Video Decoding Cross-Component Prediction Chroma Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cross-component prediction (CCP) models in video coding fail to adequately predict chroma samples due to reliance on prediction models from non-neighboring samples, leading to inefficient reconstruction of chroma blocks.
Innovation Solution
A method and device for predicting block units using a CCP merge candidate's prediction model and filtering flag, which selects and filters prediction blocks based on multiple-model cross-component prediction modes, including MM-CCCM, to refine chroma block prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If prediction models from non-neighboring samples are used for chroma block prediction, then the device complexity is reduced, but the manufacturing precision (prediction accuracy) deteriorates
Solution Approach 1:
The patent segments the prediction process into two distinct stages: first generating an initial prediction model from non-neighboring samples (reducing complexity), then refining this model using neighboring chroma samples (improving accuracy). This segmentation allows the system to benefit from both approaches without combining their drawbacks.
Solution Approach 2:
The patent performs preliminary prediction using non-neighboring samples before final refinement. The initial prediction model is established in advance, and then neighboring samples are used to refine this pre-established model, thereby improving accuracy without significantly increasing overall complexity.
2Manufacturing precision
If multiple model refinement modes are implemented for precise chroma block prediction, then the manufacturing precision is improved, but the device complexity increases
Solution Approach 1:
The patent implements dynamic model refinement where the refinement process adapts based on the specific characteristics of the chroma block and available neighboring samples. The system dynamically adjusts the refinement strength and methodology rather than applying fixed complex transformations, thereby improving precision while controlling complexity.
Solution Approach 2:
The patent changes parameters of the prediction model (such as weighting factors, refinement coefficients, and model order) based on local characteristics of the chroma block and neighboring samples. This parameter adaptation allows precise prediction without requiring complex structural changes to the overall system.
Data Source
AI summary
A method of decoding video data performed by an electronic device is provided. The method receives the video data and determines a block unit from a current frame included in the video data. The method further determines a cross-component prediction (CCP) merge list of the block unit, including several CCP merge candidates of the block unit, and selects one of the CCP merge candidates to determine a prediction model and a filtering flag of the selected CCP merge candidate. The method then predicts the block unit using the prediction model to generate a prediction block of the block unit, determines, based on the filtering flag, whether the prediction block of the block unit is further filtered to generate a predicted block of the block unit, and reconstructs the block unit based on the predicted block of the block unit.