Grid Model Generation via 3D to 2D Point Cloud Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing grid model generation algorithms are complex and inefficient, leading to low performance in generating virtual structures that simulate three-dimensional objects.
Innovation Solution
A method and system for generating a grid model by obtaining a three-dimensional contour point cloud, determining two-dimensional contour point clouds, and creating grid sections between them, which can include triangular or quadrangular sections, to simplify the process and improve efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing algorithms are used for generating grid models, then the generation process can be completed, but the algorithm complexity is high and generation efficiency is low
Solution Approach 1:
The patent segments the 3D contour point cloud into multiple 2D contour point clouds at different heights, then processes each 2D contour independently to generate grid sections. This division transforms a complex 3D processing problem into simpler 2D problems, reducing algorithm complexity while improving generation efficiency through parallel processing capability
Solution Approach 2:
The patent introduces height as an additional dimension parameter, slicing the 3D point cloud into multiple 2D layers at different heights. By adding this dimensional parameter, the system can process contours layer by layer, simplifying the overall algorithm structure and enabling more efficient grid model generation
2Manufacturing precision
If complex algorithms are used to ensure accurate grid model generation, then generation accuracy can be maintained, but processing time increases significantly
Solution Approach 1:
By segmenting the 3D point cloud into multiple 2D contours at different heights, the patent enables independent processing of each contour layer. This segmentation maintains accurate representation of the original 3D structure while allowing parallel processing of individual layers, thereby reducing total processing time without sacrificing grid model accuracy
Solution Approach 2:
The patent performs preliminary sorting of contour points within each 2D contour before grid section generation. This preliminary organization of data prepares the contours in an optimal format for subsequent grid generation, ensuring accurate results while minimizing processing time through efficient data preparation
Data Source
AI summary
The present disclosure provides a grid model generation method, device, and storage medium. The grid model generation method includes obtaining a three-dimensional contour point cloud related to an object; determining a plurality of two-dimensional (2D) contour point clouds based on the three-dimensional contour point cloud; generating a plurality of grid sections between two adjacent two-dimensional contour point clouds of the plurality of two-dimensional contour point clouds; and generating a grid model of the object based on the plurality of grid sections