Video Coding Color Prediction Using Selective PDPC Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of information processing in video coding affects the fluency of video services, particularly in video coding and decoding processes, which impacts the user's viewing experience.
Innovation Solution
Implementing a Position-Dependent Prediction Combination (PDPC) application mode for predicting colour components in video coding blocks, determining residual blocks, and signaling these in a bitstream to reduce processing complexity without compromising video coding performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional video coding methods are used with comprehensive prediction mode refinement, then prediction accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent applies partial refinement action by selectively refining prediction values only for specific prediction modes (planar, DC, and angular modes with indices ≤10 or ≥58) rather than all modes. This partial application maintains prediction accuracy where it matters most while avoiding unnecessary processing complexity for other modes, directly resolving the contradiction between accuracy and complexity.
Solution Approach 2:
The patent implements local quality by applying different processing treatments to different prediction modes based on their characteristics. High-refinement modes (planar, DC, and specific angular modes) receive full PDPC refinement to maintain accuracy, while other modes use direct residual calculation to reduce complexity. This localized differentiation optimizes the balance between prediction accuracy and processing complexity for each mode type.
2Reliability
If Position-Dependent Prediction Combination (PDPC) refinement is applied to all prediction modes, then prediction performance is improved, but computational complexity increases significantly
Solution Approach 1:
The patent applies PDPC refinement selectively to only certain prediction modes (planar mode index 0, DC mode index 1, and angular modes with indices ≤10 or ≥58) rather than all 67 intra prediction modes. This partial application maintains prediction performance for modes that benefit most from refinement while significantly reducing computational complexity by avoiding refinement for the remaining modes.
Solution Approach 2:
The patent changes the parameter of refinement application from universal to conditional by introducing mode-index-based selection criteria. The refinement process is activated only when specific conditions are met (mode index ∈ {0, 1} or mode index ≤10 or mode index ≥58), transforming the computational complexity from O(N) to O(k) where k<<N is the number of refined modes.
3Manufacturing precision
If comprehensive prediction value refinement is performed, then video coding performance is improved, but processing time increases
Solution Approach 1:
The patent performs refinement partially by skipping the PDPC refinement step for prediction modes that do not meet the selection criteria. For non-refined modes, the system directly calculates residuals using the original prediction values, significantly reducing processing time while maintaining acceptable coding performance through the selective refinement approach.
Solution Approach 2:
The patent segments the set of 67 intra prediction modes into three groups: refined modes (indices 0, 1, ≤10, ≥58), non-refined modes (other indices), and excluded modes (specific angular modes 11-49). This segmentation allows parallel processing paths where refined modes undergo full PDPC processing while non-refined modes use direct residual calculation, optimizing the overall processing time-performance balance.
Data Source
AI summary
An information processing method and device, an apparatus, and a storage medium. The method includes: for inputted source video data, predicting a colour component of a coding block in the source video data according to a prediction mode to obtain a first prediction block, wherein the prediction mode is preset and is a Position-Dependent Prediction Combination (PDPC) application mode; determining a difference between the colour component of the coding block and a prediction value of the first prediction block to obtain a residual block; and signalling the residual block and the prediction mode in a bitstream.