Abstract:The hyperspectral image superresolution (HSI-SR) problem aims to improve the spatial quality of a low spatial resolution HSI by fusing the LR-HSI with the corresponding high spatial resolution multispectral image. The generated HSI with high spatial quality, i.e., the target high spatial resolution hyperspectral image (HR-HSI), generally has some fundamental latent properties, e.g., the sparsity, and the piecewise smoothness along with the three modes (i.e., width, height, and spectral mode). However, limited works consider both properties in the HSI-SR problem. In this work, a novel unidirectional total variation (TV) based approach is been proposed. On the one hand, we consider that the target HR-HSI exhibits both the sparsity and the piecewise smoothness on the three modes, and they can be depicted well by the ℓ 1 -norm and TV, respectively. On the other hand, we utilize the classical Tucker decomposition to decompose the target HR-HSI (a three-mode tensor) as a sparse core tensor multiplied by the dictionary matrices along with the three modes. Especially, we impose the ℓ 1 -norm on core tensor to characterize the sparsity and the unidirectional TV on three dictionaries to characterize the piecewise smoothness. The proximal alternating optimization scheme and the alternating direction method of multipliers are used to iteratively solve the proposed model. Experiments on three common datasets illustrate that the proposed approach has better performance than some current state-of-the-art HSI-SR methods. Please find source code from: https://liangjiandeng.github.io/.

Low-rank total variation for image super-resolution.

Remote Sensing Image Reconstruction Method Based on Non-Local Similarity and Low Rank Matrix

A Novel Multi-Frame Super Resolution Method Based On Low Rank Matrix Completion And Bilateral Total Variation Regularization

Single Image Super-Resolution Using Combined Total Variation Regularization by Split Bregman Iteration

Single Image Super-Resolution Based on Nonlocal Sparse and Low-Rank Regularization.

A Single-Image Super-Resolution Method Via Low-Rank Matrix Recovery And Nonlinear Mappings

Adaptive total variation and second-order total variation-based model for low-rank tensor completion

Compressive total variation for image reconstruction and restoration

Video super-resolution using low rank matrix completion

Nonlocal Low-Rank and Total Variation Constrained PET Image Reconstruction

Single Image Super-Resolution Via Locally Regularized Anchored Neighborhood Regression and Nonlocal Means.

Single-image super-resolution using directional total variation regularization and alternating direction method of multiplier solver

Robust Super-Resolution Reconstruction Algorithm Based on Total Variation

Sparse Representation based Super-Resolution of MRI Images with Non-Local Total Variation Regularization

A Total Fractional-Order Variation Model for Image Super-Resolution and Its SAV Algorithm

Low-rank Sparse Representation for Single Image Super-Resolution Via Self-Similarity Learning

Regularized Super-resolution Reconstruction Based on Bilateral Total Variation Model with Adaptive Constraints

Low-Rank Tensor Completion with Total Variation for Visual Data Inpainting

Self-updating NonLocal Total Variation for highly undersampled variable density spiral reconstruction

A Super-Resolution Imaging Method for Forward-Looking Scanning Radar Based on Improved Total Variation

Hyperspectral Image Superresolution Using Unidirectional Total Variation with Tucker Decomposition