Nighttime Hazy Image Enhancement via Progressively and Mutually Reinforcing Night-Haze Priors

2026-01-05

arXiv.org

AI Summary

Nighttime hazy images suffer from simultaneous degradation due to low illumination and atmospheric haze, making it challenging for existing methods to effectively restore overall visibility. This work proposes a progressive prior-enhancement framework that, for the first time, explicitly models the mutual reinforcement mechanism between low-light and haze priors. By introducing a frequency-domain-aware routing strategy, the method synergistically integrates visual and spectral information across image, patch, and pixel levels, enabling multi-scale, multi-domain joint restoration. The proposed approach achieves state-of-the-art quantitative and qualitative performance on standard nighttime dehazing benchmarks and demonstrates strong generalization capabilities on daytime dehazing and low-light enhancement tasks.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Security and Privacy: Privacy-enhancing technologies

Abstract

Enhancing the visibility of nighttime hazy images is challenging due to the complex degradation distributions. Existing methods mainly address a single type of degradation (e.g., haze or low-light) at a time, ignoring the interplay of different degradation types and resulting in limited visibility improvement. We observe that the domain knowledge shared between low-light and haze priors can be reinforced mutually for better visibility. Based on this key insight, in this paper, we propose a novel framework that enhances visibility in nighttime hazy images by reinforcing the intrinsic consistency between haze and low-light priors mutually and progressively. In particular, our model utilizes image-, patch-, and pixel-level experts that operate across visual and frequency domains to recover global scene structure, regional patterns, and fine-grained details progressively. A frequency-aware router is further introduced to adaptively guide the contribution of each expert, ensuring robust image restoration. Extensive experiments demonstrate the superior performance of our model on nighttime dehazing benchmarks both quantitatively and qualitatively. Moreover, we showcase the generalizability of our model in daytime dehazing and low-light enhancement tasks.

Problem

• nighttime hazy image
• image enhancement
• low-light
• haze removal
• degradation

Innovation

• mutual reinforcement
• nighttime dehazing
• low-light enhancement
• frequency-aware routing
• multi-level experts