Motion Vector Range Derivation for Video Coding Boundary Clipping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video coding, the motion vector range is significantly reduced near picture boundaries, leading to inefficiencies in compression and decompression processes, especially when using Enhanced bi-linear Interpolation Filter (EIF) in Essential Video Coding (EVC), due to clipping requirements that limit the range to a single point.
Innovation Solution
A method is introduced to derive and update motion vector ranges for coding blocks, ensuring they remain within valid boundaries by adjusting the minimum and maximum values based on the block size and position relative to the picture boundaries, thereby maintaining a double value of the motion vector spread and the area size, which prevents clipping and allows for effective pixel-based motion compensation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If motion vector clipping is applied to fulfill memory limitation requirements, then memory usage is reduced, but motion vector range is significantly reduced (sometimes to one point) near picture boundaries
Solution Approach 1:
The patent applies different motion vector range adjustments based on the local position within the picture. Specifically, it calculates separate horizontal and vertical motion vector ranges (hor_min, hor_max, ver_min, ver_max) that are adapted to the local boundary conditions of each coding block, rather than applying a uniform clipping rule throughout the entire picture. This allows the motion vector range to be optimized locally at each position while maintaining overall memory efficiency.
2Reliability
If motion vector range is reduced near picture boundaries, then memory limitations are satisfied, but compression efficiency deteriorates
Solution Approach 1:
The patent implements dynamic motion vector range adjustment where the horizontal and vertical ranges (hor_min, hor_max, ver_min, ver_max) are calculated adaptively based on the center motion vector position and the deviationMV value. This dynamic calculation ensures that the motion vector range is optimized for each specific coding block's position and characteristics, thereby maintaining compression efficiency while satisfying memory constraints.
3Quantity of substance
If motion vector range is clipped to a single point, then memory requirements are met, but interpolation accuracy is lost
Solution Approach 1:
The patent changes the parameters defining the motion vector range by calculating hor_min, hor_max, ver_min, and ver_max based on the center motion vector and deviationMV. This parameter adjustment maintains a meaningful range (not reduced to a single point) while adapting to memory constraints, thereby preserving the precision needed for accurate interpolation and prediction.
Data Source
AI summary
A method for coding video data is disclosed. The method comprises: obtaining a center motion vector of a coding block; deriving a first motion vector range for the coding block based on the center motion vector and a motion vector spread, wherein the motion vector spread depends on a size of the coding block; if the first motion vector range is at least partially pointing outside a first area including a reference picture, updating the first motion vector range to point within the first area, such that a minimum value or a maximum value of the updated first motion vector range is pointing at a boundary of the first area.