Secondary Spatial Merge Candidates for Incomplete Motion Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In existing video coding standards like VVC and ECM, the availability of motion information from neighboring blocks may be limited, leading to incomplete spatial merge candidates for current coding units, which affects compression efficiency.
Innovation Solution
Extend the derivation of spatial merge candidates by incorporating motion information from secondary neighboring blocks when primary neighboring blocks are unavailable or not in inter mode, ensuring the merge candidate list is fully populated with unique motion information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion information is derived only from primary neighboring blocks, then the derivation process is simple, but the spatial merge candidate list may be incomplete when primary neighboring blocks are unavailable or not in inter mode
Solution Approach 1:
The patent segments the neighboring blocks into two categories: primary neighboring blocks (immediate neighbors) and secondary neighboring blocks (next-nearest neighbors). This segmentation allows the system to first attempt derivation from primary blocks and only recourse to secondary blocks when necessary, maintaining simplicity while ensuring completeness.
Solution Approach 2:
The patent establishes a preliminary derivation process that first attempts to obtain motion information from primary neighboring blocks before falling back to secondary neighboring blocks. This preliminary action ensures that the simpler primary block derivation is used when possible, while preparing the secondary block mechanism as a backup for completeness.
2Productivity
If motion information from secondary neighboring blocks is incorporated, then the spatial merge candidate list becomes complete, but the derivation process becomes more complex
Solution Approach 1:
The patent applies partial action by only incorporating secondary neighboring blocks when the primary neighboring blocks are unavailable or not in inter mode. This selective application ensures that the additional complexity of secondary block derivation is justified only when necessary to achieve complete spatial merge candidates, thereby improving coding efficiency without unnecessarily increasing process complexity.
3Productivity
If the merge candidate list is fully populated with unique motion information, then compression efficiency improves, but the processing time increases
Solution Approach 1:
The patent performs preliminary derivation from primary neighboring blocks, which are the most likely to provide valid motion information. Only when this preliminary process yields insufficient candidates does the system proceed to the more time-consuming secondary neighboring block derivation, thereby minimizing overall processing time while ensuring complete candidate lists when necessary.
Solution Approach 2:
The patent changes the parameter of derivation scope dynamically: using primary neighboring blocks when available (faster, simpler process) and switching to include secondary neighboring blocks only when primary blocks are unavailable (slower, more comprehensive process). This adaptive parameter change optimizes the trade-off between processing time and compression efficiency.
Data Source
AI summary
An apparatus includes at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to: determine whether a number of spatial merge candidates comprising motion information derived from primary neighboring blocks of a current coding unit is less than a number of allowed spatial merge candidates; determine a spatial merge candidate comprising motion information derived from at least one secondary neighboring block of the current coding unit, in response to the number of spatial merge candidates comprising motion information derived from the primary neighboring blocks of the current coding unit being less than the number of allowed spatial merge candidates; and code the current coding unit using the spatial merge candidate.


