Chroma Intra Prediction Blending Without Linear Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video encoding and decoding, when linear-model based chroma prediction is disabled due to low latency or low power requirements, existing methods cannot effectively leverage chroma fusion, leading to reduced coding efficiency.
Innovation Solution
Derive a first predictor based on a non-linear-model based mode and a second predictor based on another non-linear-model based mode, and determine a final predictor by blending these two predictors, even if linear-model based mode is disabled.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If linear-model based chroma prediction is disabled to reduce latency or power consumption, then coding efficiency deteriorates
Solution Approach 1:
The patent merges multiple non-linear prediction modes (first non-linear-model based mode and second non-linear-model based mode) and blends their predictors to create a final predictor. This combining approach allows the system to achieve chroma fusion benefits without requiring linear-model based mode, thus maintaining coding efficiency while satisfying low latency or low power constraints.
Solution Approach 2:
The patent changes the prediction model parameters from linear-based to non-linear-based modes. By deriving predictors from non-linear-model based modes instead of linear-model based modes, the system adapts to constrained environments (low latency/power) while preserving prediction effectiveness through the blending mechanism.
2Use of energy by moving object
If linear-model based chroma prediction is disabled to reduce power consumption, then coding efficiency deteriorates
Solution Approach 1:
The patent combines multiple non-linear prediction approaches and blends their results to achieve effective chroma prediction. This merging of non-linear modes provides an energy-efficient alternative to linear-model based prediction, maintaining coding efficiency while reducing power consumption as required by low-power applications.
Solution Approach 2:
The patent changes the prediction methodology from linear-model based to non-linear-model based parameters. This parameter change enables the system to operate in low-power modes while preserving prediction accuracy through the blended non-linear predictors, thus resolving the contradiction between power consumption and coding efficiency.
3Device complexity
If chroma fusion is disabled to simplify processing, then coding efficiency deteriorates
Solution Approach 1:
The patent merges multiple non-linear prediction modes and blends their predictors to achieve chroma fusion effects. This approach provides a simplified implementation of chroma fusion that does not require complex linear-model based processing, thus reducing processing complexity while maintaining or improving coding efficiency.
Solution Approach 2:
The patent uses copying and blending of predictors from different non-linear modes to create the final prediction. By copying predictors from multiple non-linear-model based modes and blending them, the system achieves chroma fusion functionality with simpler processing requirements compared to traditional linear-model based approaches.
Data Source
AI summary
A method of intra predicting a chroma block of a video picture includes: deriving a first predictor of the chroma block based on a first non-linear-model based mode; in response to deriving a chroma predictor based on a linear-model based mode being disabled, deriving a second predictor of the chroma block based on a second non-linear-model based mod; and determining a final predictor of the chroma block is determined by blending the first and second predictors.