采用拉普拉斯尺度混合先验的结构化近似消息传递算法

doi:10.3969/j.issn.0372-2112.2018.03.002

电子学报 ›› 2018, Vol. 46 ›› Issue (3): 520-528.DOI: 10.3969/j.issn.0372-2112.2018.03.002

采用拉普拉斯尺度混合先验的结构化近似消息传递算法

谢中华, 马丽红

华南理工大学电子与信息学院, 广东广州 510641

收稿日期:2016-10-17 修回日期:2017-04-07 出版日期:2018-03-25
- 通讯作者:
- 马丽红
- 作者简介:
- 谢中华,男,1985年生于广西钦州.博士研究生,研究方向为压缩感知、多维信号重建
- 基金资助:
- 国家自然科学基金 (No.61471173）

Structured Approximate Message Passing Algorithm with a Laplacian Scale Mixture Prior

XIE Zhong-hua, MA Li-hong

School of Electronic and Information Engineering, South China University of Technology, Guangzhou, Guangdong 510641, China

Received:2016-10-17 Revised:2017-04-07 Online:2018-03-25 Published:2018-03-25
- Corresponding author:
- MA Li-hong
- Supported by:
- National Natural Science Foundation of China (No.61471173)

摘要/Abstract

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.

关键词: 压缩感知, 近似消息传递, 拉普拉斯尺度混合先验, 非局部相似性, 期望最大化

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.

Key words: compressive sensing (CS), approximate message passing (AMP), Laplacian scale mixture (LSM) prior, nonlocal similarity, expectation-maximization

中图分类号:

TP391

谢中华, 马丽红. 采用拉普拉斯尺度混合先验的结构化近似消息传递算法[J]. 电子学报, 2018, 46(3): 520-528.

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.

参考文献

[1] 余晓东,雷英杰,岳韶华,等.基于粒子群优化的直觉模糊核匹配追踪算法[J].电子学报,2015,43(7):1308-1314. Yu Xiao-dong,Lei Ying-jie,Yue Shao-hua,et al.Research of PSO-based intuitionistic fuzzy kernel matching pursuit algorithm[J].Acta Electronica Sinica,2015,43(7):1308-1314.(in Chinese)
[2] Beck A,Teboulle M.A fast iterative shrinkage-thresholding algorithm for linear inverse problems[J].SIAM Journal on Imaging Sciences,2009,2(1):183-202.
[3] He L,Carin L.Exploiting structure in wavelet-based Bayesian compressive sensing[J].IEEE Trans on Signal Processing,2009,57(9):3488-3497.
[4] Donoho D L,Maleki A,Montanari A.Message passing algorithms for compressed sensing[J].Proceedings of the National Academy of Sciences,2009,106(45):18914-18919.
[5] Donoho D L,Maleki A,Montanari A.Message passing algorithms for compressed sensing:I.Motivation and construction[A].Proceedings of Information Theory Workshop[C].Cairo,Egypt:IEEE Press,2010.1-5.
[6] Rangan S.Generalized approximate message passing for estimation with random linear mixing[A].Proceedings of IEEE Int Symp Information Theory[C].St Petersburg,Russia:IEEE Press,2011.2168-2172.
[7] Vila J P,Schniter P.Expectation-maximization Gaussian-mixture approximate message passing[J].IEEE Trans on Signal Processing,2013,61(19):4658-4672.
[8] Basarab,Kouame D,Bull D R,et al.Compressive imaging using approximate message passing and a Cauchy prior in the wavelet domain[A].Proceedings of IEEE Int Conf Image Processing (ICIP)[C].Phoenix,Arizona:IEEE Press,2016.2514-2518.
[9] Donoho D L,Johnstone I,Montanari A.Accurate prediction of phase transitions in compressed sensing via a connection to minimax denoising[J].IEEE Trans on Information Theory,2013,59(6):3396-3433.
[10] Chen C,Huang J.Exploiting the wavelet structure in compressed sensing MRI[J].Magnetic Resonance Imaging,2014,32(10):1377-1389.
[11] Som S,Schniter P.Compressive imaging using approximate message passing and a Markov-tree prior[J].IEEE Trans Signal Processing,2012,60(7):3439-3448.
[12] Hua K,Hidayati S C,He F,et al.Context-aware joint dictionary learning for color image demosaicking[J].Journal of Visual Communication & Image Representation,2016,38(C):230-245.
[13] 沈燕飞,朱珍民,张勇东,等.基于秩极小化的压缩感知图像恢复算法[J].电子学报,2016,44(3):572-579. Shen Yan-fei,Zhu Zhen-min,Zhang Yong-dong,et al.Compressed sensing image reconstruction algorithm based on rank minimization[J].Acta Electronica Sinica,2016,44(3):572-579.(in Chinese)
[14] Dong W,Shi G,Li X,et al.Compressive sensing via nonlocal low-rank regularization[J].IEEE Trans Image Processing,2014,23(8):3618-3632.
[15] Metzler C,Maleki A,Baraniuk R G.BM3D-AMP:A new image recovery algorithm based on BM3D denoising[A].Proceedings of IEEE Int Conf Image Processing (ICIP)[C].Quebec City,QC:IEEE Press,2015.3116-3120.
[16] Egiazarian K,Foi A,Katkovnik V.Compressed sensing image reconstruction via recursive spatially adaptive filtering[A].Proceedings of IEEE Int Conf Image Processing (ICIP)[C].San Antonio,Texas:IEEE Press,2007.I-549-I-552.
[17] Dabov K,Foi A,Katkovnik V,et al.Image denoising by sparse 3-d transform-domain collaborative filtering[J].IEEE Trans Image Processing,2007,16(8):2080-2095.
[18] Mairal J,Bach F,Ponce J,Sapiro G.Non-local sparse models for image restoration[A].Proceedings of IEEE Int Conf Computer Vision (ICCV)[C].Kyoto,Japan:IEEE Press,2009.2272-2279.
[19] Dong W,Shi G,Ma Y.Image restoration via simultaneous sparse coding:where structured sparsity meets gaussian scale mixture[J].International Journal of Computer Vision,2015,114(2):217-232.
[20] Garrigues P,Olshausen B A.Group sparse coding with a Laplacian scale mixture prior[A].Proceedings of Advances in Neural Information Processing Systems[C].Hyatt Regency,Vancouver:IEEE Press,MIT Press,2010.676-684.
[21] Tibshirani R.Regression shrinkage and selection via the lasso[J].Journal of the Royal Statistical Society:Series B,1996,58(1):267-288.
[22] Dong W,Shi G,Li X.Nonlocal image restoration with bilateral variance estimation:a low-rank approach[J].IEEE Trans Image Processing,2013,22(2):700-711.
[23] Dempster A,Laird N M,Rubin D B.Maximum-likelihood from incomplete data via the EM algorithm[J].Journal of the Royal Statistical Society,1977,39(1):1-17.
[24] Combettes P L,Pesquet J C.A proximal decomposition method for solving convex variational inverse problems[J].Inverse Problems,2008,24(6):1-27.
[25] Wright S,Nowak R,Figueiredo M.Sparse reconstruction by separable approximation[J].IEEE Trans Signal Processing,2009,57(7):2479-2493.
[26] Gu S,Xie Q,Meng D,et al.Weighted nuclear norm minimization and its applications to low level vision[J].International Journal of Computer Vision,2017:121(2):183-208.
[27] Metzler C,Maleki A,Baraniuk R.From denoising to compressed sensing[J].IEEE Trans on Information Theory,2016,62(9):5117-5144.

