Research On Buildings′ Three-dimensional Reconstruction Based On Air Borne And Vehicle Borne LiDAR Data

Buildings are the most important parts of the city, With the development of "digital city", the 3 d model of buildings not only require timely update, but also require that they can truly reflect the accurate detail characteristics of buildings. In order to effectively use the air-vehicle borne LiDAR data for the three-dimensional reconstruction of buildings, based on air-vehicle borne LiDAR point cloud,this paper research on the extraction of buildingsˊtop and fa?ade features and the three-dimensional reconstruction of buildings, including the following three aspects:(1)The extraction of buildingsˊtop features. In this section,this paper precisely extracts the point cloud of buildingsˊtop, taking the point cloud of buildingsˊtop as the research object, this paper extracts the outline vectors of buildingsˊtop,and segment the complex building roof with the regional growth method,then, obtains the plane equation according to the least squares criterion.(2) The extraction of buildingsˊfacade features. taking the segmented LiDAR data as the research object, this section research on using projection point density and elevation for fine filtering,and then extracts the outline vectors of buildingsˊtop and the window feature by the point cloud projected on the horizontal plane and fa?ade.(3)Firstly,this section research on contradistinction analysis between buildingsˊtop and bottom outline vectors,and then counts the result of the contradistinction. Combine with Google Street View image,this paper find out detail which needs to be payed attention in the three-dimensional reconstruction of buildings.Finally, this paper selects three kinds of buildings, on the basis of complementary use,this paper builds their 3D models by jointing buildingsˊtop and bottom features.