Chroma Prediction Weighting for Cross-Component Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video coding standards like H.266/VVC, the prediction accuracy of chroma values is reduced due to deviations in neighboring reference luma and chroma values, leading to inaccuracies in the linear model used for cross-component prediction.
Innovation Solution
A method for predicting video color components that considers correlations and similarities between neighboring reference and reconstructed values to determine weight coefficients, constructing a more accurate linear model for chroma prediction by weighting neighboring samples based on their deviation from the current coding block.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a linear regression method is used to construct a linear model for neighboring luma parameters and chroma parameters, then chroma prediction can be performed, but prediction accuracy is reduced when neighboring reference values are greatly deviated from current coding block parameters
Solution Approach 1:
The patent applies local quality by introducing weight coefficients that vary for different neighboring reference samples based on their correlation with the current coding block. Instead of treating all reference samples uniformly, the method assigns different weights to different samples according to their local relevance, thereby improving prediction accuracy in regions where reference values deviate significantly from current block parameters.
Solution Approach 2:
The patent changes the parameter representation by introducing weight coefficients that modify the contribution of each reference sample. The weight coefficients are calculated based on correlation measures between neighboring reference values and current block parameters, dynamically adjusting the prediction model parameters to adapt to local conditions and reduce deviation effects.
2Adaptability or versatility
If neighboring reference luma values and chroma values are considered in linear model construction, then chroma prediction is enabled, but model deviation increases when reference values greatly differ from current block parameters
Solution Approach 1:
The patent introduces dynamics by making the linear model adaptive through weight coefficients that are calculated based on correlation measures. The model transitions from a static uniform treatment of reference samples to a dynamic system where weights are adjusted according to the specific characteristics of each reference sample and its relevance to the current coding block, thereby maintaining reliability across varying conditions.
Solution Approach 2:
The patent implements feedback by calculating correlation measures between neighboring reference values and current block parameters, then using these correlation measures to determine weight coefficients. This feedback mechanism allows the model to self-adjust based on the actual relationship between reference and current values, reducing model deviation when reference values differ significantly from current block parameters.
Data Source
AI summary
A method for decoding a video includes: a luma component neighboring reference value, a chroma component neighboring reference value and a luma component reconstructed value corresponding to a current block are obtained; a chroma component predicted value corresponding to the current block is generated at least based on a linear model corresponding to a factor, where the factor is determined by at least one of a weight coefficient, the luma component neighboring reference value, or the chroma component neighboring reference value; and the video is decoded based on the chroma component predicted value. The method further includes: a subset of chroma component neighboring reference samples is determined based on position information.