Iterative Low-rank Network for Hyperspectral Image Denoising

2025-08-30

AI Summary

Hyperspectral image (HSI) denoising faces dual challenges: insufficient exploitation of low-rank priors and difficulty in preserving fine spatial-spectral details. To address these, we propose a model-driven and data-driven hybrid iterative low-rank network. Built upon a U-Net backbone, it incorporates a learnable Rank Minimization Module (RMM) that applies adaptive Singular Value Thresholding (SVT) to low-frequency wavelet-domain features, coupled with an iterative residual fusion mechanism for dynamic correction of intermediate reconstructions. The RMM models spectral low-rankness via differentiable wavelet transforms and parameterized thresholds, enabling end-to-end optimization. Extensive experiments on both synthetic and real-world noisy HSIs demonstrate state-of-the-art performance: our method significantly outperforms existing approaches in quantitative metrics and visual quality, achieving superior noise suppression while faithfully recovering both spatial structures and spectral signatures.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Search and Retrieval-Augmented AI: Web search models and ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Web Mining and Content Analysis: Web data integration and cleaning

Abstract

Hyperspectral image (HSI) denoising is a crucial preprocessing step for subsequent tasks. The clean HSI usually reside in a low-dimensional subspace, which can be captured by low-rank and sparse representation, known as the physical prior of HSI. It is generally challenging to adequately use such physical properties for effective denoising while preserving image details. This paper introduces a novel iterative low-rank network (ILRNet) to address these challenges. ILRNet integrates the strengths of model-driven and data-driven approaches by embedding a rank minimization module (RMM) within a U-Net architecture. This module transforms feature maps into the wavelet domain and applies singular value thresholding (SVT) to the low-frequency components during the forward pass, leveraging the spectral low-rankness of HSIs in the feature domain. The parameter, closely related to the hyperparameter of the singular vector thresholding algorithm, is adaptively learned from the data, allowing for flexible and effective capture of low-rankness across different scenarios. Additionally, ILRNet features an iterative refinement process that adaptively combines intermediate denoised HSIs with noisy inputs. This manner ensures progressive enhancement and superior preservation of image details. Experimental results demonstrate that ILRNet achieves state-of-the-art performance in both synthetic and real-world noise removal tasks.

Problem

• Addressing hyperspectral image denoising with physical priors
• Integrating model-driven and data-driven denoising approaches effectively
• Preserving image details while removing synthetic and real noise

Innovation

• Iterative low-rank network integrating model and data
• Wavelet domain singular value thresholding for low-frequency
• Adaptive parameter learning from data for refinement