Affine Mode Motion Vector Prediction for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies, such as H.265/HEVC and H.266/VVC, face challenges in optimizing motion vector prediction accuracy and reducing code amount in affine modes, which affects encoding efficiency and complexity.
Innovation Solution
An encoder and decoder system that derives motion vectors for control points of a current block by combining motion vectors from neighboring blocks, using affine mode techniques to improve prediction accuracy and reduce code amount by projecting motion vectors onto control points and encoding difference vectors selectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vectors from multiple neighboring blocks are combined and projected onto control points, then prediction accuracy is improved, but device complexity increases
Solution Approach 1:
The current block is divided into multiple control points (e.g., four control points at corners), and motion vectors are separately derived for each control point based on different neighboring blocks. This segmentation allows independent optimization of prediction accuracy for each control point while managing complexity through localized processing.
Solution Approach 2:
Motion vectors from multiple different neighboring blocks (first neighboring block, second neighboring block, etc.) are combined and merged to derive motion vectors for control points. This merging of information from multiple sources improves prediction accuracy by utilizing diverse motion information while the systematic combination method controls complexity.
2Loss of substance
If difference vectors are encoded selectively for motion vector predictors, then code amount is reduced, but manufacturing precision deteriorates
Solution Approach 1:
Difference vectors are encoded selectively rather than for all motion vector predictors. The patent encodes difference vectors only when necessary (when prediction accuracy requires it), performing partial encoding action. This reduces code amount while maintaining sufficient encoding precision through conditional application.
Solution Approach 2:
The patent changes the parameter of motion vector representation by using difference vectors (residuals) instead of encoding full motion vectors. This parameter transformation allows more efficient code representation while maintaining precision, as difference vectors typically require fewer bits to represent the deviation from predicted values.
Data Source
AI summary
Provided is an encoder including circuitry and memory. In operation, the circuitry: derives one or more first motion vectors based on at least one of one or more motion vectors of a first block encoded in an affine mode; derives one or more second motion vectors based on at least one of one or more motion vectors of a second block that has been encoded and is different from the first block; and determines, as motion vector predictors at control points of a current block to be encoded, motion vectors including at least one of the first motion vectors and at least one of the second motion vectors.