一种改进的基于模糊判别的非局部均值算法  

An improved non-local means algorithm based on fuzzy discrimination

作　　者：任平红[1] 陈矗[1] 张元科[1] 

机构地区：[1]曲阜师范大学信息科学与工程学院,山东日照276826

出　　处：《电子技术（上海）》2016年第4期28-30,共3页Electronic Technology

基　　金：国家自然科学基金(61572283);山东省优秀中青年科学家科研奖励基金(BS2014DX005);山东省高等学校科技计划项目(J13LN31);曲阜师范大学校级基金(xkj201313)

摘　　要：针对传统的非局部均值去噪算法计算量较大以及噪声严重时均方误差较大的问题,提出了一种改进的基于绝对差值排序模糊判别的非局部均值去噪算法。首先由绝对差值排序判断像素属于噪声的隶属度,然后根据隶属度由像素和与像素具有相似窗口结构的像素重构像素。实验结果表明,该算法的均方误差和信噪比优于传统的非局部均值算法。As the traditional non-local means denosing algorithm involves a large amount of calculation and large mean square error, an improved non-local means denoising algorithm based on fuzzy discrimination of rank-ordered absolute differences is proposed in this paper. First, the fuzzy membership of pixels belongs to the noise is calculated according to the ranked-order absolute differences. Then pixels are reconstructed in accordance with the membership and those pixels that have the similar window structures with the pixels to be reconstructed. The experimental results show that the mean square error and signal-to-noise ratio of the proposed algorithm are superior to the traditional nonlocal means algorithm.

关 键 词：非局部均值 绝对差值排序 模糊判别 加权平均 

分 类 号：O224[理学—运筹学与控制论]