FCM和LS-SVM相结合的粮虫图像分割算法  

A New Stored-Product Pests Recognition Algorithm based on FCM and LS-SVM

作　　者：党培[1] 谭联[2] 

机构地区：[1]河南工业大学机电工程学院,郑州450007 [2]华北水利水电学院,郑州450011

出　　处：《电脑开发与应用》2009年第10期7-9,共3页Computer Development & Applications

基　　金：河南工业大学科研基金资助项目(06XJC005)

摘　　要：模糊C均值(FCM)聚类算法广泛用于图像的自动分割,但是该算法对噪声十分敏感。因此提出一种改进的算法,在传统的FCM聚类的基础上,结合LS-SVM对图像进行分类。该方法对噪声有很强的鲁棒性和对像素聚类的正确性。Fuzzy c-means （FCM） clustering algorithm has been widely used in automated image segmentation. However, the conventional FCM algorithm is sensitive to noise. A new image segmentation＇s algorithm based on with the FCM cluster and LSSVM is proposed. Experimental results on the test indicate that the segmentations are more aceurate and robust than the standard FCM algorithm.

关 键 词：模糊C均值 图像分割 最小二乘支持向量机 灰度共生矩阵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]