基于SVM和AdaBoost的棉叶螨危害等级识别
|
摘要
针对自然条件下棉叶螨虫害等级识别难的问题,在自然条件下以普通手机采集棉叶图像作为实验对象,首先使用大津法和连通区域标记算法,将棉花叶片图像与背景分离,然后,提取不同棉叶螨危害等级棉叶图像的颜色、纹理和边缘特征数据,使用支持向量机(Support vector machine,SVM)单独进行分类实验,得到平均识别正确率为76. 25%,最后,采用SVM和AdaBoost相结合的算法,生成最优判别模型,实现对棉叶螨危害等级的识别,平均识别正确率为88. 75%。对比实验表明,提出的棉叶螨危害等级识别方法比BP神经网络的平均识别正确率高13. 75个百分点,比单独采用SVM算法高12. 5个百分点,比单独采用AdaBoost算法高8. 75个百分点,SVM和AdaBoost相结合的算法可较好地对棉叶螨危害等级进行识别,为棉叶螨数字化防治和变量喷药提供了数据支持。
Aiming at the difficulty in identifying the level of cotton spider mites under natural conditions,an automatic identification method was proposed for rapid detection of cotton spider mites damage under natural conditions. The damaged cotton leaves images collected by mobile phone under natural conditions were used as the object. Firstly,the Otsu method and the regional interconnection marking algorithm were used to separate image of cotton leaf from background. Then,the authors combined the colors,textures,and edge features of the image of damaged cotton spider mites. The support vector machine( SVM) was used to classify the data separately. The average recognition rate of 76. 25% was obtained.Finally,it was tried to recognize the mode based on combining the SVM and AdaBoost algorithm to classify the cotton spider mites hazard criteria. With this mode,the average recognition accuracy rate finally reached 88. 75%,which was 13. 75 percentage points higher than that of BP neural network,12. 5 percentage points higher than that of the SVM algorithm alone and 8. 75 percentage points higher than that of the AdaBoost algorithm alone with comparative experiments. In conclusion,it was fully proved that the identification method mentioned can be used to better identify the cotton spider mites damage level,which provided data support for the digital control of cotton spider mites and variable spraying.
Aiming at the difficulty in identifying the level of cotton spider mites under natural conditions,an automatic identification method was proposed for rapid detection of cotton spider mites damage under natural conditions. The damaged cotton leaves images collected by mobile phone under natural conditions were used as the object. Firstly,the Otsu method and the regional interconnection marking algorithm were used to separate image of cotton leaf from background. Then,the authors combined the colors,textures,and edge features of the image of damaged cotton spider mites. The support vector machine( SVM) was used to classify the data separately. The average recognition rate of 76. 25% was obtained.Finally,it was tried to recognize the mode based on combining the SVM and AdaBoost algorithm to classify the cotton spider mites hazard criteria. With this mode,the average recognition accuracy rate finally reached 88. 75%,which was 13. 75 percentage points higher than that of BP neural network,12. 5 percentage points higher than that of the SVM algorithm alone and 8. 75 percentage points higher than that of the AdaBoost algorithm alone with comparative experiments. In conclusion,it was fully proved that the identification method mentioned can be used to better identify the cotton spider mites damage level,which provided data support for the digital control of cotton spider mites and variable spraying.
引文
[1]段利.关于新疆棉花产业发展的战略思考[J].新疆社会科学,2011(1):43-45.DUAN Li.Strategic thinking about the development of cotton industry in Xinjiang[J].Xinjiang Social Science,2011(1):43-45.(in Chinese)
[2]杨忠娜,唐继军,喻晓玲.新疆棉花产业对国民经济的影响及对策研究[J].农业现代化研究,2013,34(3):298-302.YANG Zhongna,TANG Jijun,YU Xiaoling.Xinjiang cotton industry present situation and countermeasure research[J].Research on Agricultural Modernization,2013,34(3):298-302.(in Chinese)
[3]武予清,刘芹轩,钟昌珍.不同棉花品种苗期对朱砂叶螨抗性的筛选鉴定[J].河南农业大学学报,1997,31(3):18-21.WU Yuqing,LIU Qinxuan,ZHONG Changzhen.Screening and identification of the resistance of different cotton varieties to Tetranychus cinnabaris at seedling stage[J].Acta Agriculturae Univeritatis Henanensis,1997,31(3):18-21.(in Chinese)
[4]陈鹏程,张建华,李眉眉,等.土耳其斯坦叶螨为害棉叶的生理变化及光谱特征分析[J].昆虫知识,2007,44(1):61-65.CHEN Pengcheng,ZHANG Jianhua,LI Meimei,et al.Physiological change and hyperspectral character analysis of cotton leaves infested by Tetranvchus turkestani[J].Chinese Bulletin of Entomology,2007,44(1):61-65.(in Chinese)
