An Optimized Modeling Method Based on Hierarchical Reconstruction for Three-dimensional Building Point Cloud

Abstract

In this paper, we propose an optimized modeling method based on hierarchical reconstruction for 3D building point cloud. In contrast to existing approaches, this method can make full use of the geometric information of the original building facade point cloud and roof point cloud, which will provide a more complete description for the model of buildings. Firstly, we segment and extract the facade and the roof point cloud of the building based on the difference in planar slices. Further, we use the P-Linkage algorithm to segment the facade point cloud into planar patches, and cluster the planar patches to layer the facade point cloud. Therefore, we reduce the complexity of reconstruction from three-dimensional to two-dimensional. Then we reconstruct the building facade based on the hierarchical reconstruction method and the building roof based on the block reconstruction method. Finally, we merge the facade model and roof model and clip unnecessary polygons, leading to the final model. We experiment with different data and the results compared with the commercial modeling software show that the method is feasible and can improve the three-dimensional building modeling effect.