Video Coding Motion Information via Geometric Partitioning Merge
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video coding techniques, such as MPEG-2, MPEG-4, ITU-T.263, ITU-T.264/AVC, HEVC, and VVC, suffer from low coding efficiency and high computational complexity due to limitations in motion information derivation and merge candidate construction.
Innovation Solution
The proposed method enhances video coding efficiency by determining motion information using a geometric partitioning-based merge mode, incorporating motion vector differences, and applying a pruning process during merge list construction, allowing for improved coding efficiency and reduced complexity by considering local characteristics and non-adjacent block motion information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video coding techniques are used, then the coding process is simple, but the coding efficiency is low
Solution Approach 1:
The patent divides the current block into multiple sub-blocks and performs separate motion information derivation for each sub-block. This segmentation allows the use of geometric partitioning to capture local motion characteristics more accurately, thereby improving coding efficiency without excessively increasing overall complexity through structured processing.
Solution Approach 2:
The patent applies different motion derivation strategies to different regions by using geometric partitioning. Each partition can have its own motion vector and reference picture selection, allowing the coding process to adapt to local motion characteristics and improve efficiency while maintaining manageable complexity through localized processing.
2Productivity
If geometric partitioning mode is added to improve coding efficiency, then the coding efficiency improves, but the computational complexity increases
Solution Approach 1:
The patent performs preliminary classification to determine whether geometric partitioning should be applied to each block before detailed processing. This preliminary decision step avoids unnecessary computational overhead for blocks that do not benefit from geometric partitioning, thereby improving coding efficiency where needed while controlling overall computational complexity.
Solution Approach 2:
The patent applies geometric partitioning selectively to only those blocks that benefit from it, rather than uniformly to all blocks. This partial application approach captures the efficiency gains where motion characteristics warrant it while avoiding the computational complexity overhead in regions where conventional methods suffice.
3Productivity
If motion information from non-adjacent blocks is used, then the coding efficiency improves, but the derivation process becomes more complex
Solution Approach 1:
The patent creates a unified motion candidate list that incorporates both adjacent and non-adjacent block motion information through geometric partitioning. This universal approach allows the same derivation framework to handle diverse motion patterns from different block positions, improving coding efficiency while maintaining consistent processing complexity through a standardized multi-functional procedure.
Data Source
AI summary
Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: determining, during a conversion between a target block in a target picture of a video and a bitstream of the video, first motion information of the target block based on a regular merge mode, the first motion information comprising a regular merge list of regular merge candidates for the target block; determining second motion information from the first motion information based on a geometric partitioning-based merge mode and a redundancy check, the second motion information comprising a geometric partitioning-based merge list of geometric partitioning-based merge candidates for the target block; and performing the conversion based on the first motion information and the second motion information. Compared with the conventional solution, the proposed method can advantageously improve the coding efficiency and compression ratios.