[5]杨德松,姬华,王星,等.影响新疆棉叶螨发生因素的研究[J].中国棉花,2004,31(6):10-11.YANG Desong,JI Hua,WANG Xing,et al.Study on the factors affecting the occurrence of mites in Xinjiang cotton[J].Chinese Cotton,2004,31(6):10-11.(in Chinese)
[6]张建华,孔繁涛,李哲敏,等.基于最优二叉树支持向量机的蜜柚叶部病害识别[J].农业工程学报,2014,30(19):222-231.ZHANG Jianhua,KONG Fantao,LI Zhemin,et al.Recognition of honey pomelo leaf diseases based on optimal binary tree support vector machine[J].Transactions of the CSAE,2014,30(19):222-231.(in Chinese)
[7]何青海,马本学,瞿端阳,等.基于机器视觉棉叶螨自动监测与分级方法研究[J].农机化研究,2013,35(4):152-155.HE Qinghai,MA Benxue,QU Duanyang,et al.Research of cotton spider mites automatic detection and classification which based on machine vision[J].Journal of Agricultural Mechanization Research,2013,35(4):152-155.(in Chinese)
[8]魏丽冉,岳峻,李振波,等.基于核函数支持向量机的植物叶部病害多分类检测方法[J/OL].农业机械学报,2017,48(增刊):166-171.WEI Liran,YUE Jun,LI Zhenbo,et al.Multi-classification detection method of plant leaf disease based on kernel function SVM[J/OL].Transactions of the Chinese Society for Agricultural Machinery,2017,48(Supp.):166-171.http:∥www.j-csam.org/jcsam/ch/reader/view_abstract.aspx?file_no=2017s027&flag=1&journal_id=jcsam.DOI:10.6041/j.issn.1000-1298.2017.S0.027.(in Chinese)
[9]MARTIN D,LATHEEF M.Active optical sensor assessment of spider mite damage on greenhouse beans and cotton[J].Experimental and Applied Acarology,2018,74(2):147-158.
[10]CAMARGO A,SMITH J.Image pattern classification for the identification of disease causing agents in plants[J].Computers and Electronics in Agriculture,2009,66(2):121-125.
[11]BARBEDO J,KOENIGKAN L,SANTOS T T.Identifying multiple plant diseases using digital image processing[J].Biosystems Engineering,2016,147:104-116.
[12]张建华,冀荣华,袁雪,等.基于径向基支持向量机的棉花虫害识别[J].农业机械学报,2011,42(8):178-183.ZHANG Jianhua,JI Ronghua,YUAN Xue,et al.Recognition of pest damage for cotton leaf based on RBF-SVM algorithm[J].Transactions of the Chinese Society for Agricultural Machinery,2011,42(8):178-183.(in Chinese)
[13]GB/T 15802-2011棉花叶瞒测报技术规范[S].2011.
[14]李凯,冯全,张建华.棉花苗叶片复杂背景图像的联合分割算法[J].计算机辅助设计与图形学报,2017,29(10):1871-1880.LI Kai,FENG Quan,ZHANG Jianhua.Co-segmentation algorithm for complex background image of cotton seedling leaves[J].Journal of Computer-Aided Design&Computer Graphics,2017,29(10):1871-1880.(in Chinese)
[15]符祥,郭宝龙.图像插值技术综述[J].计算机工程与设计,2009,30(1):141-144.FU Xiang,GUO Baolong.Overview of image interpolation technology[J].Computer Engineering&Design,2009,30(1):141-144.(in Chinese)
[16]卜文斌,游福成,李泉,等.一种基于大津法改进的图像分割方法[J].北京印刷学院学报,2015,23(4):76-78.BU Wenbin,YOU Fucheng,LI Quan,et al.An improved image segmentation method based on Otsu[J].Journal of Beijing Institute of Graphic Communication,2015,23(4):76-78.(in Chinese)
[17]张善文,张晴晴,齐国红,等.基于改进K中值聚类的苹果病害叶片分割方法[J].江苏农业科学,2017,45(18):205-208.ZHANG Shanwen,ZHANG Qingqing,QI Guohong,et al.Segmentation method of apple disease leaves based on improved Kmedian clustering[J].Jiangsu Agricultural Sciences,2017,45(18):205-208.(in Chinese)
[18]孙君顶,崔江涛,毋小省,等.基于颜色和形状特征的彩色图像检索方法[J].中国图象图形学报,2004,9(7):56-63.SUN Junding,CUI Jiangtao,WU Xiaosheng,et al.Color image retrieval based on color and shape features[J].Journal of Image and Graphics,2004,9(7):56-63.(in Chinese)
[19]高程程,惠晓威.基于灰度共生矩阵的纹理特征提取[J].计算机系统应用,2010,19(6):195-198.GAO Chengcheng,HUI Xiaowei.GLOM-based texture feature extraction[J].Computer Systems&Applications,2010,19(6):195-198.(in Chinese)
[20]刘丽,匡纲要.图像纹理特征提取方法综述[J].中国图象图形学报,2009,14(4):622-635.LIU Li,KUANG Gangyao.Overview of image textural of feature extraction methods[J].Journal of Image and Graphics,2009,14(4):622-635.(in Chinese)
[21]杨婷婷,顾梅花,章为川,等.彩色图像边缘检测研究综述[J].计算机应用研究,2015,32(9):2566-2571.YANG Tingting,GU Meihua,ZHANG Weichuan,et al.Review of color image edge detection algorithms[J].Application Research of Computers,2015,32(9):2566-2571.(in Chinese)
[22]YANG J,YANG X,ZHANG J.A parallel multi-class classification support vector machine based on sequential minimal optimization[C]∥International Multi-Symposiums on Computer and Computational Sciences,2008:443-446.
