Latent Image Denoising Guided by Object Structural Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image enhancement methods often result in a loss of significant original features due to uncontrolled denoising, leading to poor enhancement effects.
Innovation Solution
An image enhancement method that involves obtaining a latent variable, adding noise, extracting object structural features, denoising with reference to these features, and performing reconstruction to retain more structural details.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If denoising is performed on the noisy latent variable without reference to object structural features, then the noise is removed, but the object structural features are lost
Solution Approach 1:
The patent applies feedback by extracting object structural features from the original image and using them as a reference guide during the denoising process. The denoising operation continuously refers back to these extracted structural features to ensure that noise removal does not eliminate important structural information, thereby preserving object structural features while improving enhancement effects.
Solution Approach 2:
The patent introduces object structural features as an intermediary element between the noisy latent variable and the denoising operation. These structural features act as a mediator that guides the denoising process, allowing the system to remove noise while maintaining structural integrity. The structural features serve as a reference that mediates the interaction between noise removal and feature preservation.
2Loss of information
If denoising intensity is increased to remove more noise, then noise removal improves, but object structural features are lost
Solution Approach 1:
The system uses feedback by continuously referencing the extracted object structural features during denoising operations. This feedback mechanism allows the system to adjust denoising intensity dynamically, ensuring that noise is removed effectively while structural features are preserved. The feedback from structural feature extraction guides the denoising process to maintain an optimal balance.
Solution Approach 2:
The patent changes the parameter of denoising intensity by introducing a reference to object structural features. Instead of applying a fixed or uniform denoising intensity, the system adjusts the denoising operation based on the structural feature reference, effectively changing the denoising parameter to preserve features while removing noise. This parameter adjustment is achieved through the relationship between the noisy latent variable and the structural feature reference.
Data Source
AI summary
The present disclosure provides an image enhancement method and apparatus, an electronic device, a computer-readable storage medium, and a computer program product, and may be applied to various scenarios such as a cloud technology, artificial intelligence, smart transportation, and assisted driving. The method includes: obtaining a latent variable of a to-be-enhanced object image, and adding noise to the latent variable to obtain a noisy latent variable of the object image, the object image being an image of a target object; extracting an object structural feature of the target object in the object image; denoising the noisy latent variable with reference to the object structural feature, to obtain a denoised latent variable of the object image; and performing image reconstruction on the denoised latent variable to obtain a first enhanced object image of the object image.