Intra Prediction Mode Selection for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in reducing the complexity of mode selection during intra-prediction, particularly in real-time or near-real-time encoding, due to the computational intensity of calculating approximate costs for each of the 35 intra-prediction modes, which exceeds design constraints in some environments.
Innovation Solution
A method that initially selects a subset of pre-defined intra-prediction modes statistically more likely to provide low distortion, and then refines this subset by searching for additional modes with lower approximate costs, reducing the number of full rate distortion analyses performed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If approximate cost calculations are performed for all 35 intra-prediction modes, then coding precision is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments the set of 35 intra-prediction modes into multiple subsets and processes them in stages. First, modes are divided into groups based on their probability of being optimal. Then, approximate cost calculations are performed selectively on subsets rather than all modes simultaneously, reducing computational burden while maintaining precision through systematic evaluation of likely candidates.
Solution Approach 2:
The patent performs preliminary actions by pre-calculating and storing probability information for different intra-prediction modes based on statistical analysis of video data. This preliminary probability assessment allows the encoder to prioritize modes that are more likely to be optimal, performing full rate-distortion analysis only on these pre-identified candidates rather than all 35 modes, thus reducing computational complexity while maintaining coding precision.
2Productivity
If full rate distortion analysis is performed for all intra-prediction modes, then coding efficiency is improved, but processing time increases beyond real-time constraints
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different intra-prediction modes based on their individual characteristics and probabilities. Instead of uniformly applying full rate-distortion analysis to all modes, the method applies stricter evaluation criteria to modes with higher probability of optimality and more relaxed criteria for less likely modes, optimizing processing time while maintaining coding efficiency through targeted analysis.
Solution Approach 2:
The patent employs partial action by performing full rate-distortion analysis only on a subset of intra-prediction modes rather than all 35 modes. The method identifies and analyzes only the most probable modes that are likely to yield optimal coding efficiency, accepting that some modes are not fully evaluated but this partial analysis is sufficient to achieve real-time processing constraints while maintaining acceptable coding performance.
3Speed
If the number of intra-prediction modes to evaluate is reduced, then processing speed is improved, but coding precision may deteriorate
Solution Approach 1:
The patent implements feedback mechanisms by continuously monitoring the performance of different intra-prediction modes and adjusting the evaluation strategy accordingly. Statistical information from previously processed blocks is fed back to inform the probability assessments for current blocks, allowing the system to refine which modes warrant full analysis based on actual performance data, thus maintaining coding precision while reducing the number of modes requiring full evaluation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In general, techniques are described for reducing the complexity of mode selection when selecting from multiple, different prediction modes. A video coding device comprising a processor may perform the techniques. The processor may compute approximate costs for a pre-defined set of intra-prediction modes identified in a current set. The current set of intra-prediction modes may include fewer modes than a total number of intra-prediction modes. The processor may compare approximate costs computed for one or more most probable intra-prediction modes to a threshold and replace one or more of the intra-prediction modes of the current set with one or more most probable intra-prediction modes. The processor may perform rate distortion analysis with respect to each intra-prediction mode identified in the current set and perform intra-prediction coding with respect to the current block using a mode of the current set.