基于图像处理的玉米叶部染病程度的研究  

Investigation on the Infectious Degree of Maize Leaf Based on Image Processing

作　　者：马晓丹[1] 关海鸥[1] 黄燕[2] 

机构地区：[1]黑龙江八一农垦大学信息学院,黑龙江大庆163319 [2]黑龙江八一农垦大学工程学院,黑龙江大庆163319

出　　处：《农机化研究》2009年第11期102-103,110,共3页Journal of Agricultural Mechanization Research

基　　金：黑龙江省教育厅资助项目(11531256)

摘　　要：以玉米小斑病为例,探讨利用图像技术实现玉米叶部染病程度的计算方法。根据玉米叶部病害特点,综合应用阈值法、区域标记方法,对玉米叶部病害图片进行图像分割、统计病斑个数、去除冗余斑点、计算病斑形状特征。研究结果表明,利用图像技术进行玉米叶部染病程度的计算是可行的,为有的放矢地进行农药喷洒提供了理论依据。Set maize leaf spot for example, recognition and diagnosis methods of infectious degree of maize leaf using machine vision were studied in the ＂paper. A threshold method was adopted for image segmentation. An area - marking method was used calculating the num of disease as well as wiping off redundancy dots. It shows that this method is available for calculating the infectious degree, and provides theory support for insecticide spraying.

关 键 词：玉米叶部病害 区域标记 图像分割 形状特征 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]