Edge-guided Low-light Image Enhancement with Inertial Bregman Alternating Linearized Minimization

2024-03-02

arXiv.org

AI Summary

To address the challenge of extracting effective prior information in low-light image enhancement, this paper proposes an edge-guided Retinex decomposition model that explicitly incorporates edge features as structural priors into the Retinex framework for joint optimization of illumination and reflectance components. To ensure convergence in the non-convex optimization problem, we design an Inertial Bregman Alternating Linearized Minimization (IBALM) algorithm and provide theoretical proof of its convergence to a stationary point. The method synergistically integrates edge-aware priors, physically interpretable Retinex modeling, and accelerated optimization. Extensive experiments on multiple real-world low-light datasets demonstrate significant improvements over state-of-the-art methods, with consistent gains in PSNR and SSIM. Moreover, the approach achieves a favorable trade-off between detail preservation and noise suppression. Both convergence behavior and enhancement performance are thoroughly validated.

Technology Category

• Search and Optimization: Non-convex Optimization
• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Learning & Optimization for CV

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Prior-based methods for low-light image enhancement often face challenges in extracting available prior information from dim images. To overcome this limitation, we introduce a simple yet effective Retinex model with the proposed edge extraction prior. More specifically, we design an edge extraction network to capture the fine edge features from the low-light image directly. Building upon the Retinex theory, we decompose the low-light image into its illumination and reflectance components and introduce an edge-guided Retinex model for enhancing low-light images. To solve the proposed model, we propose a novel inertial Bregman alternating linearized minimization algorithm. This algorithm addresses the optimization problem associated with the edge-guided Retinex model, enabling effective enhancement of low-light images. Through rigorous theoretical analysis, we establish the convergence properties of the algorithm. Besides, we prove that the proposed algorithm converges to a stationary point of the problem through nonconvex optimization theory. Furthermore, extensive experiments are conducted on multiple real-world low-light image datasets to demonstrate the efficiency and superiority of the proposed scheme.

Problem

• Extracting prior information from dim low-light images
• Decomposing low-light images into illumination and reflectance
• Optimizing edge-guided Retinex model for image enhancement

Innovation

• Edge extraction network captures fine features
• Edge-guided Retinex model enhances low-light images
• Inertial Bregman algorithm optimizes nonconvex problem