Motion Vector Prediction in Multiview Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in efficiently predicting motion vectors across multiple views in Multiview Video Coding (MVC), which hinders effective data compression and transmission of 3D video data.
Innovation Solution
The techniques involve using disparity and temporal motion vectors from different views to predict motion vectors for current blocks, with the option to scale these vectors based on view identifiers and picture order counts, and incorporating them into candidate lists for motion vector prediction and merge modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion vectors are predicted using only traditional spatial and temporal methods, then the coding process remains simple, but coding efficiency is insufficient for multiview video data
Solution Approach 1:
The patent introduces inter-view prediction as a new dimension for motion vector prediction in multiview video coding. Instead of only using spatial (same picture) and temporal (different time) prediction, the invention utilizes motion vectors from different views to predict motion in the current view, thereby improving coding efficiency for 3D video data without significantly increasing implementation complexity
2Loss of information
If motion vectors from different views are used for prediction, then data redundancy is reduced, but the complexity of motion vector prediction increases
Solution Approach 1:
The patent performs preliminary scaling of motion vectors from different views based on view identifiers and picture order counts before using them for prediction. This pre-processing step adjusts the motion vectors to account for different camera positions and timing, reducing data redundancy while keeping the prediction algorithm manageable by establishing a standardized reference framework
3Measurement precision
If motion vectors are scaled based on view identifiers and picture order counts, then prediction accuracy improves, but computational overhead increases
Solution Approach 1:
The patent scales motion vectors from different views by adjusting parameters such as view identifiers and picture order counts. This parameter-based scaling approach improves prediction accuracy by accounting for geometric relationships between views and temporal relationships between frames, while maintaining computational efficiency through straightforward mathematical transformations rather than complex iterative algorithms
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Aspects of this disclosure relate to a method of coding video data. In an example, the method includes identifying a first block of video data in a first temporal location from a first view, wherein the first block of video data is associated with a first temporal motion vector. The method also includes determining, when a second motion vector associated with a second block of video data comprises a temporal motion vector and the second block is from a second view, a motion vector predictor for the second motion vector based on the first temporal motion vector. The method also includes coding prediction data for the second block using the motion vector predictor.