Luma-Chroma Image Artifact Reduction via CNN and Luma-Aware Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Compression of images by methods like JPEG and MPEG often results in distracting artifacts such as ringing and block artifacts, which degrade image quality and viewer experience.
Innovation Solution
The method involves separating an image into luma and chroma components, using a convolutional neural network to reduce artifacts in the luma component, and applying a luma-aware filter to the chroma component based on the cleaned luma component's coefficients, followed by combining the cleaned components to form a cleaned image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If image compression is applied to reduce storage and transmission requirements, then the number of bits required is reduced, but compression artifacts are introduced that degrade image quality
Solution Approach 1:
The image is separated into luma and chroma components, allowing different processing approaches for each component. The luma component undergoes artifact reduction using a convolutional neural network, while the chroma component uses a luma-aware filter, enabling targeted artifact removal without compromising overall image quality
Solution Approach 2:
A convolutional neural network acts as an intermediary between the compressed image and the final output, learning to map compressed representations to high-quality images. The network serves as a mediator that translates compressed data into artifact-reduced images while preserving important visual information
2Object-affected harmful factors
If compression quality is increased to reduce artifacts, then image quality is improved, but the number of bits required for storage and transmission increases
Solution Approach 1:
The convolutional neural network learns to create a high-quality copy of the image from the compressed representation. Instead of storing or transmitting multiple versions at different qualities, the system learns to generate a high-quality reconstruction from the compressed data, effectively creating a virtual high-quality copy without the bit cost
Solution Approach 2:
The system changes the representation parameters by transforming the image into different color spaces (luma-chroma separation) and applying different processing parameters to each component. This allows efficient artifact reduction in the luma component while maintaining compact representation
3Object-affected harmful factors
If artifact reduction processing is applied to an image, then image quality is enhanced, but processing complexity increases
Solution Approach 1:
By segmenting the image into luma and chroma components, the system can apply simpler, more targeted processing to each component rather than applying complex general-purpose artifact reduction to the entire image. The luma component receives CNN-based processing while the chroma component uses lighter luma-aware filtering
Solution Approach 2:
The image is pre-processed by separating luma and chroma components before artifact reduction, preparing the data in a format that enables more efficient processing. This preliminary segmentation simplifies subsequent artifact reduction operations by allowing component-specific optimization
Data Source
AI summary
Methods and apparatus to reduce compression artifacts in images are disclosed herein. An example method includes separating at least a portion of an image into a first component and a second component, reducing a first artifact in the first component to form a first cleaned component, reducing, using the first cleaned component, a second artifact in the second component to form a second cleaned component, and combining the first cleaned component and the second cleaned component to form a cleaned image.