Multi-Model Linear Chroma Prediction for Efficient Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in achieving high coding efficiency, simplifying complexity, and improving the accuracy of intra prediction, particularly in handling chroma blocks within video data.
Innovation Solution
The implementation of a multi-model linear model (MMLM) that utilizes a first and second linear model between threshold luma values and maximum/minimum luma values from reference luma and chroma samples to reconstruct chroma block values, combining these with intra prediction modes for improved accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single linear model is used for chroma prediction, then the device complexity is reduced, but the manufacturing precision (prediction accuracy) deteriorates
Solution Approach 1:
The patent divides the chroma prediction process into multiple linear models (first linear model and second linear model) that operate on different luma value ranges. This segmentation allows each model to be optimized for specific luminance conditions, improving overall prediction accuracy while keeping individual model complexity manageable.
Solution Approach 2:
The patent applies different linear model parameters (slope and intercept) to different regions of the luma-chroma relationship. By calculating separate first and second linear models with distinct parameters for different luma value ranges, the system achieves local optimization of prediction accuracy without requiring a single overly complex global model.
2Manufacturing precision
If multiple linear models are used for chroma prediction, then the manufacturing precision (prediction accuracy) is improved, but the device complexity increases
Solution Approach 1:
The patent changes the parameters (slope and intercept) of the linear models based on the luma value range. By adapting the linear model parameters to different luminance conditions through threshold-based selection, the system achieves high prediction accuracy without requiring an excessive number of models, thus controlling complexity.
Solution Approach 2:
The patent implements a dynamic selection mechanism that chooses between different linear models based on the actual luma values in the reference samples. This dynamic adaptation allows the system to optimize prediction accuracy for the current block's characteristics while maintaining a manageable set of pre-defined models.
3Manufacturing precision
If complex prediction methods are used, then the manufacturing precision (prediction accuracy) is improved, but the productivity (coding efficiency) deteriorates due to increased computational load
Solution Approach 1:
The patent segments the prediction process into distinct linear models for different luma ranges, which simplifies the computational approach compared to using a single complex non-linear model. Each linear model requires simple arithmetic operations, maintaining coding efficiency while improving accuracy through targeted prediction.
Solution Approach 2:
The patent uses parameter changes (selecting different linear model parameters based on luma thresholds) rather than complex computations to adapt to different content characteristics. This approach achieves high prediction accuracy through simple parameter selection and linear calculations, preserving coding efficiency.
Data Source
AI summary
A computing device performs a method of decoding video data by generating a multi-model linear model (MMLM) including a first linear model between the minimum luma value and the threshold luma value, and a second linear model between the threshold luma value and the maximum luma value from a group of reference luma samples and a group of reference chroma samples; and reconstructing a respective sample value of the chroma block from a weighted combination of a respective first corresponding reconstructed sample value of the luma block using the multi-model linear model, and a respective second reconstructed sample value of a neighboring chroma block from an intra prediction mode.