基于各向异性扩散的降噪新方法

上海大学学报(自然科学版) ›› 2009, Vol. 15 ›› Issue (4): 346-350.

基于各向异性扩散的降噪新方法

赖新平胡海平

上海大学理学院,上海 200444

收稿日期:2008-04-22 出版日期:2009-08-30 发布日期:2009-08-30
通讯作者: 胡海平(1966~),男,副教授,研究方向为图像处理. E-mail:hu_jack@staff.shu.edu.cn
基金资助:
国家自然科学基金资助项目（10572080）

New Method for Noise Removal Based on Anisotropic Diffusion

College of Sciences, Shanghai University, Shanghai 200444, China

Received:2008-04-22 Online:2009-08-30 Published:2009-08-30

摘要/Abstract

摘要：

提出一种基于各向异性扩散的图像降噪新方法.该方法能够避免Perona和Malik以及You和Kaveh等人提出的方法中对参数选择的困难，而且也能够克服Catte方案对图像进行预平滑的过程中所面临的问题.实验结果表明，该方法在图像增强过程中的降噪和细节保留方面都能有较好的效果.

关键词: 图像去噪；各向异性扩散；偏微分方程；细节保留

Abstract:

In this paper, we discuss deficiency of the PeronaMalik scheme and some other schemes derived from it, and propose a new operator. The proposed method can avoid difficulties in choosing parameters encountered in Perona-Malik and You-Kaveh schemes, and overcome the pre-smoothing problems in the Catte scheme. Experimental results show that the new scheme can achieve good results in noise removal and detail preservation in image smoothing.

Key words: image removal; anisotropic diffusion; partial differential equation; detail preservation

中图分类号:

TP391

赖新平胡海平. 基于各向异性扩散的降噪新方法[J]. 上海大学学报(自然科学版), 2009, 15(4): 346-350.

LAI Xin-ping, HU Hai-ping. New Method for Noise Removal Based on Anisotropic Diffusion[J]. Journal of Shanghai University（Natural Science Edition）, 2009, 15(4): 346-350.