G-PCC Point Cloud Transmission with Temporal Sample Grouping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for processing point cloud data are inefficient, leading to high latency and encoding/decoding complexity, and lack support for temporal scalability and efficient storage and access of geometry-based point cloud compressed data.
Innovation Solution
A method and device for processing point cloud data that includes generating and manipulating a sample group description box in a geometry-based point cloud compression (G-PCC) file, ensuring efficient storage and access, and supporting temporal scalability by managing entry counts in the sample group description box.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If point cloud data is processed using existing methods, then the data can be transmitted, but the processing efficiency is low resulting in high latency and encoding/decoding complexity
Solution Approach 1:
The patent segments the G-PCC bitstream into multiple temporal levels (e.g., base layer and enhancement layers) with different resolutions and qualities. This segmentation allows receivers to process only the necessary portion of data based on their capabilities and network conditions, reducing overall processing complexity and latency while maintaining the ability to handle complete high-quality data when needed.
2Adaptability or versatility
If point cloud data is compressed without temporal scalability, then the file structure is simpler, but temporal scalability and efficient access to different quality levels are not supported
Solution Approach 1:
The patent introduces a dynamic file structure with sample group description boxes that can adapt to different temporal levels. The structure includes flexible fields such as entry_count that dynamically indicates the number of samples at each temporal level, allowing the system to scale from simple base layers to complex multi-level structures based on actual data requirements without imposing fixed complexity on all cases.
3Reliability
If all point cloud data is stored at highest quality, then quality is maximized, but storage efficiency and network bandwidth utilization are reduced
Solution Approach 1:
The patent changes the quality parameter of point cloud data by organizing it into temporal levels with different resolution and quality characteristics. The base layer provides lower-quality data suitable for bandwidth-constrained scenarios, while enhancement layers provide progressively higher quality. This allows the system to transmit and store data at appropriate quality levels based on network conditions and receiver capabilities, reducing overall data volume while maintaining necessary quality.
4Reliability
If the sample group description box includes all possible entries, then completeness is ensured, but redundant entries increase processing overhead
Solution Approach 1:
The patent applies partial action by including only the necessary sample group description entries corresponding to actual temporal levels present in the data. The entry_count field dynamically indicates how many entries are included, avoiding the inclusion of redundant entries for temporal levels that do not exist. This ensures data completeness for present levels while improving processing efficiency by eliminating unnecessary entries.
Data Source
AI summary
A transmission device for point cloud data, a method performed by the transmission device, a reception device, and a method performed by the reception device are provide. The method performed by the reception device for point cloud data, according to the present disclosure, may comprise the steps of: acquiring a geometry-based point cloud compression (G-PCC) file including the point cloud data; and acquiring, from the G-PCC file, a sample group description box having a predetermined group type, wherein based on the sample group description box being present in a track within the G-PCC file, a count of entries within the sample group description box is equal to a value obtained by adding 1 to a highest temporal level identifier value included in the track.