基于P-M扩散和中值滤波的椒盐噪声去除方法  

Salt-and-pepper noise removal based on Anisotropic diffusion and Median filtration

作　　者：刘继承 王春艳[1] 赵磊[1] 王延民[2] 

机构地区：[1]大庆石油学院电气信息工程学院,黑龙江大庆163318 [2]大庆油田文化集团综合协调部,黑龙江大庆163453

出　　处：《大庆石油学院学报》2008年第1期77-79,123,共3页Journal of Daqing Petroleum Institute

基　　金：黑龙江省普通高等学校青年学术骨干支持计划项目(1151G001);黑龙江省教育厅科学技术研究项目(11511007)

摘　　要：结合中值滤波和P-M扩散,提出先定位、后滤波的椒盐噪声去除方法.该方法通过选取适当的阈值,将图像像素分为噪声像素和信号像素.对噪声像素采取中值滤波算法除噪;对信号像素采取P-M扩散除噪,以保留更多的图像细节.实验表明,该方法能有效去除椒盐噪声,且能较好地保持图像边缘.Based on the combination of Median filtration with Anisotropic diffusion(P-M diffusion),a simple algorithm is presented to remove salt-and-pepper noise from the image.Each pixel is classified to be signal pixel or noise pixel according to the threshold value.Median filtration is employed to the noise pixels,while the signal pixels are adopted P-M diffusion to preserve the detail of the image.Practical experiment proves that the improved algorithm can achieve a better edge preservation than Median filtration and a better noise reduction than P-M diffusion.

关 键 词：椒盐噪声 图像去噪 边缘保持 P—M扩散 中值滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]