Motion Candidate Lookup Table Updates for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face challenges in achieving efficient compression ratios due to limitations in motion information coding, particularly in advanced motion vector prediction modes, which affect bandwidth usage in digital video transmission.
Innovation Solution
Implementing a video processing method that maintains and updates tables of motion candidates and information, using conversion techniques and additional motion information to enhance encoding and decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional motion information coding is used, then device complexity is reduced, but video compression ratio deteriorates
Solution Approach 1:
The patent pre-calculates and stores motion candidate information in lookup tables before actual video encoding. By preparing motion vectors, reference picture indices, and prediction modes in advance based on historical data and spatial-temporal correlations, the system reduces real-time coding complexity while improving compression efficiency through better motion prediction accuracy.
Solution Approach 2:
The patent implements feedback mechanisms where motion information from previously encoded blocks is continuously fed back into the lookup tables. The system uses historical motion data, spatial neighboring block information, and temporal reference picture data to dynamically update and refine motion candidates, creating a self-improving system that enhances compression ratio without proportionally increasing complexity.
2Productivity
If advanced motion vector prediction modes are implemented, then video compression ratio is improved, but bandwidth usage increases
Solution Approach 1:
The patent uses copying by storing and reusing motion candidate information from lookup tables that contain pre-computed motion vectors and prediction data. Instead of calculating motion information de novo for each block, the system copies relevant motion parameters from the lookup table based on matching criteria, reducing the amount of data that needs to be transmitted while maintaining advanced prediction capabilities.
Solution Approach 2:
The patent implements selective discarding of redundant motion information and recovery of essential prediction parameters. By identifying and discarding duplicate or less relevant motion candidates from the lookup table, and recovering only the necessary motion vectors and reference indices for current block prediction, the system reduces bandwidth requirements while preserving compression performance.
3Measurement precision
If lookup tables are continuously updated with additional motion information, then motion prediction accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies partial action by selectively updating lookup tables with only the most relevant motion information rather than continuously processing all possible candidates. The system updates motion candidates based on specific criteria such as spatial proximity, temporal relevance, and prediction accuracy thresholds, performing partial updates that maintain accuracy while reducing processing overhead compared to exhaustive table maintenance.
Data Source
AI summary
A method of video processing is provided to include maintaining one or more tables, wherein each table includes one or more motion candidates and each motion candidate is associated with corresponding motion information; performing a conversion between a current block and a bitstream representation of a video including the current block by using motion information in a table; and updating, after performing of the conversion, one or more tables based on M sets of additional motion information associated with the current block, M being an integer.