Motion Field Storage in Video Coding Using Partition Mode
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motion field storage processes in video coding are complex and inefficient, requiring numerous comparisons to determine whether subblocks are in uni-predicted or bi-predicted areas, leading to increased computational complexity and resource utilization.
Innovation Solution
The proposed method simplifies motion field storage by determining whether a subblock is in a bi-predicted area based on its position and storing either uni-prediction or bi-prediction motion vectors accordingly, using larger subblocks such as 8×8 or 16×16 units to reduce the number of comparisons and streamline the process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current motion field storage processes are used to ensure accurate motion vector storage for each subblock, then motion coding precision is maintained, but computational complexity and resource utilization increase significantly
Solution Approach 1:
The patent divides the block into larger subblocks (8×8 or 16×16) rather than using finer granularity, and further segments the bi-predicted area identification into two cases based on reference picture list relationships. This segmentation strategy reduces the total number of comparisons needed while maintaining adequate motion compensation precision for each segment.
Solution Approach 2:
The patent performs preliminary classification by checking whether the first and second reference picture lists are the same before proceeding to detailed motion vector storage. This preliminary action avoids unnecessary comparisons for subblocks that can be handled by the simplified rule, reducing overall computational complexity while preserving accuracy where needed.
2Measurement precision
If numerous comparisons are performed to determine uni-predicted or bi-predicted areas for each subblock, then motion field accuracy is maintained, but processing time and resource utilization increase
Solution Approach 1:
The patent segments the motion field processing into two distinct cases: (1) when reference picture lists are the same, and (2) when they differ. This segmentation allows the use of simpler processing rules for the first case, significantly reducing processing time while maintaining accuracy through the more rigorous second case handling.
Solution Approach 2:
The patent changes the processing parameter based on the reference picture list relationship. When lists are the same, it uses a simplified storage rule; when they differ, it applies the standard comparison-based approach. This parameter change optimizes processing time by avoiding unnecessary comparisons while preserving motion field accuracy through context-appropriate processing.
3Productivity
If larger subblocks such as 8×8 or 16×16 are used for motion vector storage, then the number of comparisons is reduced, but motion field granularity and precision may be compromised
Solution Approach 1:
The patent uses 8×8 or 16×16 subblocks as segmentation units, which are larger than traditional 4×4 blocks. This segmentation reduces the number of motion vectors to store and process. The patent compensates for the reduced granularity by implementing case-based logic that ensures accurate motion field representation within each larger subblock, particularly through the two-case classification system.
Solution Approach 2:
The patent adjusts the motion field storage parameter based on the reference picture list relationship. For subblocks where reference lists are the same, it stores motion vectors more efficiently. This parameter adaptation allows larger subblocks to maintain adequate precision by adjusting the storage and processing behavior according to the specific case, balancing productivity gains with precision requirements.
Data Source
AI summary
The present disclosure provides method and apparatus for motion field storage in video coding. An exemplary method includes: determining whether a first uni-prediction motion vector for a first partition of a block and a second uni-prediction motion vector for a second partition of the block are from a same reference picture list; and in response to the first uni-prediction motion vector and the second uni-prediction motion vector being determined to be from the same reference picture list, storing, in a motion field of the block, one of the first uni-prediction motion vector and the second uni-prediction motion vector for a subblock located in a bi-predicted area of the block.