图像高椒盐噪声的迭代滤除算法  

Iterative Algorithm for Removing Salt-pepper Noise From Highly Corrupted Images

作　　者：李天翼[1,2] 王明辉[1] 袁培根[2] 刘新国[1] 

机构地区：[1]四川大学计算机学院,成都610065 [2]四川大学制造学院,成都610065

出　　处：《北京工业大学学报》2012年第5期749-754,共6页Journal of Beijing University of Technology

基　　金：民航局科研基金资助项目(MHRD0713)

摘　　要：为有效滤除图像中的高椒盐噪声,提出一种迭代滤波算法.首先采用极值方法检测出噪声点,然后对噪声点以迭代方式逐步滤波,直到噪声点全部清除.利用迭代方式中每一噪声点都能直接或间接利用到图像有用信息的特点,滤波输出始终采用恒定的3×3小邻域,避免了大邻域窗口的诸多弊端.基于图像相关特性,在滤波输出上采用一种基于灰度差的加权均值方式.仿真结果表明,该算法能有效滤除图像中的高椒盐噪声,性能优于其他许多同类算法.An iterative filter algorithm is proposed to filter high density salt-pepper noise in images.It detects the noise pixel with minimum-maximum inspection,and then filters iteratively until all noise pixels are suppressed.To overcome the drawback of enlarged neighborhood,the algorithm takes a uniform 3×3 filtering window all the time,which ensures that all noise pixels be properly restored with enough information because of the trait of iterative manner.The output value is calculated using weighted mean based on each pixel＇s correlation with the filtering pixel.Compared with other methods,simulation results show that the proposed algorithm gives better signal-to-noise ratios（SNR） and more satisfactory images.

关 键 词：迭代滤波 椒盐噪声 加权均值滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]