基于改进模糊聚类算法鲁棒的图像分割

被引:15
作者
张扬
王士同
韩斌
机构
[1] 江南大学信息工程学院
关键词
图像分割; 模糊聚类; 鲁棒性; Voronoi cell;
D O I
暂无
中图分类号
TP391.41 [];
学科分类号
080203 ;
摘要
对噪声图像提出了一种改进的模糊聚类分割算法。因为模糊C均值聚类(FCM)算法具有对噪声数据敏感的缺点,该算法通过提升意义更趋明晰的模糊隶属度来改变模糊聚类中的目标函数,即通过在标准的FCM算法中使用到类的Voronoi cell的距离来取代到类的原型的欧氏距离,从而增强了聚类结果的鲁棒性。实验结果表明,改进的算法较之于FCM对于噪声图像的分割有更好的鲁棒性。
引用
收藏
页码:911 / 917
页数:7
相关论文
共 8 条
[1]  
Fuzzy classification of remote sensing images. Wang F. IEEE Tranactions on Geoscience and Remote Sensing . 1990
[2]  
A Possibilistic Approach to Clustering. Krishnapuram R,Keller J M. IEEE Transactions on Fuzzy Systems . 1993
[3]  
Image segmentation based onadaptive cluster prototype estimation. Liew A W,Yan H,Law N F. IEEE Transactions onFuzzy Systems . 2005
[4]  
Thresholding using two-dimen-sional histogram and fuzzy entropy principle. Cheng H D,Chen Y H,Jiang X H. IEEE trans.Ima-ge Processing . 2000
[5]  
A convergence theorem for the fuzzy ISODATA clustering algorithms. Bezdek J C. IEEE Transactions on Pattern Analysis and Machine Intelligence . 1980
[6]  
Alternating cluster estimation:a new tool for clustering and function approximation. Runkler TA,Bezdek J C. IEEE Transactions on Fuzzy Systems . 1999
[7]  
A technique of three-level thresholding based on probabilistic partition and fuzzy three-partition. Zhao M S,Fu A,Yan H. IEEE Transactions on Fuzzy Systems . 2001
[8]  
Improved fuzzy partitions for fuzzy regression models. Hoppner F,Klawonn F. International Journal of Approximate Reasoning . 2003