Poisson Informed Retinex Network for Extreme Low-Light Image Enhancement

TL;DR

This paper presents a lightweight deep learning model that effectively enhances extremely low-light images by integrating Poisson noise denoising with Retinex-based illumination correction, without requiring prior reflectance or illumination information.

Contribution

It introduces a novel unified encoder-decoder network that handles Poisson noise and illumination enhancement simultaneously, improving low-light image quality without color distortion.

Findings

Significantly improves visibility and brightness in low-light images.

Preserves image structure and color constancy.

Demonstrates effectiveness and practicality through experiments.

Abstract

Low-light image denoising and enhancement are challenging, especially when traditional noise assumptions, such as Gaussian noise, do not hold in majority. In many real-world scenarios, such as low-light imaging, noise is signal-dependent and is better represented as Poisson noise. In this work, we address the problem of denoising images degraded by Poisson noise under extreme low-light conditions. We introduce a light-weight deep learning-based method that integrates Retinex based decomposition with Poisson denoising into a unified encoder-decoder network. The model simultaneously enhances illumination and suppresses noise by incorporating a Poisson denoising loss to address signal-dependent noise. Without prior requirement for reflectance and illumination, the network learns an effective decomposition process while ensuring consistent reflectance and smooth illumination without causing any form of color distortion. The experimental results demonstrate the effectiveness and practicality of the proposed low-light illumination enhancement method. Our method significantly improves visibility and brightness in low-light conditions, while preserving image structure and color constancy under ambient illumination.