Simplified Motion Vector Predictor Derivation in HEVC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing motion vector prediction (MVP) derivation process in High Efficiency Video Coding (HEVC) is inefficient due to the high number of checking operations required for redundancy removal, which can lead to increased system operation time and reduced coding efficiency.
Innovation Solution
A simplified method that reduces redundancy checking operations by only applying it to spatial MVP candidates, excluding the temporal MVP, and directly adding the zero motion vector to the candidate set when the number of candidates falls below a threshold, thereby reducing the maximum number of checking operations from 3 to 1 and eliminating unnecessary checks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If redundancy checking is applied to all MVP candidates (spatial and temporal), then the accuracy of MVP candidate set is improved, but the number of checking operations increases from 1 to 3
Solution Approach 1:
The patent segments the MVP candidate set into spatial MVP candidates and temporal MVP candidate, applying redundancy checking only to the spatial candidates. This segmentation allows the system to maintain accuracy for spatial predictions while reducing overall checking operations by excluding temporal candidates from the redundancy check process.
Solution Approach 2:
The patent extracts the temporal MVP candidate from the redundancy checking process entirely. By taking out the temporal candidate from the checking operations, the system reduces the number of checks from 3 to 1 while still maintaining a complete MVP candidate set that includes both spatial and temporal predictors.
2Reliability
If comprehensive redundancy checking is performed on all MVP candidates, then the reliability of motion vector coding is improved, but the system operation time increases
Solution Approach 1:
The patent segments the checking process to apply redundancy checking only to spatial MVP candidates, excluding temporal candidates. This segmentation maintains reliability for spatial predictions (which are more critical) while reducing overall operation time by eliminating unnecessary checks on temporal candidates.
3Adaptability or versatility
If redundancy checking is applied to spatial and temporal MVP candidates, then the completeness of candidate set is improved, but the number of operations exceeds the threshold
Solution Approach 1:
The patent extracts the temporal MVP candidate from the redundancy checking operations, allowing the system to maintain a complete candidate set (including both spatial and temporal predictors) while reducing the number of operations to meet efficiency thresholds.
Solution Approach 2:
The patent applies partial redundancy checking only to spatial MVP candidates rather than performing excessive checking on all candidates. This partial action is sufficient to ensure candidate set completeness while improving coding efficiency by reducing unnecessary operations.
Data Source
AI summary
A method and apparatus for deriving a motion vector predictor (MVP) candidate set for motion vector coding of a current block. Embodiments according to the present invention determine a redundancy-removed spatial-temporal MVP candidate set. The redundancy-removed spatial-temporal MVP candidate set is derived from a spatial-temporal MVP candidate set by removing any redundant MVP candidate. The spatial-temporal MVP candidate set includes a top spatial MVP candidate, a left spatial MVP candidate and one temporal MVP candidate. The method further checks whether candidate number of the redundancy-removed spatial-temporal MVP candidate set is smaller than a threshold, and adds a zero motion vector to the redundancy-removed spatial-temporal MVP candidate set if the candidate number is smaller than the threshold. Finally, the method provides the redundancy-removed spatial-temporal MVP candidate set for encoding or decoding of the motion vector of a current block.