Customizable Restoration in Multi-Degradation Scenarios: Joint Deraining and Low-Light Image Enhancement via Perceptual Decoupling

In real-world nighttime photography, images often suffer from both low-light conditions and rain-induced degradations simultaneously. Traditional methods typically address these issues separately, which limits their effectiveness when both types of degradation are present. To address this challenge, we propose the Perceptual Decoupling-based Learning Network (PDLN), designed to separate these degradations by leveraging their distinct physical properties. PDLN models rain and low-light priors and integrates them through the Decoupled Perception Transformer (DPformer), enabling a robust representation of these degradations. This approach harnesses interactions in both the spatial and frequency domains, resulting in state-of-the-art performance and allowing for personalized enhancement styles. The code and dataset will be available after receipt in <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/lajiaolajiaolajiao/PDLN</uri>.