Point Cloud Data Processing via Component Track Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing methods for point cloud data, particularly in immersive media applications, suffer from inefficiencies in storage, transmission, and processing due to direct encoding and packing, leading to data loss and reduced processing accuracy.
Innovation Solution
A data processing method and apparatus that involves obtaining component tracks and atlas segments, packing these components based on component track data boxes, and transmitting them as a target file, allowing for precise unpacking and decoding by user devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If point cloud data is directly encoded and packed, then the storage and transmission process is simple, but data loss occurs and processing accuracy deteriorates
Solution Approach 1:
The patent segments point cloud data into multiple component tracks (geometry, texture, semantics, etc.) with different priorities and characteristics. Each component is independently encoded and packed with appropriate parameters, allowing selective processing and reconstruction without losing overall data integrity, thus resolving the contradiction between simple processing and high accuracy.
2Reliability
If all point cloud data is processed and decoded, then complete data recovery is achieved, but time and computational resources are wasted
Solution Approach 1:
The patent enables partial decoding by allowing users to selectively reconstruct only the necessary component tracks based on their specific needs. For example, geometry-only reconstruction for rough modeling or selective reconstruction of specific semantic components, avoiding the time-consuming process of decoding all data while maintaining reliability for the required portion.
3Productivity
If component tracks are organized with detailed information boxes, then data organization and processing efficiency are improved, but the complexity of data structure increases
Solution Approach 1:
The patent creates a universal data structure where component tracks can contain multiple types of information boxes (segment quantity, space region, time region, parameter information) that serve different functions. This multi-functional structure standardizes the organization of diverse data types while maintaining processing efficiency through consistent access methods, balancing complexity and productivity.
Data Source
AI summary
Embodiments of this application provides a data processing method performed by a computer device, the method including: obtaining k component tracks in a target file, each component track having a component track data box, which comprises a segment quantity and space region information; the segment quantity referring to a quantity of atlas segments corresponding to the component track to which the component track data box belongs; the space region information referring to a space region of the component track to which the component track data box belongs; k being a positive integer; and unpacking the target file according to each component track data box, to obtain a corresponding segment display content.