Overlapped Block Motion Compensation With Selective Neighbor Skipping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems face challenges in efficiently performing overlapped block motion compensation (OBMC) due to the complexity and resource overhead associated with applying OBMC to video blocks with similar motion information, leading to suboptimal performance and increased computational demands.
Innovation Solution
The system determines whether to apply OBMC based on the similarity of motion vectors between current and neighboring blocks, omitting OBMC when the difference is not substantial, and using simplified filters for luma and chroma components, while extending the prediction block for improved compensation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OBMC is applied to all video blocks, then motion compensation accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies OBMC selectively based on local motion characteristics. By comparing motion vectors between current and neighboring blocks, the system determines whether OBMC is needed for each block boundary, applying the technique only where motion discontinuities exist rather than uniformly across all blocks.
Solution Approach 2:
The patent uses simplified filters with fewer taps than traditional OBMC implementations. Instead of applying full-length filters to all blocks, it uses reduced-complexity filters (e.g., 2-tap bi-linear filters) that provide sufficient compensation for boundary artifacts while significantly reducing computational burden.
2Stability of the object's composition
If OBMC is applied based on motion vector difference, then blocking artifacts are reduced, but processing overhead increases
Solution Approach 1:
The patent performs motion vector comparison before applying OBMC to identify blocks that require compensation. By preliminarily assessing motion vector differences and determining OBMC necessity in advance, the system avoids unnecessary processing overhead while ensuring blocking artifacts are reduced only where needed.
Solution Approach 2:
The patent changes the threshold parameter for motion vector difference comparison to optimize the balance between artifact reduction and processing overhead. By adjusting this threshold, the system can control the number of blocks subjected to OBMC, thereby managing processing overhead while maintaining effective blocking artifact reduction.
3Measurement precision
If extended prediction blocks are used for OBMC, then compensation accuracy at boundaries is improved, but memory requirements increase
Solution Approach 1:
The patent segments the extended prediction block into relevant and irrelevant portions. Only the necessary samples from the extended block are stored in memory and used for OBMC operations, while redundant samples are discarded. This segmentation approach maintains boundary compensation accuracy while reducing memory requirements.
Solution Approach 2:
The patent uses temporary extended prediction blocks that are created only when needed for OBMC operations and then discarded. These extended blocks are not permanently stored but generated on-demand and used immediately, reducing overall memory requirements while maintaining boundary compensation accuracy during critical operations.
Data Source
AI summary
Overlapped block motion compensation (OBMC) may be performed for a current video block based on motion information associated with the current video block and motion information associated with one or more neighboring blocks of the current video block. Under certain conditions, some or all of these neighboring blocks may be omitted from the OBMC operation of the current block. For instance, a neighboring block may be skipped during the OBMC operation if the current video block and the neighboring block are both uni-directionally or bi-directionally predicted, if the motion vectors associated with the current block and the neighboring block refer to a same reference picture, and if a sum of absolute differences between those motion vectors is smaller than a threshold value. Further, OBMC may be conducted in conjunction with regular motion compensation and may use simplified filters than traditionally allowed.