[23]周志华.机器学习[M].北京:清华大学出版社,2016.
[24]LI T,ZHO S,OGIHARA M.Using discriminant analysis for multi-class classification:an experimental investigation[J].Knowledge&Information Systems,2006,10(4):453-472.
[25]BATISTA G,BAZZAN A,MONARD M.Balancing training data for automated annotation of keywords:a case study[C]∥Brazilian Workshop on Bioinformatics,DBLP,2003:10-18.
[26]RTSCH G,ONADA T,MULLER K.Soft margins for Ada Boost[J].Machine Learning,2001,42(3):287-320.
[2]杨忠娜,唐继军,喻晓玲.新疆棉花产业对国民经济的影响及对策研究[J].农业现代化研究,2013,34(3):298-302.YANG Zhongna,TANG Jijun,YU Xiaoling.Xinjiang cotton industry present situation and countermeasure research[J].Research on Agricultural Modernization,2013,34(3):298-302.(in Chinese)
[3]武予清,刘芹轩,钟昌珍.不同棉花品种苗期对朱砂叶螨抗性的筛选鉴定[J].河南农业大学学报,1997,31(3):18-21.WU Yuqing,LIU Qinxuan,ZHONG Changzhen.Screening and identification of the resistance of different cotton varieties to Tetranychus cinnabaris at seedling stage[J].Acta Agriculturae Univeritatis Henanensis,1997,31(3):18-21.(in Chinese)
[4]陈鹏程,张建华,李眉眉,等.土耳其斯坦叶螨为害棉叶的生理变化及光谱特征分析[J].昆虫知识,2007,44(1):61-65.CHEN Pengcheng,ZHANG Jianhua,LI Meimei,et al.Physiological change and hyperspectral character analysis of cotton leaves infested by Tetranvchus turkestani[J].Chinese Bulletin of Entomology,2007,44(1):61-65.(in Chinese)
[5]杨德松,姬华,王星,等.影响新疆棉叶螨发生因素的研究[J].中国棉花,2004,31(6):10-11.YANG Desong,JI Hua,WANG Xing,et al.Study on the factors affecting the occurrence of mites in Xinjiang cotton[J].Chinese Cotton,2004,31(6):10-11.(in Chinese)
[6]张建华,孔繁涛,李哲敏,等.基于最优二叉树支持向量机的蜜柚叶部病害识别[J].农业工程学报,2014,30(19):222-231.ZHANG Jianhua,KONG Fantao,LI Zhemin,et al.Recognition of honey pomelo leaf diseases based on optimal binary tree support vector machine[J].Transactions of the CSAE,2014,30(19):222-231.(in Chinese)
[7]何青海,马本学,瞿端阳,等.基于机器视觉棉叶螨自动监测与分级方法研究[J].农机化研究,2013,35(4):152-155.HE Qinghai,MA Benxue,QU Duanyang,et al.Research of cotton spider mites automatic detection and classification which based on machine vision[J].Journal of Agricultural Mechanization Research,2013,35(4):152-155.(in Chinese)
[8]魏丽冉,岳峻,李振波,等.基于核函数支持向量机的植物叶部病害多分类检测方法[J/OL].农业机械学报,2017,48(增刊):166-171.WEI Liran,YUE Jun,LI Zhenbo,et al.Multi-classification detection method of plant leaf disease based on kernel function SVM[J/OL].Transactions of the Chinese Society for Agricultural Machinery,2017,48(Supp.):166-171.http:∥www.j-csam.org/jcsam/ch/reader/view_abstract.aspx?file_no=2017s027&flag=1&journal_id=jcsam.DOI:10.6041/j.issn.1000-1298.2017.S0.027.(in Chinese)
[9]MARTIN D,LATHEEF M.Active optical sensor assessment of spider mite damage on greenhouse beans and cotton[J].Experimental and Applied Acarology,2018,74(2):147-158.
