Subblock Affine Motion Model for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding and decoding technologies face inefficiencies in inter prediction methods, particularly in affine motion compensation, where storing control-point motion vectors (CPMVs) requires significant buffer storage, and determining subblock motion vectors as CPMVs can lead to increased complexity and prediction errors.
Innovation Solution
The technique involves selecting subblock motion vectors as control-point motion vectors for a current block, generating an affine motion model using these subblock motion vectors, and predicting the current block based on the model, thereby reducing the need for separate CPMV storage and simplifying the prediction process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If control-point motion vectors are stored separately in buffer, then affine motion compensation can be performed, but buffer storage requirements increase significantly
Solution Approach 1:
The patent merges the storage of control-point motion vectors with the subblock motion vector storage. Instead of maintaining separate CPMV buffers, the system uses the existing subblock motion vector storage to also hold the control-point motion vectors, thereby eliminating the need for additional buffer storage while preserving affine motion compensation capability
Solution Approach 2:
The patent makes the subblock motion vector storage serve multiple functions: it stores both the subblock motion vectors needed for inter prediction and the control-point motion vectors needed for affine motion compensation. This multi-functional approach eliminates the need for dedicated CPMV storage space
2Quantity of substance
If subblock motion vectors are determined as control-point motion vectors, then storage requirements are reduced, but prediction complexity and errors increase
Solution Approach 1:
The patent applies local quality by selecting specific subblocks (corner subblocks) to serve as control points for affine motion compensation. Instead of using all subblocks uniformly, it identifies and uses only the corner subblocks' motion vectors as control-point motion vectors, thereby reducing complexity while maintaining prediction accuracy
Solution Approach 2:
The patent segments the block into corner subblocks and uses only those specific segments' motion vectors for affine motion compensation. This segmentation approach allows the system to use a subset of subblock motion vectors (corner ones) as control points, reducing the complexity of determining and storing CPMVs while maintaining sufficient prediction accuracy
Data Source
AI summary
A device for processing video data includes a memory configured to store video data and one or more processors implemented in circuitry. The one or more processors are configured to select two or more subblocks of one or more blocks of video data and generate an affine motion model for a current block of video data using a subblock motion vector from each of the two or more subblocks as control-point motion vectors for the current block. The one or more processors are further configured to determine one or more prediction blocks based on the affine motion model and predict the current block with the one or more prediction blocks.