3D Cadaster Creation with CAD Generalization and Graph Mining

It is difficult and costly to survey the 3D boundary points for the cadaster management. In this paper, a method to extract the boundaries of a 3D cadaster is proposed based on the CAD map generalization. Furthermore, a graph mining algorithm is studied to automatically generate the 3D cadaster from the extracted boundary points. We improve the graph cycle base detection algorithm to generate the semantic information of 3D cadaster. The experimental results suggest that the proposed method can correctly detect the boundary points and create the semantic 3D cadaster map for complex buildings.