【图像降噪】基于定向M波段双树（希尔伯特）小波对多分量/彩色图像进行降噪（Matlab代码实现）

 👨‍🎓个人主页：研学社的博客    

💥💥💞💞欢迎来到本博客❤️❤️💥💥

🏆博主优势：🌞🌞🌞博客内容尽量做到思维缜密，逻辑清晰，为了方便读者。

⛳️座右铭：行百里者，半于九十。

📋📋📋本文目录如下：🎁🎁🎁

目录

💥1 概述

📚2 运行结果

🎉3 参考文献

🌈4 Matlab代码实现

---

💥1 概述

文献来源：

 我们提出了基于希尔伯特小波对的双树分解结构的M波段情况的二维推广（最初由Kingsbury提出，由Selesnick进一步研究）。我们特别关注：1）对偶基础的构建和2）由此产生的方向分析。我们还回顾了M波段情况下必要的预处理阶段。虽然由于表示的冗余，可以进行多次重建，但我们提出了一种新的最佳信号重建技术，该技术最大限度地减少了潜在的估计误差。通过对几种图像类型（自然、纹理、地震）的去噪比较，以及各种M波段小波和阈值策略，证明了所提出的M波段分解的有效性。在整体降噪和方向保留方面都取得了显著改善。

原文摘要：

Abstract:

We propose a two-dimensional generalization to the M-band case of the dual-tree decomposition structure (initially proposed by Kingsbury and further investigated by Selesnick) based on a Hilbert pair of wavelets. We particularly address: 1) the construction of the dual basis and 2) the resulting directional analysis. We also revisit the necessary pre-processing stage in the M-band case. While several reconstructions are possible because of the redundancy of the representation, we propose a new optimal signal reconstruction technique, which minimizes potential estimation errors. The effectiveness of the proposed M-band decomposition is demonstrated via denoising comparisons on several image types (natural, texture, seismics), with various M-band wavelets and thresholding strategies. Significant improvements in terms of both overall noise reduction and direction preservation are observed.

实现了一个参数非线性估计器，该估计器推广了几种小波收缩去噪方法。它致力于加性高斯噪声，采用多元统计方法考虑不同小波子带之间存在的空间和分量间相关性，使用斯坦无偏风险估计器（SURE）原理推导出最佳参数。小波选择是一个稍微冗余的多波段几何双小波框架。多光谱遥感影像实验优于传统的小波去噪技术（包括曲线）。

📚2 运行结果

 

 

部分代码：

%description
disp('This program decomposes an image via a dual-tree M-band transform and reconstructs the obtained coeffients.');

m='haar';
M1=2;M2=2;
rm=2;

load('barbpgm.mat')
x=barb;

x=x(1:M1^rm*floor(size(x,1)/M1^rm),1:M2^rm*floor(size(x,2)/M2^rm));

%decomposition
[wp,wm,H1,H2,HH1,HH2,G,Gh]=dualtreeM2D_freq(x,m,rm);

%plots of the coefficients
plot_dualtree_Mband2D(wp)
plot_dualtree_Mband2D(wm)

%reconstruction
xrec = idualtreeM2D_freq(wp,wm,rm,H1,H2,HH1,HH2,G,Gh);

disp('Original image norm:');
norm(x,'fro').^2
disp('Reconstructed image norm:');
norm(xrec,'fro').^2
disp('Norm of the difference between the original image and the reconstructed one:');
norm(x-xrec,'fro').^2

figure
imagesc(x)
title('Original image')
colormap(gray(256))

figure
imagesc(xrec)
title('Reconstructed image')
colormap(gray(256))

figure
imagesc(x-xrec)
title('Original image minus reconstructed image')
colormap(gray(256))
colorbar

🎉3 参考文献

部分理论来源于网络，如有侵权请联系删除。

[1] several dual-tree M-band wavelet transforms from: Image analysis using a dual-tree M-band wavelet transform, IEEE TRANSACTIONS ON IMAGE PROCESSING, 2006, Image analysis using a dual-tree M-band wavelet transform | IEEE Journals & Magazine | IEEE Xplore
[2]a neighborhood choice from: Noise covariance properties in dual-tree wavelet decompositions, IEEE TRANSACTIONS ON INFORMATION THEORY, 2007, Noise Covariance Properties in Dual-Tree Wavelet Decompositions | IEEE Journals & Magazine | IEEE Xplore
[3]the non-linear Stein estimator: A nonlinear Stein-based estimator for multichannel image denoising, IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2008, A Nonlinear Stein-Based Estimator for Multichannel Image Denoising | IEEE Journals & Magazine | IEEE Xplore
[4]relative merits of different directional 2D wavelets are detailed in: A Panorama on Multiscale Geometric Representations, Intertwining Spatial, Directional and Frequency Selectivity, SIGNAL PROCESSING, 2011, Redirecting

🌈4 Matlab代码实现