Intra Mode OBMC Blending for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video processing technologies, such as those in the HEVC and VVC standards, face limitations in performing overlapped block motion compensation (OBMC) for intra mode blocks, which can lead to suboptimal coding efficiency and increased bandwidth usage due to the inability to apply OBMC on blocks predicted using intra modes.
Innovation Solution
The method involves performing OBMC on blocks predicted with an intra mode by obtaining prediction signals based on both the current block vector and neighboring block vectors, blending these signals to generate a final prediction, and applying this process at a sub-block level for boundaries, thereby extending OBMC capabilities to intra mode blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If OBMC is not applied to intra mode blocks, then device complexity is reduced, but coding efficiency deteriorates and artifacts increase
Solution Approach 1:
The patent applies OBMC selectively only to blocks at picture boundaries (top and left boundaries) that are predicted using intra modes, rather than applying it uniformly to all blocks. This localized application maintains coding efficiency improvements while limiting the increase in processing complexity to only where boundary artifacts are most problematic.
Solution Approach 2:
The patent divides the block processing into sub-blocks and applies different prediction strategies: intra prediction for the current block and OBMC using neighboring block vectors for boundary sub-blocks. This segmentation allows the system to leverage both intra prediction accuracy and OBMC artifact reduction where needed, improving overall coding efficiency without applying full OBMC complexity everywhere.
2Manufacturing precision
If OBMC is applied to intra mode blocks, then coding efficiency is improved, but processing complexity increases
Solution Approach 1:
The patent applies OBMC selectively only to blocks at picture boundaries (top and left boundaries) that are predicted using intra modes, rather than applying it uniformly to all blocks. This localized application maintains coding efficiency improvements while limiting the increase in processing complexity to only where boundary artifacts are most problematic.
Solution Approach 2:
The patent applies OBMC partially - only to boundary sub-blocks of intra mode blocks, not to the entire block or to all blocks. This partial application provides sufficient artifact reduction at boundaries where it is most needed, improving compression performance without incurring the full processing cost of applying OBMC to all blocks.
3Manufacturing precision
If OBMC is applied to intra mode blocks, then artifacts are reduced, but bandwidth usage increases
Solution Approach 1:
The patent applies OBMC selectively only to blocks at picture boundaries (top and left boundaries) that are predicted using intra modes, rather than applying it uniformly to all blocks. This localized application maintains coding efficiency improvements while limiting the increase in processing complexity to only where boundary artifacts are most problematic.
Solution Approach 2:
The patent uses neighboring block vectors as predictors for boundary sub-blocks, effectively copying motion information from adjacent blocks. This allows the generation of accurate prediction signals for OBMC without requiring additional transmitted data, reducing bandwidth usage while still achieving artifact reduction through the blended prediction.
Data Source
AI summary
A method for processing video includes receiving a bitstream; and decoding, using coded information of the bitstream, one or more pictures. The decoding includes performing overlapped block motion compensation (OBMC) on a block predicted with an intra mode.