Structured Approximate Message Passing Algorithm with a Laplacian Scale Mixture Prior

XIE Zhong-hua; MA Li-hong

doi:10.3969/j.issn.0372-2112.2018.03.002

您当前的位置：

首页 >

文章列表页 >

Structured Approximate Message Passing Algorithm with a Laplacian Scale Mixture Prior

更新时间：2025-07-16

- Structured Approximate Message Passing Algorithm with a Laplacian Scale Mixture Prior
- Acta Electronica Sinica Vol. 46, Issue 3, Pages: 520-528(2018)
- 作者机构：
  
  华南理工大学电子与信息学院,广东,广州,510641
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China (No.61471173)
- DOI：10.3969/j.issn.0372-2112.2018.03.002
  CLC： TP391
- Published Online：25 March 2018，
  
  Published：2018
- 稿件说明：
移动端阅览
XIE Zhong-hua, MA Li-hong. Structured Approximate Message Passing Algorithm with a Laplacian Scale Mixture Prior[J]. Acta Electronica Sinica, 2018, 46(3): 520-528.
DOI：

XIE Zhong-hua, MA Li-hong. Structured Approximate Message Passing Algorithm with a Laplacian Scale Mixture Prior[J]. Acta Electronica Sinica, 2018, 46(3): 520-528. DOI： 10.3969/j.issn.0372-2112.2018.03.002.

摘要

为了准确有效地实现自然图像的压缩感知重构，提出一种使用拉普拉斯尺度混合（Laplacian Scale Mixture，LSM）先验的结构化近似消息传递（Approximate Message Passing，AMP）算法.利用LSM模型构建AMP算法的高阶统计约束，将压缩感知重构问题转化为先验信息估计问题和奇异值最小化问题.首先，用LSM分布刻画相似块矩阵奇异值的稀疏性，其中该稀疏性指示了图像块的相似性，因此LSM模型被用来描述图像的非局部相似结构；然后，通过期望最大化算法估计LSM模型的尺度参数，得到可靠的先验信息；最后，由AMP算法求解奇异值最小化问题，实现图像的精确重构.实验结果表明，提出的结构化AMP算法的图像重构质量优于多种主流的压缩感知图像重构算法.

Abstract

In order to reconstruct natural images from compressive sensing (CS) measurements accurately and effectively

a novel structured approximate message passing algorithm using a Laplacian scale mixture (LSM) prior is proposed. The higher-order statistical constraint of the AMP algorithm is created by the LSM model

turning the CS recovery problem into a prior information estimation problem and a singular value minimization problem. Firstly

we use the LSM distribution to model the sparsity of the singular values of the matrices built by similar patches

which denotes the similarity of image patches

and thus utilize the LSM model to describe the nonlocal similarity of images. Secondly

to obtain reliable prior information

the scale parameters of the LSM model are estimated through the use of the expectation-maximization (EM) algorithm. Finally

the singular value minimization problem is solved by the AMP algorithm to achieve the accurate image reconstruction. Experimental results show that the reconstruction quality of our structured AMP algorithm is superior to the state of art CS reconstruction algorithms.

关键词

Keywords

references

Views

358

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

A Nonlocal Sparse Representation Method for Color Demosaicking

Research on Recovery of Clipping and HPA Nonlinear Distortion Based on Compressive Sensing in OFDM Systems

Mutual Support Dual Watermark Algorithm Based on Compressive Sensing

Construction of Measurement Matrix in Compressive Sensing Based on Composite Chaotic Mapping

Research on Structural Similarity Based Inter-Frame Group Sparse Representation for Compressed Video Sensing

Related Author

HUANG Li-li

XIAO Liang

WEI Zhi-hui

YANG Lin

SONG Kun

ZHAO Chun-hui

LIU Wei

ZHOU Wei

Related Institution

School of Science, Guangxi University of Science and Technology

School of Computer Science and Technology, Nanjing University of Science and Technology

School of Science Guangxi University of Science and Technology Liuzhou Guangxi China

School of Computer Science and Technology Nanjing University of Science and Technology Nanjing Jiangsu China

Key Laboratory of National Communication Technology, University of Electronic Science and Technology

⁰