首页 | 本学科首页   官方微博 | 高级检索  
     检索      

几种图像去噪算法的对比分析
引用本文:谭攀,姜超.几种图像去噪算法的对比分析[J].测绘与空间地理信息,2014(7):39-42.
作者姓名:谭攀  姜超
作者单位:[1]61287部队,四川成都610036 [2]信息工程大学地理空间信息学院,河南郑州450052
基金项目:国家863项目(2012AA7032031D)资助
摘    要:图像去噪作为一种重要的预处理手段,一直以来都是图像处理和计算机视觉领域的研究热点。去噪算法的难点在于去除噪声的同时尽可能保持原始信息的完整性。本文主要研究TV,BLS-GSM,NLM以及BM3D四种经典去噪算法的基本原理,并用模拟数据验证分析四种算法在图像去噪中的性能,实验证明,四种算法均能在保留图像细节的前提下很好的去除噪声,其中,BM3D算法在保证去噪质量的同时效率最高。

关 键 词:图像去噪  总变分  高斯尺度混合  非局部均值  块匹配三维滤波

Analysis of Several Kinds of Image Denoising Algorithm
TAN Pan,JIANG Chao.Analysis of Several Kinds of Image Denoising Algorithm[J].Geomatics & Spatial Information Technology,2014(7):39-42.
Authors:TAN Pan  JIANG Chao
Institution:1. Troops 61287, Chengdu 610036, China; 2. Information Engineering University, Zhengzhou 450052, China)
Abstract:Image denoising is an important method of preprocessing, it is one of the forelands in the field of Computer Graphic and Computer Vision. The difficulties of denoising algorithm is not only to remove the noise, but also to keep the original information as well. This paper mainly research on the principle of four classic denoising algorithm, which are TV, BLS - GSM, NLM and BM3D, we use simulate data for image denoising to analysis the performance of the four algorithms, experiments demonstrate that the four algorithms can remove the noise, the BM3D algorithm not only have high quality of denosing, but also have the highest efficiency at the same time.
Keywords:image denoising  total variation  gaussian scale mixtures  non-local means  block-matching and 3D filtering
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号