A method for automatic separation of ground model and non-ground model of real 3D model

Abstract

The invention discloses an automatic separation method for a ground model and a non-ground model for a three-dimensional model of a real scene. The present invention simulates the free fall of the cloth on the inverted three-dimensional model of the real scene by using the particle physics model with the distance constraint for the surface element, and determines the final form of the cloth through the collision detection between the particle and the three-dimensional model of the real scene, which is used as an approximation of the ground model, The ground model and the non-ground model are locally distinguished by comparing the spatial relationship between the particle and the surfel of the 3D model of the real scene. The invention has the advantages of high efficiency, high robustness and high precision, and can be used for the separation of the ground model and the non-ground model of the real scene three-dimensional model of urban, woodland, rural and other scenes and any surface element density.

Description

technical field [0001] The invention belongs to the field of surveying and mapping data processing, and is an automatic separation method for a ground model and a non-ground model for a three-dimensional model of a real scene. Different 3D scenes such as cities, villages, woodlands, etc. Background technique [0002] In the past two decades, benefiting from the rapid development of UAV technology and optical sensor technology, airborne LiDAR and oblique photogrammetry technology have been able to obtain large-scale 3D surface information quickly and accurately at low cost. In particular, the UAV oblique photogrammetry technology can effectively collect the coordinates and texture information of the top and side elevations of buildings, which is more and more important in the construction of 3D digital cities and smart cities. The collected 3D surface information exists in the form of point clouds after being processed, and these point clouds can be used to further generate ...

AI Technical Summary

Problems solved by technology

At present, a large number of studies have emerged on the ground filtering, object extraction, and scene segmentation of DSM, DOM, and point cloud data. blank

After the 3D model is generated from the point cloud, its vertex density is lower than that of the point cloud, and there is a problem of density imbalance. Therefore, it is difficult to obtain the desired effect by directly applying the relevant algorithm of the point cloud to the model vertices.

Due to the unbalanced density of bins in different areas of the real-world 3D model, the bins in flat areas or artificially repaired areas are generally larger. Therefore, if the vertices of the real-world 3D model are processed directly using the point cloud cloth simulation method to separate the ground bins , the separation effect of the unbalanced bin density and the artificially repaired real-world 3D model is poor

Images