Affine Motion Vector Prediction for Complex Video Block Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in achieving further bitrate reduction without compromising image quality, particularly in predicting motion vectors for video images.
Innovation Solution
A method for predicting motion vectors using an affine transform model-based approach, where a decoder side constructs a candidate motion vector list for a current block by utilizing spatially adjacent decoded blocks, allowing for different affine transform models to better match the actual motion status of the current block, thereby improving coding efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional motion vector prediction methods are used, then the coding process is simple, but the bitrate reduction is insufficient and coding efficiency is limited
Solution Approach 1:
The current block is divided into multiple subblocks, and different affine transform models (4-parameter or 6-parameter) are selectively applied to different subblocks based on motion characteristics. This segmentation allows the system to achieve high coding efficiency for complex motion regions while maintaining simplicity for uniform motion regions, thus resolving the contradiction between coding efficiency and model complexity.
Solution Approach 2:
The patent implements dynamic selection of affine transform models (4-parameter or 6-parameter) based on the actual motion status of each block. The system adaptively chooses the appropriate model complexity for each region, making the prediction process dynamic rather than static. This dynamic approach enables the system to achieve high coding efficiency when complex motion is present while avoiding unnecessary complexity when simple motion suffices.
2Measurement precision
If a single affine transform model is used for all blocks, then the processing is uniform and simple, but the accuracy in matching actual motion status is insufficient
Solution Approach 1:
Different subblocks within the current block can use different affine transform models (4-parameter or 6-parameter) according to their local motion characteristics. This local quality approach ensures that each region is predicted with the most appropriate model for its specific motion type, thereby improving motion vector prediction accuracy without requiring a single complex model for the entire block.
Solution Approach 2:
The patent changes the parameters of the affine transform model (number of parameters: 4 or 6) based on the motion characteristics of different blocks. By adjusting the model parameters adaptively, the system achieves higher prediction accuracy for complex motions while maintaining efficiency for simpler motions, thus resolving the contradiction between precision and adaptability.
3Reliability
If more bitrate is allocated for motion information, then the image quality is preserved, but the compression ratio is reduced
Solution Approach 1:
The patent applies the more complex 6-parameter affine transform model only to subblocks that exhibit complex motion characteristics, while using the simpler 4-parameter model for subblocks with uniform motion. This partial application of excessive modeling complexity only where needed reduces the overall bitrate requirement while maintaining image quality in regions that demand it, thus resolving the contradiction between reliability and quantity of substance.
Data Source
AI summary
A motion vector prediction method includes parsing a bitstream to obtain an index value of a candidate motion vector list and constructing the candidate motion vector list that includes. candidate motion vectors of K control points of a current block. The candidate motion vectors of K control points are obtained based on a 2N-parameter affine transform model used for a neighboring block of the current block, where N and K are integers greater than or equal to 2 and less than or equal to 4 with N not being equal to K. The method further includes determining, in the candidate motion vector list, target candidate motion vectors of the K control points based on the index value and obtaining a predicted motion vector of each subblock of the current block based on the target candidate motion vectors of the K control points.