Component-Wise Adaptive Spatial Resolution for Video Motion Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques face inefficiencies in transmitting motion vectors due to the limitation of having the same spatial resolution for both horizontal and vertical components, leading to increased transmission overhead and potential data loss when component values differ significantly.
Innovation Solution
A method for adaptively determining spatial resolution for each component of a motion vector, allowing separate resolution settings for horizontal and vertical components based on flags and precision indices in the bitstream, enabling more efficient coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a single spatial resolution is used for both horizontal and vertical components of motion vectors, then device complexity is reduced and ease of operation is improved, but transmission overhead increases and data loss occurs when component values differ significantly
Solution Approach 1:
The motion vector transmission is segmented into two independent parts: horizontal component and vertical component. Each component can have its own spatial resolution setting, allowing independent optimization. The encoder transmits separate resolution indicators for horizontal and vertical components, enabling the decoder to reconstruct motion vectors with appropriate precision for each direction without being constrained by a unified resolution setting.
Solution Approach 2:
Different spatial resolutions are applied locally to different components of the motion vector based on their specific requirements. The horizontal component may use one resolution while the vertical component uses another, allowing each component to have the quality appropriate for its characteristics and the motion patterns it represents, rather than forcing a uniform quality level across both components.
2Measurement precision
If adaptive motion vector resolution is applied to motion vectors with greatly different component values, then motion vector precision is improved, but transmission overhead increases when using the resolution of the smaller component as reference
Solution Approach 1:
The spatial resolution parameter is changed independently for horizontal and vertical components based on their respective value magnitudes and requirements. Instead of selecting a single resolution that fits both components, the encoder adjusts the resolution parameter separately for each component, allowing optimal precision for each while avoiding the overhead of transmitting unnecessary precision information for components that don't require it.
3Productivity
If component-wise adaptive spatial resolution is implemented for multiple motion vectors, then coding efficiency is improved, but device complexity increases due to separate resolution determination for each component
Solution Approach 1:
The spatial resolution for horizontal and vertical components is determined in advance during the encoding process, before motion vector transmission. The encoder pre-calculates and stores the appropriate resolution values for each component based on the block characteristics and motion patterns. This preliminary determination simplifies the transmission phase, as the decoder can directly use these pre-established resolution settings without performing complex real-time calculations.
Data Source
AI summary
A video coding method and apparatus adaptively determine spatial resolution for a horizontal component and a vertical component of a motion vector or block vector indicating a position of a reference block when performing inter prediction or intra block copy (IBC) of a current block. The video coding method and apparatus separately determine each spatial resolution for a plurality of motion vectors.