复小波域维纳滤波与偏微分扩散相结合的图像去噪方法

张曈; 朱虹; 张然; 沈憧; 杨向波

doi:10.11834/jig.20090511

图像处理和编码 | 浏览量 : 0 下载量: 319 CSCD: 0

PDF
导出
分享
收藏
专辑

复小波域维纳滤波与偏微分扩散相结合的图像去噪方法
Based on the Dual-tree Complex Wavelet with 6 Directional Wiener Filter Combine Partial Differential Equation in Image Denoising
2009年14卷第5期页码：848
纸质出版：2009
DOI： 10.11834/jig.20090511
稿件说明：

移动端阅览

张曈, 朱虹, 张然, 沈憧, 杨向波. 复小波域维纳滤波与偏微分扩散相结合的图像去噪方法[J]. 中国图象图形学报, 2009,14(5):848. DOI： 10.11834/jig.20090511.

Based on the Dual-tree Complex Wavelet with 6 Directional Wiener Filter Combine Partial Differential Equation in Image Denoising[J]. Journal of Image and Graphics, 2009, 14(5): 848. DOI： 10.11834/jig.20090511.

摘要

针对图像噪声的去除，提出了一种基于复数小波域上的多方向窗维纳滤波与偏微分方程保持边缘细节相结合的方法。针对小波域维纳滤波的方向性差，去噪后图像容易产生哑铃效应

该方法首先进行双树复数小波变换，集中6个方向上的图像信号能量，之后，再在该6个方向上进行方向维纳滤波，对图像进行初步去噪，再以此引导偏微分方程中的扩散函数，实现各项异性进行扩散，最大限度地在保持图像细节的同时，去除噪声。实验结果表明，该方法的峰值信噪比，以及视觉质量都较复小波去噪或各项异性非线性扩散去噪方法有明显的改善。

Abstract

A new image denoising method based on Dual-tree complex wavelet combining partial differential equation in image denoising is presented. Image denoising with Wiener filter can not keep the edges of the image well enough due to the lack of directional ability in common tensor wavelet transform. Authors use the complex wavelet which has stronger directional ability and locally 6 directional Wiener filter to get a “clearer image”

then use “clearer image” guidance the diffusion function of partial differential equation to reduce noise in the image.Results of experiment show: the recovered images have higher PSNR and better visual quality than DTCW and PDE.