Non-Parametric 3D City Modeling Using Point Cloud Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D city modeling methods face challenges with extreme computational demand and scalability issues due to their reliance on parametric approaches and predefined shapes, making them inefficient for generating detailed and accurate models of diverse urban areas.
Innovation Solution
A non-parametric and non-iterative method using a k-dimensional (kd)-tree to assess the proximity and nature of neighboring 3D points, allowing for efficient generation of detailed 3D geographical models by removing vertical structures, segmenting rooftops, and connecting points to a base for accurate representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If parametric approaches with predefined shapes are used for 3D city modeling, then the modeling process can be standardized, but extreme computational demand and scalability issues occur
Solution Approach 1:
The patent replaces traditional parametric modeling approaches with a point-cloud-based direct modeling system. Instead of using predefined shapes and iterative optimization algorithms, the system directly processes 3D sensor data points to generate city models, eliminating the computational burden of parameter optimization while maintaining modeling capability
Solution Approach 2:
The patent changes the fundamental parameters of the modeling approach by transitioning from shape-based parametric models to point-cloud-based geometric models. This parameter change allows the system to work with raw sensor data directly without requiring predefined shape parameters, thereby reducing computational demand
2Ease of manufacture
If parametric approaches with predefined shapes are used for 3D city modeling, then the modeling process can be standardized, but scalability across different architectural environments deteriorates
Solution Approach 1:
The patent segments the city modeling process into independent processing stages: point cloud acquisition, processing, and model generation. This segmentation allows the system to handle different architectural types independently without requiring global parameter adjustments, thereby improving scalability across diverse urban environments
Solution Approach 2:
The patent creates a universal point-cloud-based modeling framework that can process various architectural types using the same core algorithms. The system's ability to work with raw 3D data points makes it adaptable to different building styles, heights, and configurations without requiring architecture-specific predefined shapes
3Extent of automation
If automatic methods are used for 3D city modeling, then time and effort of specialized personnel are reduced, but computational complexity increases
Solution Approach 1:
The patent extracts the essential geometric information directly from 3D sensor point clouds without requiring complex intermediate processing steps. By taking out only the necessary data elements needed for modeling and processing them through simplified algorithms, the system maintains high automation while reducing computational complexity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An apparatus for generating 3D geographical models includes a processor and memory storing executable computer program code causing the apparatus to at least perform operations including removing 3D points of a cloud depicting vertical structures in a geographic area responsive to analyzing a set of 3D points of the cloud. The points include data indicating geocoordinates of the geographic area. The computer program code further causes the apparatus to segment rooftops of vertical structures responsive to analyzing the set. The computer program code further causes the apparatus to delineate points of segmented rooftops responsive to extracting a boundary and a portion of rooftops. The computer program code further causes the apparatus to generate 3D geographic models depicting the geographic area based on connecting delineating points to a base including a height lower than heights of delineating points. Corresponding methods and computer program products are also provided.