Building roof facet segmentation method based on three-dimensional Voronoi graph

Abstract

The invention discloses a building roof facet segmentation method based on a three-dimensional Voronoi graph, and relates to the technical field of building three-dimensional reconstruction by using the airborne laser radar data. According to the method, building roof facet segmentation in the airborne LiDAR point cloud is performed by using the three-dimensional Voronoi graph and natural divisionof space and topological connectivity of the three-dimensional Voronoi graph are fully utilized so that the problem that the neighborhood is not easy to control in construction of the point cloud data spatial topological relation can be effectively solved and a new approach can be provided for performing building roof facet segmentation by using the airborne LiDAR point cloud data.

Description

technical field [0001] The invention relates to the technical field of three-dimensional reconstruction of buildings using airborne laser radar data, in particular to a building roof patch segmentation method based on a three-dimensional Voronoi diagram. Background technique [0002] Maps are the traditional way and carrier of expressing, transmitting and researching geographical information. The emergence of digital earth and digital city has broken this traditional concept, enabling human beings to obtain information from two-dimensional to three-dimensional in describing and analyzing surface space affairs. breakthrough leap. 3D city models have been widely used in urban planning, microclimate research, air pollution analysis, noise propagation analysis, military, communications, virtual reality and many other fields. The demand for 3D models of urban buildings is increasing day by day. [0003] Airborne LiDAR (Light Detection and Ranging) is a new type of non-contact ac...

AI Technical Summary

Problems solved by technology

This type of method has a great dependence on the accuracy of digital elevation model extraction. In addition, because it is difficult to select a suitable window, it is difficult to eliminate large bushes and bushes closely related to buildings by morphological erosion methods;

[0011] The third type of remote sensing classification method is to use supervised classification, unsupervised classification, fuzzy theory, neural network, machine learning and other classification methods in remote sensing classification to classify and judge point cloud data. This type of method provides LiDAR point cloud data processing. A new method has been found, but more in-depth research is still needed to improve the versatility and practicability;

Images