Unified Low-Light Traffic Image Enhancement via Multi-Stage Illumination Recovery and Adaptive Noise Suppression

TL;DR

This paper introduces an unsupervised multi-stage deep learning framework that enhances low-light traffic images by decomposing images into illumination and reflectance, refining brightness, suppressing noise, and compensating over-exposed regions to improve visibility and perception tasks.

Contribution

It presents a novel unsupervised multi-stage network with specialized modules for illumination adaptation, reflectance restoration, and over-exposure compensation, eliminating the need for paired ground-truth images.

Findings

Outperforms state-of-the-art methods in quantitative metrics

Improves visibility and structural detail in low-light traffic images

Enhances downstream perception tasks in real-world scenarios

Abstract

Enhancing low-light traffic images is crucial for reliable perception in autonomous driving, intelligent transportation, and urban surveillance systems. Nighttime and dimly lit traffic scenes often suffer from poor visibility due to low illumination, noise, motion blur, non-uniform lighting, and glare from vehicle headlights or street lamps, which hinder tasks such as object detection and scene understanding. To address these challenges, we propose a fully unsupervised multi-stage deep learning framework for low-light traffic image enhancement. The model decomposes images into illumination and reflectance components, progressively refined by three specialized modules: (1) Illumination Adaptation, for global and local brightness correction; (2) Reflectance Restoration, for noise suppression and structural detail recovery using spatial-channel attention; and (3) Over-Exposure Compensation, for reconstructing saturated regions and balancing scene luminance. The network is trained using self-supervised reconstruction, reflectance smoothness, perceptual consistency, and domain-aware regularization losses, eliminating the need for paired ground-truth images. Experiments on general and traffic-specific datasets demonstrate superior performance over state-of-the-art methods in both quantitative metrics (PSNR, SSIM, LPIPS, NIQE) and qualitative visual quality. Our approach enhances visibility, preserves structure, and improves downstream perception reliability in real-world low-light traffic scenarios.