去除椒盐噪声的自适应开关加权均值滤波  

Adaptive Switch Weighted Mean Filtering for Salt and Pepper Noise Removal

作　　者：陈初侠[1] 丁勇[2] 刘栎莉[1] 

机构地区：[1]合肥工业大学应用物理系,合肥230009 [2]东南大学国家专用集成电路系统工程技术研究中心,南京210096

出　　处：《计算机工程》2010年第4期210-212,共3页Computer Engineering

基　　金：江苏省专用集成电路设计重点实验室开放课题基金资助项目(JSICK0405)

摘　　要：提出一种去除椒盐噪声的自适应开关加权均值滤波算法。该算法采用一种新的噪声检测方法将图像中的像素分为信号点和噪声点,对检测出的噪声点采用加权均值滤波进行处理,而信号点保持其灰度值不变直接输出。实验结果表明,该算法能在有效去除椒盐噪声的同时保护图像细节,较传统中值滤波及其改算法有更好的滤波性能。This paper proposes Adaptive Switch Weighted Mean（ASWM） filtering algorithm for salt and pepper noise removal. A new noise detection method is adopted to separate the pixels in the image into signal pixels and noise pixels. The noise pixels detected are removed by using the signal pixels with the weighted mean filtering, while signal pixels hold their gray values and are output unprocessed. Experimental results show that the algorithm can suppress noise effectively and preserve the image details. Contrasted to the Traditional Median（TM） filtering and its improved algorithms, ASWM has a better filtering performance.

关 键 词：椒盐噪声 噪声检测 噪声去除 自适应开关 加权均值滤波 

分 类 号：TN911.73[电子电信—通信与信息系统]