Motion Vector Prediction Range Constraints for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in motion vector prediction due to the unrestricted range of motion vector components, which can lead to increased memory requirements and compression inefficiencies, especially when scaled motion vectors exceed the typical 16-bit length.
Innovation Solution
The proposed solution restricts the range of motion vector components to a specified range, such as [−32768, 32767] in quarter pixel units, by scaling and clipping horizontal and vertical components, thereby ensuring they fit within 16-bit signed integer values, reducing memory usage and improving compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the range of motion vector components is unrestricted, then motion vector prediction accuracy is improved, but memory requirements increase and compression efficiency deteriorates
Solution Approach 1:
The patent applies parameter changes by imposing constraints on motion vector component ranges (e.g., limiting horizontal and vertical components to specific intervals). This transforms the unrestricted motion vector parameters into bounded parameters, reducing the memory space needed to store and process motion vector data while maintaining adequate prediction accuracy for typical video content.
Solution Approach 2:
The patent segments the motion vector prediction process by treating horizontal and vertical components separately, applying different range constraints to each component. This segmentation allows independent optimization of each dimension's memory usage while preserving the overall prediction functionality.
2Productivity
If scaled candidate motion vectors are used, then motion estimation efficiency is improved, but motion vector length may exceed 16-bit capacity
Solution Approach 1:
The patent changes the parameters of scaled motion vectors by applying clipping operations that limit the magnitude of horizontal and vertical components. This ensures that even after scaling, the resulting motion vectors fit within 16-bit signed integer ranges, maintaining data type compatibility while preserving the efficiency gains from scaling.
Solution Approach 2:
The patent applies partial scaling by selectively scaling motion vector components only when necessary and within safe ranges. The clipping operation acts as a safeguard that prevents excessive scaling beyond 16-bit capacity, allowing the system to benefit from scaling where appropriate while avoiding overflow issues.
3Quantity of substance
If motion vector range is restricted to specified intervals, then memory bandwidth utilization is improved, but motion vector prediction flexibility is reduced
Solution Approach 1:
The patent optimizes memory bandwidth utilization by changing the parameter ranges of motion vectors to fit within standard data type limits (16-bit signed integers). This parameter restriction reduces the memory bandwidth required to transfer motion vector data while maintaining adequate prediction capability for most video sequences through intelligent range selection.
Data Source
AI summary
In general, techniques are described for performing motion vector prediction for video coding. A video coding device comprising a processor may perform the techniques. The processor may be configured to determine a plurality of candidate motion vectors for a current block of the video data so as to perform the motion vector prediction process and scale one or more of the plurality of candidate motion vectors determined for the current block of the video data to generate one or more scaled candidate motion vectors. The processor may then be configured to modify the scaled candidate motion vectors to be within a specified range.