Building Boundary Extraction Based on Lidar Point Clouds Data

LIDAR (Light Detection And Ranging) is an active remote sensing system, which could provide fleetly three dimensional information of earth surface with high vertical accuracy. Now the building boundary extraction and normalization is the key approach for 3D modeling of building and city mapping. In this paper, a new algorithm named Alpha Shapes is developed to extract the building boundary. Compared with other algorithms, Alpha Shapes algorithm works effectively in inner and outer boundaries extraction from point clouds data with convex and concave polygon shape. Moreover it can keep fine features of buildings adaptively and filter the footprints of non-building. In addition, an improved boundary simplifying algorithm is suggested to refine the extracted building boundary. And two regularization algorithms are developed to make the refined boundary regular. The experiments proved that the normalized building boundary is generated perfectly by these algorithms.