Affine Motion Compensation for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies, such as HEVC and VVC, face limitations in accurately predicting complex motions like zoom, rotation, and perspective due to their reliance on translation motion models, which can lead to inefficiencies in compression and decoding processes.
Innovation Solution
The implementation of affine motion compensation techniques, including the use of affine models with control point motion vectors and sub-block based predictions, allows for more precise motion modeling and prediction, enhancing coding efficiency by storing and signaling affine model parameters and deriving new affine models for improved prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If translation motion models are used for video coding, then device complexity is reduced, but motion prediction accuracy deteriorates for complex motions like zoom, rotation, and perspective
Solution Approach 1:
The current block is divided into multiple sub-blocks, each with its own motion model parameters. This segmentation allows different regions to have different motion characteristics, improving prediction accuracy for complex motions while keeping each individual sub-model relatively simple.
Solution Approach 2:
The patent transitions from translation motion models to affine motion models by changing the mathematical parameters used for motion compensation. Affine models include additional parameters (rotation, scaling, shear) that better represent complex motions, directly improving prediction accuracy despite increased complexity.
2Productivity
If affine motion models are used to improve motion prediction accuracy, then compression efficiency improves, but computational complexity increases
Solution Approach 1:
Instead of applying full affine models to the entire block, the patent applies simplified affine models to smaller sub-blocks. This partial application reduces the overall computational burden while still capturing complex motion in regions where it occurs, maintaining compression efficiency.
Solution Approach 2:
The patent performs motion estimation and parameter derivation in advance during the encoding process, storing these parameters for use during decoding. This preliminary action shifts computational complexity to the encoding phase, allowing simpler decoding operations while maintaining high compression efficiency.
3Measurement precision
If multiple affine models are combined for prediction, then prediction accuracy improves, but processing time increases
Solution Approach 1:
The block is segmented into sub-blocks that can be processed independently with different affine models. This allows parallel processing of multiple sub-blocks, reducing overall processing time while maintaining the benefits of multiple models for improved prediction accuracy.
Solution Approach 2:
The patent merges multiple affine models into a unified prediction framework where parameters from multiple models are combined to generate the final prediction. This merging process is optimized to reduce redundant computations, balancing the need for accurate multi-model prediction with acceptable processing time.
Data Source
AI summary
A method, a video processing apparatus and computer-readable program medium for video process is disclosed. A method of video processing includes storing, during a conversion between a current block of video and a bitstream representation of the current block, affine model parameters for the current block; and performing the conversion using the affine model parameters, wherein the affine model parameters are derived at least from motion vectors of control points of the current block.