Hierarchical topometric representation of 3D robotic maps

In this paper, we propose a method for generating hierarchical, volumetric topological map from 3D point clouds. There are three basic hierarchical levels in our map: $storey - region volume$. The advantages of reflected both input and output. terms input, accept multi-storey clouds building structures with sloping roofs or ceilings. output, can generate results metric information different dimensionality, that suitable robotics applications. algorithm generates the representation by $volumes$ voxel occupancy map. We then add $passage$s (connections between $volumes$), combine small into big $region$ use 2D segmentation better representation. evaluate on several freely available datasets. experiments highlight approach.