Position: Home > Articles > Crop Disease Leaf Segmentation Method Based on Intuitional Fuzzy C-means
Journal of Anhui Agricultural Sciences
2019
(5)
233-236
基于直觉模糊C均值聚类算法的作物病害图像分割
作 者：
张晴晴;张云龙;齐国红;李瑶
单 位：
郑州大学西亚斯国际学院
关键词：
IFCM算法;模糊度;作物病害;图像分割
摘 要：
针对作物病害图像的病斑分割问题,提出一种直觉模糊C均值(Intuitional Fuzzy C-means,IFCM)聚类算法。通过引入隶属度、非隶属度和犹豫度3个参数来表示模糊集,从而定义了用来表示模糊集的模糊度的直觉模糊熵(IFE)这一概念,对传统的FCM算法进行改进,克服了FCM算法分割时计算目标函数容易陷入局部极小值,而且聚类数目需要提前设定初值的缺点。将预处理过的作物(以黄瓜为例)病害叶片图像作为研究对象采用该改进算法进行病斑图像分割,得到了很好的分割效果。与其他分割方法进行比较,结果表明该算法分割出来的作物病斑图像准确率高达94%以上,分割效果明显。
译 名：
Crop Disease Leaf Segmentation Method Based on Intuitional Fuzzy C-means
作 者：
ZHANG Qing-qing;ZHANG Yun-long;QI Guo-hong;SIAS International College,Zhengzhou University;
单 位：
ZHANG Qing-qing%ZHANG Yun-long%QI Guo-hong%SIAS International College,Zhengzhou University
关键词：
IFCM;;Fuzzy degree;;Crop diseases;;Image segmentation
摘 要：
According to the disease spots' segmentation problems detected by the images of crops diseases,a new method called Intuitional Fuzzy C-means(IFCM) has been put forward.These three parameters,such as membership degree,non-membership degree and hesitancy degree were used to express fuzzy set.Thus,the concept of IFE,which was used to represent the fuzzy degree of fuzzy set,had been defined.This method improved the traditional FCM algorithm a lot and some disadvantages of FCM,for example,it was easily trapped in local minima when calculating objective function and the clustering number needed to set the initial value in early time had been overcome.Considering the images of pretreated crops(with cucumber as the research object),this improved method was used to segment the crop disease spots,which ensured the segmentation effect.Compared with other segmentation methods,the experimental results showed that the segmentation accuracy rate of crop disease spots was higher than 94%,and the segmentation effect was obvious.
导出引用
计量
文章访问数: 18
HTML全文浏览量: 0
PDF下载量: 0
所属期刊