采用拉普拉斯尺度混合先验的结构化近似消息传递算法

Structured Approximate Message Passing Algorithm with a Laplacian Scale Mixture Prior

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	陈胜垚, 席峰, 张素玲, 刘中. 正交压缩采样系统时域分析[J]. 电子学报, 2023, 51(3): 573-584.
[2]	吕明久, 马建朝, 韦旭, 陈文峰, 杨军, 马晓岩. 基于矩阵化原子范数的高频雷达距离-多普勒二维估计方法[J]. 电子学报, 2022, 50(5): 1150-1158.
[3]	苏林林, 陈亮, 陈菲菲, 周鑫, 焦振航, 刘钊良. 面向B5G/6G的GFDM信号高精度测距与定位研究[J]. 电子学报, 2022, 50(4): 849-859.
[4]	魏志超, 杨春玲. 时域注意力特征对齐的视频压缩感知重构网络[J]. 电子学报, 2022, 50(11): 2584-2592.
[5]	陈文俊, 杨春玲. 图像压缩感知的特征域优化及自注意力增强神经网络重构算法[J]. 电子学报, 2022, 50(11): 2629-2637.
[6]	左婷, 王法松, 张建康, 李睿. 室内可见光通信系统中基于压缩感知的空移键控信号检测方法[J]. 电子学报, 2022, 50(1): 36-44.
[7]	冯西安, 寇思玮, 谭伟杰, 毕杨. 水声信号处理中的稀疏表示理论及应用[J]. 电子学报, 2021, 49(9): 1840-1851.
[8]	裴翰奇, 杨春玲, 魏志超, 曹燕. 基于SPL迭代思想的图像压缩感知重构神经网络[J]. 电子学报, 2021, 49(6): 1195-1203.
[9]	禤韵怡, 杨春玲. 基于帧间组稀疏的两阶段递归增强视频压缩感知重构网络[J]. 电子学报, 2021, 49(3): 435-442.
[10]	李峰, 詹邦成, 辛蕾, 刘洋, 刘志佳, 肖化超. 基于新型联合感知矩阵的压缩学习目标识别技术[J]. 电子学报, 2021, 49(11): 2108-2116.
[11]	郑恩明, 陈新华, 周权斌, 李嶷, 杨鹤, 孟浩. 一种复域压缩感知目标方位估计方法[J]. 电子学报, 2021, 49(11): 2117-2123.
[12]	孙泽宇, 李传锋, 阎奔. 传感网中基于压缩感知的丢包匹配数据收集算法[J]. 电子学报, 2020, 48(4): 723-733.
[13]	陈兵兵, 曹欣远, 陈明生, 齐琦, 况晓静, 吴先良, 陈辉. 基于欠定方程的频谱峰值搜索算法及其在连续毫米波雷达系统中的应用[J]. 电子学报, 2020, 48(2): 369-374.
[14]	郑学炜, 杨春玲, 禤韵怡. CVS中基于视频运动特征的多假设-双稀疏重构算法[J]. 电子学报, 2020, 48(2): 249-257.
[15]	马彬, 王宏明, 谢显中. 基于二项分布改进的宽带压缩频谱检测方案[J]. 电子学报, 2020, 48(2): 243-248.