Multi-View Video Decoding Using Depth-Based Motion Data Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for decoding video data, particularly in 3D video systems, face challenges in efficiently producing motion data, increasing data processing efficiency, and reducing complexity and memory usage during encoding and decoding.
Innovation Solution
A method and apparatus for decoding video data that involves receiving coded video data including multi-view video data and depth data, acquiring motion data for inter-view prediction from depth data, and restoring video data using this motion data, which reduces the number of depth value accesses by using a limited set of depth values for motion vector prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depth-based motion vector prediction is performed using multiple depth values, then motion data accuracy is improved, but memory access complexity and processing time increase
Solution Approach 1:
The patent extracts only the necessary depth values (first, second, third, and fourth depth values corresponding to specific pixel positions) from the depth map for motion vector prediction, rather than using all depth values. This selective extraction reduces memory access complexity and processing time while maintaining sufficient motion data accuracy for inter-view prediction.
Solution Approach 2:
The patent segments the depth-based motion vector prediction process into distinct components: acquiring specific depth values at defined pixel positions, calculating motion vectors using these segmented depth values, and performing inter-view prediction. This segmentation allows for optimized processing of each component independently, reducing overall complexity.
2Measurement precision
If depth-based motion vector prediction is performed using multiple depth values, then motion data accuracy is improved, but device complexity increases
Solution Approach 1:
The patent extracts only the necessary depth values (first, second, third, and fourth depth values corresponding to specific pixel positions) from the depth map for motion vector prediction, rather than using all depth values. This selective extraction reduces memory access complexity and processing time while maintaining sufficient motion data accuracy for inter-view prediction.
Solution Approach 2:
The patent applies different processing approaches to different regions of the video data by using view identifiers to determine which depth map picture to reference. This local quality approach optimizes processing for each specific coding unit based on its spatial and temporal characteristics.
3Measurement precision
If all depth values are accessed for motion vector calculation, then motion prediction accuracy is improved, but memory usage increases
Solution Approach 1:
The patent extracts only the necessary depth values (first, second, third, and fourth depth values corresponding to specific pixel positions) from the depth map for motion vector prediction, rather than using all depth values. This selective extraction reduces memory access complexity and processing time while maintaining sufficient motion data accuracy for inter-view prediction.
Data Source
AI summary
Disclosed is a method and apparatus for decoding video data. The method for decoding video data includes receiving coded video data including multi-view video data and depth data corresponding to the video data, acquiring motion data for inter-view prediction of a coding unit of the coded video data from the depth data, and performing inter-view prediction based on the motion data, and restoring video data according to the multi-view video data including the coding unit and the depth data based on the motion prediction.