一种基于各向同性集的模糊滤波算法  

A fuzzy filtering algorithm based on isotropy set

作　　者：邵志勇[1] 张学东[1] 

机构地区：[1]鞍山科技大学计算机科学与工程学院,辽宁鞍山114005

出　　处：《大连理工大学学报》2004年第2期301-304,共4页Journal of Dalian University of Technology

基　　金：辽宁省自然科学基金资助项目(20022163).

摘　　要：图像中的椒盐噪声是一种特殊的噪声,处理不当会使图像本身的细节变得模糊不清,从而使图像降质.中值滤波能够有效地抑制椒盐噪声,但模糊了图像中的一些细节.因此结合方向图,提出了一种基于各向同性集的模糊滤波算法,不仅能够有效地抑制椒盐噪声,而且能够较好地保持图像细节.该算法大大优于中值滤波.Salt & pepper noise is a kind of special noise. If it is processed improperly, the image′s details will be blurred, and its quality degenerated. The median filtering method can suppress salt & pepper noise, but the image′s details are blurred at the same time. To resolve this problem, a fuzzy approach is presented based on the image′s isotropy set, combined with its directional image. It can suppress salt & pepper noise effectively and keep images′ border details better. Experiments indicate that the method is more effective to suppress this noise, and has a better behaviour in holding images′ detail features than the median filtering method.

关 键 词：各向同性集 椒盐噪声 模糊逻辑 中值滤波 方向图 图象处理 

分 类 号：TN911.73[电子电信—通信与信息系统]