摘要
本文提出了一个基于二维直方图的图象分割模糊聚类方法,它除了考虑象素点的灰度信息外还考虑了象素点与其邻域的空间相关信息,利用模糊C均值(FCM)聚类算法得到象素点的隶属度,并由各象素点的隶属度实现图象分割.实验结果表明,本文提出的方法与Otsu法和熵函数法相比,错分的象素点数大约减少了四分之三.
A fuzzy clustering method for image segmentation based on two-dimensional histogram is presented. It utilizes the gray level information of each pixel and its spatial correlation information within the neighborhood. By using the fuzzy c-means (FCM) clustering algorithm, the membership of pixels is obtained and then the image is segmented in terms of the values of the membership. The experimental results show that the misclassification pixel number using the proposed method decreases about by a factor of 4 compared with the Otsu method or the entropic method.
出处
《电子学报》
EI
CAS
CSCD
北大核心
1992年第9期40-46,共7页
Acta Electronica Sinica
关键词
二维直方图
图象分割
模糊聚类法
Fuzzy c-means clustering algorithm, Two-dimensional histogram. Image segmentation