[10]CAMARGO A,SMITH J.Image pattern classification for the identification of disease causing agents in plants[J].Computers and Electronics in Agriculture,2009,66(2):121-125.
[11]BARBEDO J,KOENIGKAN L,SANTOS T T.Identifying multiple plant diseases using digital image processing[J].Biosystems Engineering,2016,147:104-116.
[12]张建华,冀荣华,袁雪,等.基于径向基支持向量机的棉花虫害识别[J].农业机械学报,2011,42(8):178-183.ZHANG Jianhua,JI Ronghua,YUAN Xue,et al.Recognition of pest damage for cotton leaf based on RBF-SVM algorithm[J].Transactions of the Chinese Society for Agricultural Machinery,2011,42(8):178-183.(in Chinese)
[13]GB/T 15802-2011棉花叶瞒测报技术规范[S].2011.
[14]李凯,冯全,张建华.棉花苗叶片复杂背景图像的联合分割算法[J].计算机辅助设计与图形学报,2017,29(10):1871-1880.LI Kai,FENG Quan,ZHANG Jianhua.Co-segmentation algorithm for complex background image of cotton seedling leaves[J].Journal of Computer-Aided Design&Computer Graphics,2017,29(10):1871-1880.(in Chinese)
[15]符祥,郭宝龙.图像插值技术综述[J].计算机工程与设计,2009,30(1):141-144.FU Xiang,GUO Baolong.Overview of image interpolation technology[J].Computer Engineering&Design,2009,30(1):141-144.(in Chinese)
[16]卜文斌,游福成,李泉,等.一种基于大津法改进的图像分割方法[J].北京印刷学院学报,2015,23(4):76-78.BU Wenbin,YOU Fucheng,LI Quan,et al.An improved image segmentation method based on Otsu[J].Journal of Beijing Institute of Graphic Communication,2015,23(4):76-78.(in Chinese)
[17]张善文,张晴晴,齐国红,等.基于改进K中值聚类的苹果病害叶片分割方法[J].江苏农业科学,2017,45(18):205-208.ZHANG Shanwen,ZHANG Qingqing,QI Guohong,et al.Segmentation method of apple disease leaves based on improved Kmedian clustering[J].Jiangsu Agricultural Sciences,2017,45(18):205-208.(in Chinese)
[18]孙君顶,崔江涛,毋小省,等.基于颜色和形状特征的彩色图像检索方法[J].中国图象图形学报,2004,9(7):56-63.SUN Junding,CUI Jiangtao,WU Xiaosheng,et al.Color image retrieval based on color and shape features[J].Journal of Image and Graphics,2004,9(7):56-63.(in Chinese)
[19]高程程,惠晓威.基于灰度共生矩阵的纹理特征提取[J].计算机系统应用,2010,19(6):195-198.GAO Chengcheng,HUI Xiaowei.GLOM-based texture feature extraction[J].Computer Systems&Applications,2010,19(6):195-198.(in Chinese)
[20]刘丽,匡纲要.图像纹理特征提取方法综述[J].中国图象图形学报,2009,14(4):622-635.LIU Li,KUANG Gangyao.Overview of image textural of feature extraction methods[J].Journal of Image and Graphics,2009,14(4):622-635.(in Chinese)
[21]杨婷婷,顾梅花,章为川,等.彩色图像边缘检测研究综述[J].计算机应用研究,2015,32(9):2566-2571.YANG Tingting,GU Meihua,ZHANG Weichuan,et al.Review of color image edge detection algorithms[J].Application Research of Computers,2015,32(9):2566-2571.(in Chinese)
[22]YANG J,YANG X,ZHANG J.A parallel multi-class classification support vector machine based on sequential minimal optimization[C]∥International Multi-Symposiums on Computer and Computational Sciences,2008:443-446.
[23]周志华.机器学习[M].北京:清华大学出版社,2016.
[24]LI T,ZHO S,OGIHARA M.Using discriminant analysis for multi-class classification:an experimental investigation[J].Knowledge&Information Systems,2006,10(4):453-472.
[25]BATISTA G,BAZZAN A,MONARD M.Balancing training data for automated annotation of keywords:a case study[C]∥Brazilian Workshop on Bioinformatics,DBLP,2003:10-18.
[26]RTSCH G,ONADA T,MULLER K.Soft margins for Ada Boost[J].Machine Learning,2001,42(3):287-320.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |