去除高密度椒盐噪声的自适应加权均值滤波算法  

Adaptive Weighted Mean Filtering Algorithm for Removing High Density Salt and Pepper Noise

作　　者：查兵 江巨浪 刘国明 韩少刚 黄忠 Zha Bing;Jiang Julang;Liu Guoming;Han Shaogang;Huang Zhong(School of Electronic Engineering and Intelligent Manufacturing,AnqingTeachers University,Anqing,Anhui 246133)

机构地区：[1]安庆师范大学电子工程与智能制造学院,安徽安庆246133

出　　处：《池州学院学报》2020年第6期43-45,共3页Journal of Chizhou University

基　　金：安徽省自然科学基金项目(1908085MF195)。

摘　　要：为了进一步提高对图像椒盐噪声的滤波性能,对现有的自适应均值滤波算法进行了改进。滤波过程分为3个阶段:首先基于像素灰度极值对图像进行噪声检测与噪声密度计算;然后基于滤波窗口中信号点数量自适应扩大窗口尺寸;最后在滤波窗口中进行基于距离权重的加权均值滤波计算,对噪声点像素值更新。采用该算法对包含10%-90%椒盐噪声的图像进行滤波,实验结果表明,该算法的峰值信噪比(PSNR)与结构相似度(SSIM)优于现有其他算法,能够更好地保持图像中的边缘细节。In order to further improve the filtering performance of image salt and pepper noise, the existing adaptive mean filtering algorithm is improved. The filtering process is divided into 3 stages. Firstly, the noise detection and noise density are calculated based on the pixel gray extrema;then the window size is expanded adaptively based on the number of signal points in the filter window;finally, the weighted mean filter calculation based on the distance weight is carried out in the filter window to update the pixel value of the noise point. The results show that the peak signal-to-noise ratio(PSNR) and the structure similarity(SSIM) of the proposed algorithm are excellent with other existing algorithms, which can better keep the edge details in the image.

关 键 词：图像去噪 椒盐噪声 距离 自适应 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]