• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2015, Vol. 37 ›› Issue (08): 1546-1550.

• 论文 • 上一篇    下一篇

基于小波阈值的非局部均值去噪

李嘉浪,李华君,徐庆   

  1. (天津大学计算机学院,天津 300072)
  • 收稿日期:2014-04-08 修回日期:2014-07-12 出版日期:2015-08-25 发布日期:2015-08-25
  • 基金资助:

    国家自然科学基金资助项目(U1333110)

Non-local means denoising based on wavelet threshold 

LI Jialang,LI Huajun,XU Qing   

  1. (School of Computer Science,Tianjin University,Tianjin 300072,China)
  • Received:2014-04-08 Revised:2014-07-12 Online:2015-08-25 Published:2015-08-25

摘要:

非局部均值去噪算法充分利用了图像的全局信息,因此比传统的局部去噪算法有着更好的去噪效果。但是,非局部均值去噪算法计算时间复杂度较高,故利用小波阈值的方法对其进行改进,改进后使用非局部均值处理的数据量大幅减小。实验表明,改进后的算法比非局部均值算法去噪效果基本持平,且运行速度更快。

关键词: 非局部均值, 小波阈值滤波, 图像去噪

Abstract:

The non-local means denoising algorithm can use the globe information of the picture, therefore it has better denoising effect than other traditional algorithms. However, since its time complexity is high, we put forth a new non-local means denoising algorithm based on wavelet threshold filter, which use much less data than the traditional non-local means.  Experimental results show that compared to the traditional non-local means, the denoising effect of our algorithm is basically the same, but the running speed is faster.

Key words: non-local means;wavelet threshold filter;image denoising