Motion Vector Candidate Ranking for Error-Resilient Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing moving picture coding and decoding methods using inter prediction lack sufficient error resistance, particularly due to reliance on reference picture information that may be lost during transmission, leading to incorrect decoding of motion vector predictor candidates.
Innovation Solution
A method for determining a maximum number of motion vector predictor candidates, deriving additional candidates if needed, and coding an index for the selected predictor, independent of reference picture information, to enhance error resistance and coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion vector predictor candidates are derived based on reference picture information, then coding efficiency is improved, but error resistance deteriorates when reference pictures are lost during transmission
Solution Approach 1:
The patent derives motion vector predictor candidates from already-coded adjacent blocks within the current picture before transmission occurs. This preliminary action creates multiple candidate predictors that can be selected during decoding even if reference picture information is lost, thereby maintaining both coding efficiency and error resistance
Solution Approach 2:
The patent changes the source of motion vector predictor candidates from reference picture information (temporal prediction) to adjacent block information (spatial prediction). This parameter change allows the system to maintain coding efficiency through multiple candidates while improving error resistance by not depending on potentially lost reference pictures
2Manufacturing precision
If multiple motion vector predictor candidates are derived and selected, then coding precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the derivation of motion vector predictors into two distinct groups: first candidates from spatially adjacent blocks and second candidates from temporally adjacent blocks. This segmentation allows systematic processing of multiple candidates while managing complexity through organized, separate derivation paths
Solution Approach 2:
The patent derives more motion vector predictor candidates than the minimum single predictor would provide, creating an excess of candidates that enables more precise selection. This partial excessive action improves coding precision by providing multiple options while the complexity is managed through standardized derivation procedures
Data Source
AI summary
A moving picture coding apparatus for coding a current block of a picture is provided. A motion vector candidate list including motion vector candidates is generated and, for each of the motion vector candidates, a value is also generated. A second candidate is arranged before a first candidate in the motion vector candidate list. The first candidate is sorted before the second candidate based on the value of the first candidate and the value of the second candidate, with the value of the first candidate being higher than the value of the second candidate. A motion vector candidate to use to code the current block is selected, and an index to identify the selected motion vector candidate is coded. For each candidate of the motion vector candidates, the value is based on a number of times the candidate is selected as a motion vector predictor for neighboring blocks.