Unified Image Restoration Model for Low-Light Super-Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing super-resolution models are limited in effectively enhancing low-light images, particularly in real low-light scenes, as they are primarily designed for images with sufficient illumination without visual enhancement.
Innovation Solution
An image restoration method utilizing a pre-trained model with an optical feature extraction network and an image feature extraction network, which extracts illumination and target image features to generate a bright image, incorporating an illumination code matrix and channel coefficients for improved low-light image processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a pre-trained super-resolution model is used for low-light images, then the model structure is simple and easy to implement, but the image quality is poor and visual effect is not improved
Solution Approach 1:
The patent merges the low-light enhancement function and super-resolution function into a single unified model. The model simultaneously performs illumination enhancement and super-resolution processing through integrated feature extraction networks and feature fusion modules, eliminating the need for separate processing stages while improving overall image quality
Solution Approach 2:
The unified model is segmented into distinct functional components: an optical feature extraction network for illumination enhancement, an image feature extraction network for super-resolution, and a feature fusion module. This segmentation allows each component to specialize in specific tasks while working together to solve the overall low-light super-resolution problem
2Manufacturing precision
If separate low-light enhancement and super-resolution models are used, then the image quality is improved, but the processing efficiency is reduced and the system becomes more complex
Solution Approach 1:
The patent combines two separate processing models into one unified model that performs both low-light enhancement and super-resolution simultaneously. This integration maintains the quality improvements of separate models while achieving faster processing through a single-pass architecture with shared feature extraction pathways
Solution Approach 2:
The unified model achieves multi-functionality by incorporating both low-light enhancement capabilities and super-resolution capabilities within a single architecture. The model can process low-light images to simultaneously improve illumination and resolution, making it universally applicable to low-light super-resolution tasks without requiring multiple specialized models
Data Source
AI summary
An image restoration method includes: acquiring an original low-light image to be restored; acquiring a pre-trained image processing model; and inputting the original low-light image into the image processing model, so that an optical feature extraction network in the image processing model extracts an illumination feature from the original low-light image, and an image feature extraction network extracts a target image feature from the original low-light image, and generating a target bright image based on the illumination feature and the target image feature. The optical feature extraction network and the image feature extraction network are respectively used to process an image feature of an original low-light image, so as to obtain an illumination feature and a target image feature, and the illumination feature is then fused with the target image feature for image restoration to obtain a bright image.