Intra Prediction Using Cross-Component Linear Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding schemes face challenges in efficiently generating intra prediction images for chrominance, particularly in terms of coding efficiency and prediction accuracy, while maintaining manageable computational complexity.
Innovation Solution
An intra prediction image generation apparatus that derives prediction parameters for luminance blocks and uses sub-block coordinates to generate intra prediction images for chrominance blocks, incorporating methods such as MPM candidate list derivation, linear prediction with weight configuration, and adaptive reference pixel selection to improve prediction accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional intra prediction methods are used for chrominance, then computational complexity is reduced, but coding efficiency and prediction accuracy deteriorate
Solution Approach 1:
The chrominance block is divided into multiple sub-blocks, and prediction parameters are derived for each sub-block independently using sub-block coordinates. This segmentation allows for more precise local prediction while maintaining manageable computational complexity through localized processing.
Solution Approach 2:
Different prediction parameters are applied to different sub-blocks of the chrominance block based on local luminance characteristics. This local quality approach improves prediction accuracy by adapting to local variations in the image content rather than applying a single global prediction.
2Productivity
If simple prediction methods are used for chrominance, then processing speed is maintained, but coding efficiency deteriorates
Solution Approach 1:
Luminance prediction parameters are derived first and stored, then chrominance prediction parameters are derived based on these pre-computed luminance parameters. This preliminary action of computing luminance parameters first enables more efficient chrominance prediction without redundant computations.
Solution Approach 2:
Luminance blocks serve as an intermediary to derive chrominance prediction parameters. The luminance-chrominance relationship is exploited where luminance prediction results act as a mediator to generate accurate chrominance predictions, improving coding efficiency through cross-component prediction.
Data Source
AI summary
An intra prediction image decoding apparatus is provided with a prediction parameter derivation circuitry that derives prediction parameters of multiple luminance blocks. A Cross-component Linear Model (CCLM) is used to specify the model and the prediction parameter is obtained, and thus an intra prediction image with reference to the prediction parameters is generated.