Affine Motion Estimation for Screen Content Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards, such as HEVC, are inefficient in handling high contrast and limited color data with duplicated regions in screen content videos, as they primarily rely on two-dimensional translation motion models for intra prediction, which do not effectively capture complex motions like scaling, rotation, and transformation.
Innovation Solution
The introduction of affine motion estimation and compensation in the prediction method, which creates candidate lists for coding blocks and performs affine motion estimation to obtain coding parameters, allowing for non-translation motion modeling such as scaling, rotation, and transformation, thereby improving coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional two-dimensional translation motion models are used for intra prediction, then the coding structure remains simple, but coding efficiency is insufficient for screen content videos with complex motions
Solution Approach 1:
The patent transforms the motion model from a simple two-dimensional translation model to an affine motion model by changing the mathematical parameters involved. The affine model introduces additional parameters (rotation angle, scaling factor, shear angle) that enable it to represent complex motions including rotation, scaling, and shearing, thereby improving coding efficiency for screen content videos while maintaining a structured approach to motion compensation
2Productivity
If affine motion estimation is introduced to model complex motions, then coding efficiency improves, but the number of coded bits increases
Solution Approach 1:
The patent implements a hybrid approach where affine motion estimation is applied selectively rather than universally. The system first attempts simpler motion models and only applies the more complex affine model when necessary, thereby capturing complex motions where needed while avoiding unnecessary bit overhead for regions that can be adequately represented by simpler models
Solution Approach 2:
The patent divides the prediction process into multiple stages or modes, allowing different motion models to be applied to different regions or under different conditions. This segmentation enables the system to use the computationally efficient two-dimensional translation model for simple regions while applying the more accurate affine model only where complex motions are detected, balancing precision and bit cost
3Measurement precision
If affine motion estimation is used for all coding blocks, then complex motions are accurately captured, but computational complexity increases significantly
Solution Approach 1:
The patent introduces dynamic adaptation mechanisms that allow the system to adjust the motion model complexity based on the actual content characteristics. The system dynamically selects between two-dimensional translation and affine motion models by analyzing motion patterns in different regions, ensuring high accuracy for complex motions while maintaining low computational complexity for simpler regions
Data Source
AI summary
A prediction method, including: creating a candidate list corresponding to a coding mode for the coding mode used by a coding block in an original picture, wherein the coding mode includes a first affine motion mode or a second affine motion mode, a reference block in the candidate list is spatially adjacent to the coding block, and the reference block is coded using the first affine motion mode or the second affine motion mode; performing affine motion estimation on the coding block according to the candidate list corresponding to the coding mode to obtain coding parameters corresponding to the coding block; and performing predictive coding on the coding block on the basis of the coding parameters corresponding to the coding block.