数字图像处理技术在农作物病虫草识别中的应用  

Application of Digital Image Processing Technology in Recognizing the Diseases,Pests,and Weeds from Crops

作　　者：管泽鑫[1] 姚青[1] 杨保军[2] 胡洁[1] 唐健[2] 

机构地区：[1]浙江理工大学信息电子学院,杭州310018 [2]中国水稻研究所水稻生物学国家重点实验室,杭州310006

出　　处：《中国农业科学》2009年第7期2349-2358,共10页Scientia Agricultura Sinica

基　　金：国家自然科学基金(30600398);“十一五”国家重大科技支撑计划课题(2006BAD08A04);中国农业科学院院所基金(2007-RG016)

摘　　要：综述了近年来国内外在农作物病虫草识别中应用的主要图像处理方法,包括颜色空间区分法、纹理特征分析法、形态特征分析法、小波分析法、多种参数结合分析法以及特殊图像的分析法,概述了各种方法的原理,比较了各种处理方法的优缺点,并对相关的模式识别方法做了简要概括,最后指出目前研究存在的问题及今后的研究方向。The major image processing methods used for the reorganization of diseases, pests, and weeds from crops were reviewed. These methods include color space partition method, texture feature analysis method, morphological feature analysis method, wavelet analysis methods, multi-parametric analysis method and special image analysis method. The principles of these methods were introduced and their advantages and disadvantages were compared. In addition, the pattern recognition methods that related to the above methods were also summarized briefly. Finally, some concerns as well as further researches in this field were discussed.

关 键 词：数字图像处理 农作物 病虫草 识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]