Image Enhancement Using Structural-Guided Latent Denoising
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
1Manufacturing precision
If noise is added to latent variable and then removed through denoising process, then image enhancement is achieved, but object structural features are lost due to uncontrolled denoising intensity
Solution Approach 1:
The patent applies feedback control by using extracted object structural features to guide and adjust the denoising process. The structural features serve as feedback signals that continuously monitor and regulate denoising intensity, ensuring that original structural information is preserved while achieving effective noise removal and image enhancement.
Solution Approach 2:
The patent performs preliminary extraction of object structural features from the input image before the denoising process begins. This preliminary action allows the structural features to be used as reference guidance throughout the subsequent denoising and generation steps, preventing structural information loss before it occurs.
2Object-generated harmful factors
If denoising intensity is increased to remove more noise, then noise removal effectiveness improves, but more original features are lost
Solution Approach 1:
The extracted structural features provide real-time feedback during the denoising process, allowing the system to dynamically adjust denoising intensity. When structural features indicate important regions, the denoising intensity is automatically reduced to preserve these features, while allowing stronger denoising in regions with less critical information.
Solution Approach 2:
The patent applies different denoising intensities to different regions of the image based on local structural importance. Regions with significant structural features receive milder denoising treatment, while regions with less critical information undergo more aggressive denoising, achieving local optimization of the noise-removal vs. feature-preservation trade-off.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
The present application provides an image enhancement method and apparatus, an electronic device, a computer-readable storage medium, and a computer program product, which can be applied to various scenarios such as cloud technology, artificial intelligence, intelligent traffic and assisted driving. The method comprises: acquiring a latent variable of an object image to be enhanced, 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 object structure features of the target object in the object image; on the basis of the object structure features, denoising the noisy latent variable to obtain a denoised latent variable of the object image; and on the basis of the denoised latent variable, performing image reconstruction to obtain a first object enhanced image of the object image.