3D Model Partitioning With Multi-LoD Loading for Low-Latency XR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for delivering 3D content in extended reality (XR) experiences suffer from high latency and inefficient data transfer, leading to prolonged loading times and interrupted user experiences.
Innovation Solution
A 3D model processing system that partitions 3D models based on location or material, generating graded partitions of varying levels of detail, allowing selective loading and rendering of model parts to optimize data transfer and reduce latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the entire 3D model is loaded and transmitted at once, then the user receives complete content, but the data transfer size is large and latency is high
Solution Approach 1:
The 3D model is divided into multiple original partitions based on location or material characteristics, and each original partition is further divided into multiple graded partitions with different levels of detail. This hierarchical segmentation allows selective transmission of only necessary model parts, significantly reducing initial data transfer size and loading time while maintaining complete content delivery when needed.
2Manufacturing precision
If high-detail 3D model data is transmitted, then rendering quality is improved, but data transfer size increases and processing time increases
Solution Approach 1:
Different partitions of the 3D model are assigned different levels of detail based on their importance and visibility requirements. Critical areas receive high-detail representations while less important areas use lower-detail representations. This local quality differentiation maintains overall rendering quality while significantly reducing total data transfer size.
Solution Approach 2:
The system dynamically selects and transmits partitions at appropriate levels of detail based on real-time requirements such as user position, viewing angle, and network conditions. This dynamic adaptation allows the system to optimize between rendering quality and data transfer size continuously during operation.
3Productivity
If the 3D model is divided into many small partitions, then data transfer efficiency is improved and latency is reduced, but system complexity increases
Solution Approach 1:
The partitioning system uses a nested hierarchical structure where the 3D model is divided into original partitions, which are further divided into graded partitions with different levels of detail. This nested organization allows efficient management of numerous partitions through a structured hierarchy, reducing the complexity burden while maintaining high data transfer efficiency.
Solution Approach 2:
The 3D model is pre-partitioned into original and graded partitions during model creation or preprocessing, establishing the hierarchical structure in advance. This preliminary action eliminates the need for complex real-time partitioning decisions during runtime, reducing system complexity while enabling efficient selective data transfer.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
This disclosure provides a 3D model processing system (100). The 3D model processing system includes a storage circuit (102) and a processor (104). The storage circuit (102) is configured to store a program code. The processor (104) is coupled to the storage circuit (102) and configured to access the program code to execute: obtaining (S210) a plurality of nodes of a 3D model; splitting (S220) the 3D model into a first original partition (320, 330) and a second original partition (320, 330) based on a location (300B) or a material (300A), wherein the first original partition (320, 330) comprises a first group of nodes and the second original partition (320, 330) comprises a second group of nodes; generating (S240) a plurality of first graded partitions (340) of different levels of detail (LoD 0, LoD 1, LoD 2) for the first original partition (320, 330) based on the first group of nodes; and generating (S250) a plurality of second graded partitions (340) of different levels of detail (LoD 0, LoD 1, LoD 2) for the second original partition (320, 330) based on the second group of nodes.