Multiple-prespecified-dictionary sparse representation for compressive sensing image reconstruction with nonconvex regularization

Li, Yunyi; Dai, Fei; Cheng, Xiefeng; Xu, Li; Gui, Guan

doi:10.1016/j.jfranklin.2018.12.013

Cited by 22 publications

(6 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Alternating optimization is applied to efficiently find the solution for model (6). The proposed model is optimized in two parts, subspace decomposition, and auto-weighted TR decomposition, which are detailed below.…”

Section: B Optimization Proceduresmentioning

confidence: 99%

“…Recently, CS has been widely applied to the compression of hyperspectral image (HSI) [3]- [5], solving the dilemma of storing and transmitting the high-resolution HSI data. Various real implementations that adhere the restricted isometry property (RIP) [6] have been developed, including dualcamera compressive hyperspectral imaging (DC-CHI) [7] and the coded aperture snapshot spectral imager (CASSI) [8].…”

Section: Introductionmentioning

confidence: 99%

“…The critical challenge of HSI-CSR lies in proposing an effective reconstruction mechanism to improve the reconstruction quality since CS is an undetermined issue. CS reconstruction essentially involves solving for the minimum l 0 sparse term, which is an NPhard problem and usually relaxed to the constrain of l 1 and l p (0 < p < 1) norm [9], [10]. Traditional model regularization methods based on sparse prior, such as sparse dictionary [11] and total variation (TV) norm [12], have been applied to HSI-CSR.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Spectral Quadratic Variation Regularized Autoweighted Tensor Ring Decomposition for Hyperspectral Image Reconstruction

Wan,

Li,

Kong

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

The structure information of hyperspectral image (HSI) is well-characterized by tensors, surpassing the capabilities of traditional compressive sensing (CS) reconstruction models based on vectors and matrices. Tensor decomposition has been integrated with other regularizations in the model-based reconstruction algorithms to capture more priors. However, the existing tensor decomposition fails to achieve the best low-rank approximation. The effectiveness of model-based reconstruction methods can be promoted. In this paper, a subspace-based model utilizing spectral quadratic variation regularized auto-weighted tensor ring (TR) decomposition (SQV-AwTR) is proposed to explore the multiple-layer spatial-spectral priors of HSI. The original HSI is decomposed into the feature image and spectral basis to explore the first-layer spectral low-rankness. To capture the second-layer low-rank prior, TR decomposition is applied to obtain the effective low-rankness approximation and low computational complexity. The tensor nuclear norm is employed to describe underlying structure priors of the TR factors, which address the deficiency in tensor rank robustness and enhance the reconstruction quality. An auto-weighted mechanism is utilized to account for the varying contributions of different TR factors to the low-rank approximation. Moreover, embedding spectral quadratic variation into subspace decomposition enhances spectral smoothness and continuity. Alternating minimization is used to optimize the spectral basis and feature HSI. Through comparative experiments on three datasets, the superiority of the proposed model is demonstrated.

show abstract

Section: B Optimization Proceduresmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Spectral Quadratic Variation Regularized Autoweighted Tensor Ring Decomposition for Hyperspectral Image Reconstruction

Wan,

Li,

Kong

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…One of the main technical challenges for CS is how to reduce the measurements, meanwhile, to obtain high-quality images. Typical applications of CS include radar imaging [2], channel estimation in communications systems [3][4][5][6], sparse recovery [7] and signal detection [8][9][10][11][12], electrocardiogram signal reconstruction [13], magnetic resonant imaging (MRI) [14][15][16], and especially in image processing [17][18][19].…”

Section: Introductionmentioning

confidence: 99%

Nonconvex Nonsmooth Low-Rank Minimization for Peneralized Image Compressed Sensing via Group Sparse Representation

Li¹,

Liu²,

Yu³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Group sparse representation (GSR) based method has led to great successes in various image recovery tasks, which can be converted into a low-rank matrix minimization problem. As a widely used surrogate function of low-rank, the nuclear norm based convex surrogate usually leads to over-shrinking problem, since the standard soft-thresholding operator shrinks all singular values equally. To improve traditional sparse representation based image compressive sensing (CS) performance, we propose a generalized CS framework based on GSR model, which leads to a nonconvex nonsmooth low-rank minimization problem. The popular 2 -norm and M-estimator are employed for standard image CS and robust CS problem to fit the data respectively. For the better approximation of the rank of group-matrix, a family of nuclear norms are employed to address the over-shrinking problem. Moreover, we also propose a flexible and effective iteratively-weighting strategy to control the weighting and contribution of each singular value. Then we develop an iteratively reweighted nuclear norm algorithm for our generalized framework via an alternating direction method of multipliers framework, namely, GSR-AIR. Experimental results demonstrate that our proposed CS framework can achieve favorable reconstruction performance compared with current state-of-the-art methods and the robust CS framework can suppress the outliers effectively.

show abstract

“…One of the main technical challenges for CS is how to reduce the measurements whereas obtain high-quality images. Typical applications of CS include radar imaging [2], channel estimation in communications systems [3][4][5][6], sparse recovery [7] and signal detection [8][9][10][11][12], electrocardiogram signal reconstruction [13], magnetic resonant imaging (MRI) [14][15][16], and especially in image processing [17][18][19].…”

Section: Introductionmentioning

confidence: 99%

Nonconvex Nonsmooth Low-Rank Minimization for Peneralized Image Compressed Sensing via Group Sparse Representation

Li¹,

Liu²,

Yu³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Group sparse representation (GSR) based method has led to great successes in various image recovery tasks, which can be converted into a low-rank matrix minimization problem. As a widely used surrogate function of low-rank, the nuclear norm based convex surrogate usually leads to over-shrinking problem, since the standard soft-thresholding operator shrinks all singular values equally. To improve traditional sparse representation based image compressive sensing (CS) performance, we propose a generalized CS framework based on GSR model, leading to a nonconvex nonsmooth low-rank minimization problem. The popular -norm and M-estimator are employed for standard image CS and robust CS problem to fit the data respectively. For the better approximation of the rank of group-matrix, a family of nuclear norms are employed to address the over-shrinking problem. Moreover, we also propose a flexible and effective iteratively-weighting strategy to control the weighting and contribution of each singular value. Then we develop an iteratively reweighted nuclear norm algorithm for our generalized framework via an alternating direction method of multipliers framework, namely, GSR-ADMM-IRNN. Experimental results demonstrate that our proposed CS framework can achieve favorable reconstruction performance compared with current state-of-the-art methods and the RCS framework can suppress the outliers effectively.

show abstract

Multiple-prespecified-dictionary sparse representation for compressive sensing image reconstruction with nonconvex regularization

Cited by 22 publications

References 27 publications

Spectral Quadratic Variation Regularized Autoweighted Tensor Ring Decomposition for Hyperspectral Image Reconstruction

Spectral Quadratic Variation Regularized Autoweighted Tensor Ring Decomposition for Hyperspectral Image Reconstruction

Nonconvex Nonsmooth Low-Rank Minimization for Peneralized Image Compressed Sensing via Group Sparse Representation

Nonconvex Nonsmooth Low-Rank Minimization for Peneralized Image Compressed Sensing via Group Sparse Representation

Contact Info

Product

Resources

About