摘要
为快速辨别植物叶片病害种类并对症下药,将四种植物常见叶片病害作为识别对象,设计搭建了一个植物叶片病害识别系统,先对病斑图像进行对比度增强等预处理,然后使用K-means聚类分割算法,在HSI颜色空间和Lab颜色空间内进行对比,找出最优方法后将病斑区域从叶片中分割出。通过提取病斑的同质性、能量、近似熵、对比度等13维特征,采用SVM算法完成四种病害种类的识别,识别率达90.67%。
In order to quickly identify plant disease types and take appropriate medicine,four common plant diseases are used as recognition objects.A plant leaf disease recognition system is designed and built.Firstly,contrast enhancement is performed on the im⁃age of disease spot,and then k-means clustering segmentation algorithm is used to compare the image in HSI color space and Lab color space.After finding out the best method,the spot area was separated from the leaf.By extracting 13 dimensional features such as homo⁃geneity,energy,approximate entropy and contrast of disease spots,SVM algorithm was used to complete the recognition of four disease types,and the recognition rate reached 90.67%.
作者
张铄
谢裕睿
董建娥
Zhang Shuo;Xie Yurui;Dong Jian’e(College of Big Data and Intelligent Engineering,Southwest Forestry University,Kunming,Yunnan 650224)
出处
《现代计算机》
2021年第34期112-116,共5页
Modern Computer
基金
国家级大学生创新创业训练计划平台项目(202010677011)
云南省农业基础研究联合专项青年项目(2018FG001-101)。
关键词
图像分割
特征提取
SVM分类器
病害识别
image segmentation
feature extraction
SVM classifier
disease recognition
