Spatial Tree Subdivision for Large CAD Model Visualization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CAD systems face challenges in efficiently managing and visualizing large models due to memory limitations and data bandwidth issues, particularly when dealing with models exceeding 10,000 parts, as existing methods consume excessive time and resources.
Innovation Solution
The implementation of a spatial tree data structure that subdivides complex cells into subcells using a multi-threaded process, allowing for efficient management and visualization of large models by limiting data loading to only visible regions and utilizing a spatial hierarchy to maintain connectivity and performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If current CAD systems use Large Model Visualization methods with semantically-organized parts, then model geometry can be loaded in organized units, but the system consumes excessive memory and time when dealing with models exceeding 10,000 parts
Solution Approach 1:
The patent applies spatial partitioning by dividing the 3D model into a hierarchical tree structure where each node represents a spatial region containing a subset of model parts. This segmentation allows the system to manage large models by processing only relevant portions at different levels of the hierarchy, reducing memory consumption while maintaining organized access to model geometry.
Solution Approach 2:
The patent transitions from semantic organization (parts-based) to spatial organization (region-based) by introducing a spatial tree data structure. This dimensional change in organization allows the system to manage models by their spatial distribution rather than their semantic structure, enabling efficient handling of arbitrarily large models through spatial queries and partitioning.
2Reliability
If current CAD systems load complete model geometry, then all model data is available for processing, but data bandwidth and processing time become excessive for large models
Solution Approach 1:
The patent extracts only the necessary portions of model geometry for processing by using spatial queries to identify and load only those parts of the model that are currently needed. The spatial tree structure enables selective extraction of model data based on spatial criteria, reducing data bandwidth and processing time while maintaining reliability through on-demand loading of required geometry.
Solution Approach 2:
The patent performs preliminary spatial partitioning and organization of model data into a hierarchical tree structure before actual processing occurs. This preliminary action pre-organizes the data in a way that enables efficient selective loading and processing, avoiding the need to load complete model geometry while ensuring all necessary data is available when needed.
3Quantity of substance
If the spatial tree is subdivided using a single-threaded process, then memory usage is controlled, but processing speed decreases for large models
Solution Approach 1:
The patent implements a dynamic processing approach where the subdivision process can adapt between single-threaded and multi-threaded modes based on the complexity of the spatial tree and available system resources. For simpler models or memory-constrained scenarios, single-threaded processing is used to maintain memory control. For complex models with sufficient resources, multi-threaded processing is activated to significantly improve subdivision speed while maintaining memory usage within acceptable limits.
Data Source
AI summary
A method and related CAD system and computer readable medium. A method includes loading an object model in a CAD system, the object model including a plurality of parts each of which may contain one or more shapes which in turn are composed of multiple polygons. The method includes adding the shapes to a spatial tree, the shapes each corresponding to at least one cell, each cell corresponding to a spatial region of the object model. The method also includes, for each cell that is too complex to process within a memory space of the CAD system, subdividing the cell into a plurality of subcells using a first subdivision process. The method also includes subdividing each cell into a plurality of subcells using a multi-threaded subdivision process, and combining the subcells into the spatial tree.