Litcius/Paper detail

Hyperspectral Image Denoising via $L_{0}$ Regularized Low-Rank Tucker Decomposition

Xin Tian, Kun Xie, Hanling Zhang

2023IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing14 citationsDOIOpen Access PDF

Abstract

This paper studies the mixed noise removal problem for hyperspectral images (HSIs), which often suffer from Gaussian noise and sparse noise. Conventional denoising models mainly employ the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{1}$</tex-math></inline-formula> -norm-based regularizers to remove sparse noise and ensure piecewise smoothness. However, the denoising performance is poor for highly structured images with severe noise since the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{1}$</tex-math></inline-formula> -norm over-penalizes large entries. To tackle this limitation, we propose a denoising model that combines tensor decomposition with two kinds of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{0}$</tex-math></inline-formula> -norm-based regularizers. Firstly, we use low-rank Tucker decomposition with the Stiefel manifold to characterize the global correlation of HSIs. Then, we utilize the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{0}$</tex-math></inline-formula> -norm to leverage the intrinsic sparsity information of the corruption domain, thereby enhancing the effectiveness of sparse noise removal. Simultaneously, we introduce a weighted <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{0}$</tex-math></inline-formula> -norm regularizer on the gradient of each pixel to promote the local spectral-spatial smoothness. To solve the proposed model, we design a hard-thresholding-based alternating direction method of multipliers (ADMM) algorithm. Instead of spending time to find a proper rank in advance, we adopt a rank-increasing scheme to dynamically adjust the tensor rank during the optimization procedure. In this way, our algorithm avoids the rank selection burden and improves computational efficiency. Finally, we test the proposed method on both synthetic and real datasets. Numerical results demonstrate its superiority, especially, the improvements of our method over the best-compared results up to 2.07dB for mixed noise removal.

Topics & Concepts

Hyperspectral imagingDecompositionNoise reductionTucker decompositionImage denoisingArtificial intelligenceComputer sciencePattern recognition (psychology)Rank (graph theory)Image (mathematics)Computer visionMathematicsTensor decompositionCombinatoricsTensor (intrinsic definition)Pure mathematicsEcologyBiologyImage and Signal Denoising MethodsSparse and Compressive Sensing TechniquesTensor decomposition and applications
Hyperspectral Image Denoising via $L_{0}$ Regularized Low-Rank Tucker Decomposition | Litcius