Motion Vector Predictor Scaling for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video compression standards, such as H.264, face challenges in accurately predicting motion vectors across frames with different reference frames and temporal distances, leading to suboptimal compression efficiency.
Innovation Solution
A method is introduced to determine candidate units for motion prediction, scale motion vectors based on temporal distances between reference frames, and select the best predictor from a ranked set of scaled motion vectors to improve prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vectors of neighboring units are used directly to predict MVP without temporal scaling, then the prediction process is simple and fast, but the prediction accuracy deteriorates due to different reference frames and temporal distances
Solution Approach 1:
The patent applies parameter changes by introducing temporal scaling factors that modify the motion vectors based on the temporal distance between reference frames. The scaling factor is calculated as the ratio of temporal distances (current frame to target frame) / (current frame to reference frame), which adjusts the motion vectors to account for different temporal relationships. This resolves the contradiction by improving prediction accuracy through parameter modification while maintaining a relatively simple computational process.
Solution Approach 2:
The patent introduces an intermediary element - the temporal scaling factor - that mediates between the motion vectors of neighboring units and the final MVP prediction. This scaling factor serves as a bridge that adjusts the motion vectors to be temporally consistent with the current frame's reference frame, thereby improving accuracy without requiring complete re-computation of motion vectors.
2Adaptability or versatility
If a fixed predefined candidate set of MVPs is used, then the encoding process is efficient and fast, but the adaptability to different video content and scenarios deteriorates
Solution Approach 1:
The patent applies dynamics by making the candidate set selection adaptive rather than fixed. The system dynamically determines which neighboring units to include in the candidate set based on their availability and spatial relationships to the current unit. This allows the prediction process to adapt to different video content and coding scenarios while maintaining encoding efficiency through a structured selection process.
Solution Approach 2:
The patent applies local quality by treating different spatial locations and coding scenarios differently. The candidate set construction is localized to the current unit's position and configuration, selecting from neighboring units that are spatially adjacent and relevant. This localized approach improves adaptability to local video content characteristics while maintaining overall encoding efficiency.
3Measurement precision
If multiple candidate MVPs are considered and ranked to select the best predictor, then the prediction accuracy improves, but the computational complexity and encoding time increase
Solution Approach 1:
The patent applies partial action by considering a limited set of candidate MVPs from specific neighboring units rather than evaluating all possible motion vectors. The system selectively constructs a candidate set from available neighboring units (A, B, C, D) and ranks only these candidates, achieving improved accuracy over single-source prediction while avoiding the computational burden of exhaustive search.
Data Source
AI summary
The invention provides a motion prediction method. First, a plurality of candidate units corresponding to a current unit of a current frame is determined. A plurality of motion vectors of the candidate units is then obtained. A plurality of scaling factors of the candidate units is then calculated according to a plurality of respective temporal distances depending on a plurality of reference frames of the motion vectors. The motion vectors of the candidate units are then scaled according to the scaling factors to obtain a plurality of scaled motion vectors. The scaled motion vectors are ranked, and a subset of highest ranking motion vectors are identified to be included in a candidate set. Finally, a motion vector predictor for motion prediction of the current unit is then selected from the candidate units.