Motion Prediction Method Using Temporal Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motion prediction methods in video processing, such as those used in the H.264 compression standard, face inefficiencies due to improper temporal scaling of motion vectors from neighboring units and a fixed candidate set for motion vector predictor selection, which can lead to suboptimal prediction accuracy and coding efficiency.
Innovation Solution
A method that involves calculating temporal scaling factors for motion vectors of candidate units based on their reference frame distances to the current frame, and adaptively selecting the best motion vector predictor from a dynamically determined candidate set, considering the characteristics of the current unit and neighboring units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vectors of neighboring units are used directly for MVP prediction without temporal scaling, then the prediction process is simple and fast, but the prediction accuracy deteriorates due to improper temporal scaling
Solution Approach 1:
The patent applies preliminary action by calculating temporal scaling factors before using motion vectors for MVP prediction. The scaling factors are pre-computed based on temporal distances between reference frames, allowing motion vectors to be properly scaled in advance, which improves prediction accuracy without adding significant complexity to the main prediction process
Solution Approach 2:
The patent changes the parameter of motion vectors by applying temporal scaling factors to adjust their magnitude according to temporal distances. This parameter transformation allows motion vectors from different reference frames to be properly normalized, resolving the accuracy issue while maintaining a relatively simple computational approach
2Productivity
If a fixed candidate set for MVP selection is used, then the selection process is simple and deterministic, but the coding efficiency deteriorates due to suboptimal predictor selection
Solution Approach 1:
The patent applies dynamics by making the candidate set adaptive rather than fixed. The candidate set is dynamically determined based on characteristics of the current unit and neighboring units, allowing the system to adapt to different video content and improve coding efficiency. This dynamic approach replaces the static fixed candidate set with a flexible, content-adaptive selection mechanism
Solution Approach 2:
The patent applies local quality by determining candidate sets based on local characteristics of specific units. Different candidate sets are selected for different current units depending on their specific properties and the properties of their neighboring units, rather than using a universal fixed set. This localized adaptation improves prediction accuracy for each specific context
3Measurement precision
If temporal scaling is applied to motion vectors of all neighboring units, then the prediction accuracy is improved, but the computational complexity increases due to additional scaling calculations
Solution Approach 1:
The patent applies partial action by selectively applying temporal scaling to motion vectors based on the specific needs of each prediction case. Rather than universally scaling all motion vectors regardless of necessity, the method applies scaling where it provides benefit while avoiding unnecessary computations in cases where it is not needed, thus balancing accuracy improvement with computational efficiency
Data Source
Figure 1
Figure 2
Figure 3
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 temporal scaling factors of the candidate units is then calculated according to a plurality of temporal distances between a plurality of reference frames of the motion vectors and the current frame. The motion vectors of the candidate units are then scaled according to the temporal scaling factors to obtain a plurality of scaled motion vectors. Finally, a motion vector predictor for motion prediction of the current unit is then selected from the candidate units according to the scaled motion vectors.