Point Cloud Decoding with Subgroup LoD and Neighbor Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for processing point cloud data are inefficient due to the high computational complexity and latency associated with encoding and decoding large volumes of data, requiring improved methods for efficient transmission and reception.
Innovation Solution
The method involves receiving and decoding point cloud data on a subgroup basis, generating levels of detail (LoD) using nodes from both the current and parent subgroups, and performing neighbor searches to derive attribute data efficiently, with separate storage for direct compression nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If point cloud data is processed using conventional encoding and decoding methods, then the data can be transmitted and received, but the computational operation becomes complicated and time-consuming due to the large volume of point cloud data
Solution Approach 1:
The patent divides the point cloud data into multiple subgroups based on spatial partitioning (e.g., octree structure). Each subgroup is processed independently through encoding and decoding operations, which reduces the computational complexity and time required compared to processing the entire point cloud as a single unit. This segmentation enables parallel processing and improves throughput.
Solution Approach 2:
The patent performs preliminary actions by pre-processing point cloud data into subgroups and preparing decoding parameters before actual transmission and decoding operations. This includes pre-calculating spatial relationships, preparing attribute data mappings, and establishing decoding sequences, which reduces the computational burden during real-time decoding operations.
2Productivity
If point cloud data is processed using conventional methods, then encoding and decoding can be performed, but latency increases due to the large volume of data requiring processing
Solution Approach 1:
By segmenting the point cloud into subgroups, the patent enables selective decoding of only the necessary subgroups based on viewing requirements or spatial priorities. This reduces the total processing time and latency compared to decoding the entire point cloud, while maintaining the ability to process large volumes of data efficiently.
Solution Approach 2:
The patent applies partial decoding by processing only the necessary portions of the point cloud data (subgroups) rather than the entire dataset. This partial action approach reduces latency and processing time while maintaining sufficient detail for the application requirements, avoiding the excessive action of processing all data when not needed.
3Quantity of substance
If conventional encoding methods are used for large point cloud data, then compression can be achieved, but the computational complexity increases significantly
Solution Approach 1:
The patent segments the encoding process into independent operations on subgroups, where each subgroup is encoded using simplified rules compared to the entire dataset. This reduces the computational complexity per operation while still achieving effective compression of the large point cloud data through hierarchical processing.
Solution Approach 2:
The patent applies different encoding strategies to different subgroups based on their local characteristics (e.g., spatial density, importance, viewing angle). This local quality approach allows the system to use more complex encoding only where necessary and simpler encoding where acceptable, reducing overall computational complexity while maintaining compression effectiveness.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A point cloud data reception method and device are provided. The point cloud data reception method according to embodiments may comprise the steps of: receiving geometry data, attribute data and signaling information; decoding the geometry data on the basis of the signaling information; decoding the attribute data on the basis of the signaling information and the decoded geometry data; and rendering point cloud data including the decoded geometry data and the decoded attribute data.