基于GLCM纹理特征提取的黄瓜叶部病害检测算法研究  

Research on cucumber leaf disease detection algorithm based on GLCM texture feature extraction

作　　者：李亚文[1] 刘爱军[1] 陈垚[1] LI Ya-wen;LIU Ai-jun;CHEN Yao(Electronic Information and Electrical Engineering College,Shangluo University/Smart Agricultural Technology and Application Research Center of Shangluo,Shangluo 726000,Shaanxi,China)

机构地区：[1]商洛学院电子信息工程与电气工程学院/商洛市智慧农业与技术应用研究中心,陕西商洛726000

出　　处：《湖北农业科学》2022年第9期141-145,共5页Hubei Agricultural Sciences

基　　金：商洛市科技计划重点项目(19SLKJ121);商洛学院科研创新团队项目(19SXC03)。

摘　　要：针对传统的植物叶部病害检测算法复杂的特点,提出了一种基于GLCM纹理特征提取的植物叶部病害检测算法。以黄瓜叶部炭疽病为研究对象,利用K-means聚类算法进行图像阈值分割,并利用灰度共生矩阵提取样本的能量均值、熵均值、对比度均值和相关均值等4种纹理特征参数,通过参数训练,确定无病害区和有病害区参数的区域,进而判定样本的病害情况。结果表明该算法实现效率高、鲁棒性较好。To the complex algorithm of traditional plant leaf disease detection,this paper proposed a plant leaf disease detection algorithm based on GLCM texture feature extraction. As the research object of cucumber leaf anthracnose,the K-means clustering algorithm was used to perform image threshold segmentation,and the gray level co-occurrence matrix was used to extract the energy mean,entropy mean,contrast mean and correlation mean of the sample. With the parameter training,the area of disease-free area and diseased area parameters were determined,and then the disease condition of the sample was judged. The results showed that the algorithm had high efficiency and good robustness.

关 键 词：纹理特征 灰度共生矩阵 聚类算法 图像分割 植物叶部病害 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]