Optimal 3D Model Trees for Quality-Aware XR Streaming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for displaying 3D geometry data in extended reality (XR) environments often result in long loading times due to the need to load all data at once or load data when a user approaches a virtual object, interrupting the user experience.
Innovation Solution
A method for generating an optimal tree for rendering or streaming a 3D model based on quality and visibility values of each node, allowing the 3D model to be rendered with reduced data amounts, prioritizing higher details for closer and more visible parts, and lower details for distant or less visible parts, optimizing data transfer and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If all 3D geometry data is loaded at once, then complete visual quality is achieved, but loading time increases significantly
Solution Approach 1:
The patent segments the complete 3D geometry data into a hierarchical tree structure with multiple levels of detail (LOD). Each node in the tree represents a portion of the 3D model at a specific detail level, allowing the system to load and render only the necessary segments based on viewer position and importance, rather than loading the entire model at once.
Solution Approach 2:
The patent applies local quality by assigning different detail levels to different parts of the 3D model based on their visual importance and visibility. Critical areas closer to the viewer are rendered with higher detail, while distant or less important areas use lower detail, optimizing the distribution of computational resources and data loading.
2Loss of time
If data is loaded when user approaches virtual object, then loading time is reduced, but user experience is interrupted
Solution Approach 1:
The patent performs preliminary action by pre-calculating and organizing the 3D model data into a hierarchical tree structure with quality and visibility values assigned to each node before runtime. This pre-processing enables the system to quickly determine which data to load based on current viewer position without interrupting the user experience during interaction.
Solution Approach 2:
The patent implements dynamics by making the data loading and rendering process adaptive to the user's real-time position and movement. The system dynamically adjusts which portions of the 3D model are loaded and rendered at what quality levels based on the viewer's current location, maintaining smooth user experience while optimizing performance.
3Manufacturing precision
If high detail is used for all parts, then visual quality is maximized, but data transfer and processing load increases
Solution Approach 1:
The patent applies local quality by assigning different detail levels to different parts of the 3D model based on their visual importance and visibility. Critical areas closer to the viewer are rendered with higher detail, while distant or less important areas use lower detail, optimizing the distribution of computational resources and data loading.
Solution Approach 2:
The patent changes the quality parameter dynamically based on viewer position, distance, and object importance. By adjusting the detail level parameter for different regions of the 3D model, the system maintains high visual quality where needed while reducing data transfer and processing requirements for less critical areas.
Data Source
Figure 1A
Figure 1B
Figure 2A
AI summary
This disclosure provides a method for generating an optimal tree for rendering or streaming a 3D model. The method includes: obtaining a complete tree of a 3D model; obtaining a quality value of a node of the complete tree of the 3D model; obtaining a visibility value of the node of the complete tree of the 3D model; determining an optimal tree of the 3D model based on the complete tree, the quality value, and the visibility value; and rendering the 3D model based on the optimal tree.