Sub-Picture Motion Vector Filtering for Independent Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems face challenges in efficiently encoding and decoding sub-pictures due to motion vectors pointing outside the sub-picture boundaries, leading to errors and inefficiencies in compression and decompression processes.
Innovation Solution
Implementing a flag to treat sub-pictures as independent entities by excluding motion vectors that point outside the sub-picture boundaries, along with mechanisms to control filtering and boundary signaling, ensuring accurate and efficient extraction and display of sub-pictures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion vectors from collocated blocks are used for sub-picture coding, then coding efficiency is improved, but errors occur when motion vectors point outside sub-picture boundaries
Solution Approach 1:
The patent extracts and removes motion vector candidates that point outside the sub-picture boundary from the candidate list. This is achieved by checking the validity of each motion vector candidate against the sub-picture boundaries and excluding invalid candidates, thereby preventing decoding errors while maintaining coding efficiency.
Solution Approach 2:
The patent applies different quality control to different parts of the motion vector candidate list. Valid motion vectors (those pointing inside the sub-picture) are retained with full quality, while invalid motion vectors (those pointing outside) are excluded. This local quality differentiation ensures reliability without sacrificing overall coding efficiency.
2Adaptability or versatility
If sub-pictures are coded as independent entities, then separate extraction and display is enabled, but motion vectors pointing outside boundaries cause errors
Solution Approach 1:
The patent segments the picture into independent sub-pictures with clearly defined boundaries. Each sub-picture is treated as an independent coding unit with its own motion vector validation, enabling separate extraction and display while preventing cross-boundary errors through boundary-aware motion vector filtering.
Solution Approach 2:
The patent applies preliminary anti-action by pre-validating motion vector candidates against sub-picture boundaries before they are used for prediction. This preventive measure ensures that no invalid motion vectors can cause decoding errors, thereby maintaining reliability while enabling independent sub-picture processing.
3Manufacturing precision
If all motion vector candidates are considered, then coding precision is improved, but resource usage increases and errors may occur
Solution Approach 1:
The patent extracts only the valid motion vector candidates from the complete set of candidates. By removing invalid candidates that point outside sub-picture boundaries, it reduces the data processing volume while maintaining the precision needed for accurate motion compensation within the valid region.
Data Source
AI summary
A video coding mechanism includes receiving a bitstream comprising a current picture including a sub-picture coded according to inter-prediction. Coded blocks contain candidate motion vectors for a current block of the sub-picture. The coded blocks include a collocated block from a different picture. A candidate list of candidate motion vectors for the current block are derived by excluding collocated motion vectors from the candidate list when the collocated motion vectors are included in the collocated block, when the collocated motion vectors point outside of the sub-picture, and when a flag is set to indicate the sub-picture is treated as a picture. A current motion vector for the current block is determined from the candidate list of candidate motion vectors. The current block is decoded based on the current motion vector. The current block is forwarded for display as part of a decoded video sequence.