基于图像的植物病变叶面特征值提取研究  

Image-based Study on Leaf Characteristic Extraction Value for Plant Disease Lesions

作　　者：程鹏飞[1] 刘静香[1] 周春娥[2] 

机构地区：[1]河南机电高等专科学校机电工程系,河南新乡453002 [2]河南师范大学生命科学学院,河南新乡453007

出　　处：《安徽农业科学》2010年第27期15000-15001,15004,共3页Journal of Anhui Agricultural Sciences

基　　金：河南省教育厅自然科学计划项目(2008B210001)

摘　　要：[目的]运用计算机图像处理技术对生产中的角斑病与斑疹病进行区分研究。[方法]利用计算机视觉技术对植物病变特征进行色度学研究,以颜色及纹理作为植物病害图像特征参数进行病斑图像周长、面积和形状的提取,从而进行病害图像的分类判断。[结果]以CIE1976HIS色调百分率直方图法提取色度特征参数,过程简单、有效,运算速度快,消除了叶片形状大小的影响;利用色调直方图的统计特征参数分析,其色调偏度能够显著地将不同病状区分出来。[结论]该研究认为色调偏度可作为区分角斑病与斑疹病的特征参数。[Objective]The study aimed to make the distinguished research on the angular leaf spot disease and typhus diseas in the agricultural production by using computer image processing technology.[Method]The computer vision technology was used to make the chromatic research on the plant disease characteristics.The color and the texture were taken as the plant disease image characteristic parameter to conduct the extraction of the perimeter,the area and the shape of the disease spot image,thus carrying on the classified judgment of the plant disease image.[Result]The characteristic parameters of the chroma was extracted with the method of CIE1976HIS hue percentage histogram,which was simple,effective and quik in operation speed and eliminated the effect of leaf shapes and sizes.The hue deviation could significantly distinguished the different symptoms by using the statistic parameter analysis of the percentage histogram.[Conclusion]The study suggested that the hue deviation could taken as the characteristic parameters for distinguishing the angular leaf spot disease and typhus diseas.

关 键 词：图像处理 轮廓跟踪 植物病害 特征值提取 色度 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]