Cross-Component Chroma Prediction With Adaptive Reference Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression methods struggle to effectively leverage cross-component redundancy between luma and chroma samples, leading to inefficiencies in predicting chroma samples, which affects compression efficiency.
Innovation Solution
A cross-component model is used to predict chroma samples based on reconstructed luma samples, where the reference area is modified based on block characteristics, and the model is derived using a combination of reference luma and chroma samples to improve prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional prediction methods are used for chroma samples, then the encoding process is simple, but the prediction accuracy is insufficient and compression efficiency is reduced
Solution Approach 1:
The patent introduces a cross-component model as an intermediary mechanism that utilizes reconstructed luma samples to predict chroma samples. This model acts as a mediator between luma and chroma components, establishing a mathematical relationship (linear model with parameters a and b) that improves prediction accuracy without requiring complex direct chroma-chroma analysis
Solution Approach 2:
The patent employs parameter-based modeling where the cross-component prediction is achieved by transforming chroma prediction into a function of luma samples with adjustable parameters (a and b). These parameters can be adapted based on block characteristics, allowing the system to optimize prediction accuracy dynamically while maintaining computational efficiency through standardized parameter update procedures
2Measurement precision
If the reference area is not modified, then the processing is faster, but the prediction accuracy is reduced due to using inappropriate reference samples
Solution Approach 1:
The patent applies local quality adaptation by modifying the reference area selection based on specific block characteristics. Instead of using a uniform reference area for all blocks, the system adapts the reference area definition to match the local properties of each block (such as its position, size, or content characteristics), ensuring that only relevant and appropriate reference samples are used for prediction
Solution Approach 2:
The reference area definition is made dynamic rather than static. The patent allows the reference area to be adjusted based on block characteristics, enabling the system to respond to different coding scenarios adaptively. This dynamic adjustment ensures optimal prediction performance for diverse block types while maintaining processing efficiency through rule-based modification rather than exhaustive search
Data Source
AI summary
A method and an apparatus for encoding and decoding a block of an image or a video wherein chroma samples of the block are intra-predicted using cross-component model and reconstructed luma samples of the block. The cross-component model is obtained for the block based on a reference area that is modified according to at least one characteristic of the block. In an embodiment, reconstructed luma and chroma samples in a reference area are selected for deriving the cross-component model based on the reconstructed luma samples of the block. In another embodiment, the reference area is modified based on an indicator signaled for the block.