Rapid Construction of 3D Models of Urban Buildings Based on UAV Photogrammetric Data
The traditional manual modeling of buildings has a long cycle,high cost,and loss of geospatial information.Although ob-lique photography and 3D laser scanning technology can quickly build 3D models of urban buildings and have rich geospatial informa-tion,they are difficult to edit and manage due to high resource occupation and low degree of individualization.In this regard,based on UAV photogrammetric data,this paper presents a new method to extract buildings from orthophoto by using U-Net semantic segmenta-tion network,building height information is obtained by DSM,and CGA rules are compiled to construct 3D models of urban buildings in batch.The effectiveness of the modeling process is verified by example analysis.