Multilevel Threshold Based Image Denoising in Curvelet Domain

被引：5

作者：

Nguyen Thanh Binh ^{[1
]}

Ashish Khare ^{[1
]}

机构：

[1] Department of Electronics and Communication, University of Allahabad

来源：

Journal of Computer Science & Technology | 2010年 / 25卷 / 03期

关键词：

curvelet transform; denoising; multilevel thresholding; cycle-spinning;

D O I：

暂无

中图分类号：

TP391.41 [];

学科分类号：

080203 ;

摘要：

In this paper, we propose a multilevel thresholding technique for noise removal in curvelet transform domain which uses cycle-spinning. Most of uncorrelated noise gets removed by thresholding curvelet coeffcients at lowest level, while correlated noise gets removed by only a fraction at lower levels, so we used multilevel thresholding on curvelet coe?cients. The threshold in the proposed method depends on the variance of curvelet coeffcients, the mean and the median of absolute curvelet coeffcients at a particular level which makes it adaptive in nature. Results obtained for 2-D images demonstrate an improved performance over other recent related methods available in literature.

引用

页码：632 / 640

页数：9