References:

[1]MEBATSION H K，PALIWAL J，JAYAS D S. Automatic classification of non-touching cereal grains in digital images using limited morphological and color features[J]. Computer&Electronics in Agriculture,2013,90(90):99-105.
[2]CHOUDHARY R，PALIWAL J. Classification of cereal grains using wavelet, morphological, colour, and textural features of non-touching kernel images[J]. Biosystems Engineering, 2008,99(3):330-337.
[3]徐赛,周志艳,罗锡文. 常规稻与杂交稻谷的仿生电子鼻分类识别[J].农业工程学报,2014,30(9):133-139.
[4]李辉,王阳恩,刘庆,等. 分段激光诱导击穿光谱的水稻种子识别[J].激光杂志,2017,38(2):8-12.
[5]HAI V. Spatial and spectral features utilization on a hyperspectral imaging system for rice seed varietal purity inspection[C]//MACIEJ P,AGNIESZKA T.2016 IEEE RIVF International Conference on Computing & Communication Technologies.Hanoi:IEEE,2016(10):169-174.
[6]FABIYI S，HAI V，TACHTATZIS C，et al. Varietal classification of rice seeds using RGB and hyperspectral images[J]. IEEE Access,2020,8:22493-22505.
[7]邓小琴,朱启兵,黄敏. 融合光谱、纹理及形态特征的水稻种子品种高光谱图像单粒鉴别[J].激光与光电子学进展,2015,52(2):128-134.
[8]李 冬. 基于图像的稻花香水稻种子鉴别方法的研究[D].大庆：黑龙江八一农垦大学,2018.
[9]郑宇. 基于机器视觉的稻谷种子特征提取与品种识别方法研究[D].武汉：华中农业大学,2009.
[10]李秀昊,马旭,李泽华,等. 基于特征融合和SVM的稻谷品种识别[J].中国农机化学报,2019,40(7):97-102.
[11]黄星奕,李剑,姜松. 基于计算机视觉的稻谷品种识别技术的研究[J].江苏大学学报(自然科学版),2004,25(2):102-104.
[12]HONG P, HAI T，LAN L. Comparative study on vision based Rice seed varieties identification[C]//NGUYEN L M,NGUYEN T T.2015 Seventh International Conference on Knowledge and Systems Engineering. Ho Chi Minh City:IEEE, 2015,11(7):377-382.
[13]LIU Z Y，CHENG F，YING Y B，et al. Identification of rice seed varieties using neural network[J].Journal of Zhejiang University Science B, 2005,6(11):53-58.
[14]BARALDI A，PARMIGGIANI F. An investigation of the textural characteristics associated with GLCM matrix statistical parameters[J]. IEEE Transactions on Geoscience and Remote Sensing, 1995, 33(2): 293-304.
[15]HARALICK R M, SHANMNGAM K, DINSTEIN I. Texture feature forimage classification[J].IEEE Transactions on Systems, Manand Cybernetics, 1973, 3 (6) :768-780.
[16]章万锋. 基于PCA与LDA的说话人识别研究[D].杭州：浙江大学,2004.
[17]周志华. 机器学习[M]. 北京：清华大学出版社, 2016.
[18]NIKITA E，NIKITAS P. Sex estimation: a comparison of techniques based on binary logistic, probit and cumulative probit regression, linear and quadratic discriminant analysis, neural networks, and nave Bayes classification using ordinal variables[J]. International Journal of Legal Medicine,2020,134(3):1213-1225.
[19]FAYYAZI S，ABBASPOURFARD M H，ROHANI A，et al. Identification and classification of three Iranian rice varieties in mixed bulks using image processing and MLP neural network[J]. International Journal of Food Engineering,2017,13(5).
[20]ZANATY E A. Support vector machines (SVMs) versus Multilayer perception (MLP) in data classification[J]. Egyptian Informatics Journal,2012,13(3):177-183.
[21]彭彦昆,赵芳,李龙,等. 利用近红外光谱与PCA-SVM识别热损伤番茄种子[J].农业工程学报,2018,34(5):159-165.
[22]梁晓莹,田光杰. 基于混合特征LLE融合与SVM的质量异常模式识别[J].组合机床与自动化加工技术,2020(3):55-59，64.
[23]刘开南,冯新扬,邵超. 一种面向图像分类的流形学习降维算法[J].计算机应用与软件,2019,36(8):210-213，229.
[24]高惠璇. 应用多元统计分析[M]. 北京：北京大学出版社, 2005.
[25]熊利荣,郑宇. 基于形态学的稻谷种子品种识别[J].粮油加工,2010(6):45-48.

相似文献/References:

[1]曲歌,陈争光,张庆华.基于无信息变量消除法的水稻种子发芽率测定[J].江苏农业学报,2019,(05):1015.[doi:doi:10.3969/j.issn.1000-4440.2019.05.002]
　QU Ge,CHEN Zheng-guang,ZHANG Qing-hua.Study on germination rate of rice seed based on uninformation variable elimination method[J].,2019,(01):1015.[doi:doi:10.3969/j.issn.1000-4440.2019.05.002]

备注/Memo

备注/Memo:
收稿日期：2020-06-29基金项目：国家自然科学基金项目（61562039)作者简介：黄琼(1997-)，女，江西赣州人，硕士研究生，主要从事图形图像处理研究。（E-mail）406274397@qq.com通讯作者：杨红云，（E-mail）nc_yhy@163.com
更新日期/Last Update: 2021-03-15