J4 ›› 2010, Vol. 48 ›› Issue (06): 987-994.

• 计算机科学 • 上一篇    下一篇

基于方向小波变换的高斯噪声图像恢复方法

张震1,2, 马驷良3, 谭琨4   

  1. 1. 东北石油大学 电气信息工程学院, 黑龙江 大庆 163318|2. 黑龙江省油田控制与信息工程省重点实验室, 黑龙江 大庆 163318|3. 吉林大学 数学
    研究所, 长春 130012;4. 大庆师范学院 计算机科学与信息技术学院, 黑龙江 大庆 163712
  • 收稿日期:2010-04-23 出版日期:2010-11-26 发布日期:2010-11-26
  • 通讯作者: 马驷良 E-mail:masl@jlu.edu.cn

Gauss Noise Image Recovering Method Based onDirectional Wavelet Transform

ZHANG Zhen1,2, MA Siliang3, TAN Kun4   

  1. 1. School of Electrical Engineering and Information, Northeast Petroleum University, Daqing 163318, Heilongjiang Province, China|2. Provincial Key Laboratory of Control and Information Engineering in Oil Field, Daqing 163318, Heilongjiang Province, China|3. Institute of Mathematics, Jilin University, Changchun 130012, China|4. College ofComputer Science and Informational Technology, Daqing Normal University, Daqing 163712, Heilongjiang Province, China
  • Received:2010-04-23 Online:2010-11-26 Published:2010-11-26
  • Contact: MA Siliang E-mail:masl@jlu.edu.cn

摘要:

提出一种基于方向小波变换的相关和阈值去噪方法, 恢复高斯噪声图像. 方向小波去掉了标准二维小波变换仅沿水平和竖直两个方向的限制, 可沿任意方向变换, 这种多方向组合变换方法有利于削弱Gibbs效应, 对去噪后图像的边界保护具有积极作用. 实验结果表明, 相对于标准小波变换, 该方法无论是PSNR值还是视觉效果都较原有的方法更好.

关键词: 方向小波变换; 高斯噪声; 图像恢复

Abstract:

A correlation and threshold denoising method based on the directional wavelet transform was proposed in order to restore the images which were contaminated by the Gaussian noises. The directional wavelet breakthroughs  the restriction of the standard 2D wavelet transforms only along horizontal and vertical directions, can transform along any directions. Not only does it weaken the Gibbs effect, but it can protect image edges.
 The experiment results show our method has better performances than the denoising method based on the standard wavelet in both PSNR and visual effects.

Key words: directional wavelet transform, Guass noise, image denoising

中图分类号: 

  • TP391.41