一种改进的自适应中值滤波算法  

An Improved Method of Adaptive Median Filter

作　　者：王敏[1] 程京[1] 张帆[2] 

机构地区：[1]湖南大学软件学院,长沙410082 [2]湖南大学电气学院,长沙410082

出　　处：《微计算机信息》2010年第10期109-110,119,共3页Control & Automation

摘　　要：传统的自适应中值滤波算法,利用极值判别像素点是否为噪声点。该方法容易将边缘点误判为噪声点,将噪声点误判为信号点。本文针对这一问题,提出一种利用边缘点与噪声点的差异判别噪声点,并将噪声点分为平滑区噪声点与边缘区域噪声点,对不同类型噪声点进行不同处理。从而达到更好的抑制噪声与保护图像细节效果。Traditional adaptive median filter algorithm use extreme value to determine whether pixel the noise points.The method is too easy to misjudge the edge points for the noise,the noise points for the signal.In response to this question,this article use difference between marginal points and noise points to discriminant noise points,the noise points are divided into smooth and edge areas noise points,for different types of noise points in the different treatment.So it can achieve better noise suppression and protection of image detail.

关 键 词：自适应中值滤波 噪声检测 边缘保护 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]