论文部分内容阅读
3D building reconstruction in urban area is a multidisciplinary research topic in spatial information processing and applications.Aerial photogrammetry is one of the traditional means for building model collection,which usually consists of two steps: firstly,extracting 2D building geometric features from aerial imageries,and then recovering 3D information and key parameters of a model by matching and grouping algorithms.However,there are several problems arising when the traditional aerial photogrammetry techniques are used in automatic 3D building model reconstruction,these include the difficulties of automatic building feature extraction,the long processing cycle of stereo image matching and the loss of geometry information of the building.Airborne Light Detection And Ranging (LiDAR),a recently developed active sensor with Positioning and Orientation System (POS) and laser range finder assembled,can directly acquire 3D point cloud data with high density.These datasets accurately describe the topological relationships between objects without the need for image matching.Airborne LiDAR has been becoming an economically saving but yet efficient system for three dimensional surface point data acquisition.