Multi-feature-weighted Building Extraction Modeling Under Regular Grid Projection

The vehicle laser scanning technology is widely used in the extraction of surface geographic information because of its high precision and rapid acquisition of the three-dimensional geographic information data of the surface.The classification,segmentation and modeling of the objects need to take into account the user's needs and applications,and because the laser scanning data is a collection of large discrete points,they are discontinuous in space,and have a lot of elevation information,the above factors determine the Importance of Classification Extraction.In this paper,we consider the diversity of vehicle laser data acquisition form,the diversity of building spatial distribution and the diversity of geometric features,and propose a multi-characteristic building based on regular grid projection And the method of modeling for domestic buildings is studied on this basis.Mainly applicable to the vehicle point cloud data building rapid segmentation extraction modeling.In this paper,the relationship between the points of the point cloud data is taken as the starting point,the projection point cloud data is mapped to the XOY plane,and the feature data is generated based on the grid point center and its eigenvalues.After the geometric feature segmentation method,The image is divided from the building point cloud data.On this basis,according to the building configuration of the building,a reasonable modeling method is selected to reverse the mathematical modeling of the segmented building point cloud data.The buildings with regular rules in various environments can be extracted by this method.Because the algorithm contains elevation parameters,the data of the building point cloud can be modeled quickly and effectively.In this study,a government building in Fengqing County,Lincang City,Yunnan Province as the experimental object,the application of the rules of the grid design under the multi-feature classification algorithm,split the building point cloud data,and finally get the government building The three-dimensional digital model of.The experimental results show that the proposed algorithm can quickly and accurately extract the cloud data of the building from the point cloud data of the vehicle and complete the reverse modeling of the building.