Adaptive Loop Filtering with Chroma Classification for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems, such as VVC, do not effectively utilize chroma classification in their Adaptive Loop Filters (ALF), leading to suboptimal video quality, particularly in handling chroma components.
Innovation Solution
Implement chroma classification methods within the ALF framework, allowing for adaptive filtering of both luma and chroma components by deriving filtered outputs from specific sets of luma and chroma ALFs based on classification results, using techniques like geometric transformations and cross-component ALF to refine chroma components with luma channel data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If chroma classification is not applied in ALF, then device complexity is reduced, but video quality deteriorates due to suboptimal chroma component handling
Solution Approach 1:
The patent divides the chroma component processing into multiple classification blocks, where each block is independently classified and filtered using appropriate ALF parameters. This segmentation allows targeted filtering of chroma blocks based on their local characteristics, improving video quality while managing complexity through localized processing rather than global processing of the entire chroma component.
2Measurement precision
If chroma classification is implemented, then filtering precision for chroma components is improved, but computational complexity increases
Solution Approach 1:
The patent applies local quality by deriving separate ALF parameters for different chroma classification blocks based on their local characteristics. Each chroma block is classified into categories (e.g., flat, edge, textured) and assigned specific filtering parameters appropriate to its local content, thereby improving filtering precision while managing computational complexity through selective processing of only the necessary blocks.
Solution Approach 2:
The patent implements a unified chroma classification framework that can handle different chroma block types (Cb, Cr, and combined) using the same classification mechanism. The classification system serves multiple functions: identifying block characteristics, selecting appropriate filter parameters, and determining which blocks require filtering, thereby reducing overall computational complexity through a multi-functional approach.
3Manufacturing precision
If separate chroma ALF parameters are derived for each chroma block, then video quality is improved, but processing time increases
Solution Approach 1:
The patent performs chroma classification and parameter derivation in advance during the encoding process, storing the classified chroma blocks and their associated ALF parameters in buffers. During decoding, the pre-classified blocks and parameters are retrieved and applied directly, eliminating the need for real-time classification and parameter derivation, thereby significantly reducing processing time while maintaining improved chroma filtering quality.
Data Source
AI summary
Method and apparatus for video coding using chroma classification for chroma ALF. According to this method, reconstructed pixels are received, where the reconstructed pixels comprise a current colour block and the current colour block comprises a luma block and one or more chroma blocks. A filtered luma output is derived from one luma ALF selected from a set of luma ALFs according to luma block classification derived for each luma ALF classification block of the luma block. A filtered chroma output is derived from one chroma ALF selected from a set of chroma ALFs according to chroma block classification derived for each chroma ALF classification block of each of said one or more chroma blocks. Filtered-reconstructed pixels are provided, where the filtered-reconstructed pixels comprise the filtered luma output and the filtered chroma output.