Neighboring Block Availability Ordering for Parallel Motion Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding and decoding methods face challenges in determining the usability of neighboring blocks for motion candidate lists and achieving parallelism between prediction units within coding units, which affects video encoding and decoding efficiency.
Innovation Solution
A method and apparatus for determining the availability of neighboring blocks in configuring a motion candidate list by deriving block availability information and motion information based on a predetermined order restriction, ensuring parallel processing of merge candidate lists.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If availability determination for neighboring blocks is performed without order restriction, then processing parallelism is improved, but determination accuracy deteriorates due to dependency conflicts
Solution Approach 1:
The patent applies preliminary action by establishing a predetermined order restriction for determining availability of neighboring blocks before parallel processing begins. This order restriction is pre-configured to eliminate dependency conflicts, allowing each prediction unit to determine availability independently without encountering circular dependencies during parallel execution.
Solution Approach 2:
The patent segments the availability determination process by dividing it into distinct ordered steps for different neighboring blocks. Instead of attempting to determine all neighboring block availabilities simultaneously, the method segments them into a specific sequence (e.g., left block first, then upper blocks in predetermined order), enabling parallel processing while maintaining determination accuracy through the segmented approach.
2Measurement precision
If motion candidate list is configured using all neighboring blocks, then prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies local quality by selectively determining availability for different neighboring blocks based on their specific positions and relationships to the current block. Rather than uniformly processing all neighboring blocks, the method applies different availability determination rules to different locations (e.g., left block vs. upper blocks), incorporating local spatial characteristics to improve prediction accuracy while managing complexity.
Solution Approach 2:
The patent implements partial action by determining availability only for the necessary subset of neighboring blocks required for motion candidate list configuration. The predetermined order restriction enables the system to stop availability determination once sufficient candidates are identified, avoiding the computational overhead of evaluating all possible neighboring blocks while still achieving accurate prediction.
3Speed
If parallel processing is implemented for merge candidate lists, then encoding/decoding speed is improved, but processing correctness deteriorates due to dependency conflicts
Solution Approach 1:
The patent applies preliminary action by pre-establishing an order restriction for neighboring block availability determination before parallel processing of merge candidate lists begins. This preliminary ordering prevents dependency conflicts during parallel execution, ensuring that each prediction unit can independently and correctly determine availability without interfering with other parallel processes, thus maintaining processing correctness while achieving speed improvement.
Solution Approach 2:
The patent applies copying by creating independent copies of the availability determination logic for each prediction unit processing merge candidate lists in parallel. Each prediction unit has its own instance of the determination algorithm that operates independently according to the predetermined order restriction, eliminating conflicts between parallel processes while maintaining identical correctness guarantees.
Data Source
AI summary
The image decoding method according to the present invention comprises the steps of: determining the availability of a block adjacent to a current block; configuring a merge candidate list for the current block on the basis of the result of the determined availability; and performing motion compensation for the current block on the basis of the merge candidate list, wherein the step of determining availability comprises the steps of inducing block availability information on the adjacent block and inducing the availability of motion information on the adjacent block, and the step of determining the availability of the adjacent block is such that when there is more than one adjacent block, the availability of adjacent blocks can be determined in accordance with a predetermined order restriction.


