基于直觉模糊散度的自适应非局部均值去噪算法  

An Adaptive Nonlocal Mean Denoising Algorithm Based on Intuitionistic Fuzzy Divergence

作　　者：张宁[1] 宋小鹏[1] 刘祎[1] 张鹏程[1] 桂志国[1] ZHANG Ning;SONG Xiaopeng;LIU Yi;Zhang Pengcheng;Gui Zhiguo(Shanxi Provincial Key Laboratory for Biomedical Imaging and Big Data,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学生物医学成像与影像大数据山西省重点实验室

出　　处：《测试技术学报》2020年第1期1-8,共8页Journal of Test and Measurement Technology

基　　金：国家自然科学基金资助项目（61671413,61801438）;国家重大科学仪器设备开发专项资助项目（2014YQ24044508）;中北大学青年学术带头人资助项目（QX201801）

摘　　要：针对传统的非局部均值(NLM)算法在图像去噪时会出现边缘会模糊的问题,提出了一种基于直觉模糊散度的自适应非局部均值(IFD-NLM)去噪算法.该算法利用图像块之间的直觉模糊散度度量非局部图像块的相似性,修正NLM算法的相似性权重,降低不相似图像块之间的干扰,提高了NLM算法相似性权重的准确性.此外,根据图像块内容和直觉模糊散度特征图像,设定相关阈值,自适应地选择滤波参数.实验结果表明,所提算法相对于传统NLM算法能更有效地去除噪声,同时可以保留更多的纹理细节特征和几何结构特征,图像细节信息更为丰富.The edges become blurry while removing noise in image denoising by traditional non-local mean(NLM)algorithm,so this paper propose an adaptive non-local mean denoising algorithm based on intuitionistic fuzzy divergence(IFD-NLM).The algorithm,which uses the intuitionistic fuzzy divergence between image blocks to measure the similarity of non-local image blocks,corrects the similarity weight of the NLM algorithm,reduces interference between dissimilar image blocks,and improves the accuracy of similarity weight of NLM algorithm.In addition,according to image block content and intuitionistic fuzzy divergence feature image,this experiment sets the relevant threshold and adaptively select the filtering parameters.The experimental results show that the proposed algorithm can remove noise more effectively than the traditional NLM algorithm,at the same time,it can retain more texture details and geometric structure features,the image details are more abundant.

关 键 词：图像去噪 直觉模糊散度 非局部均值 滤波参数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]