RG Mapping : Building Object-Oriented Representations of Structured Human Environments

We present a new approach to mapping of indoor environments, where the environment structure in terms of regions and gateways is automatically extracted, while the robot explores. Objects, both in 2D and 3D, are modelled explicitly in those maps and allow for robust localization. We refer to those maps as object-oriented environment representations or region and gateway maps. RG mapping is capable of acquiring very compact, structured, and semantically annotated maps. We show that those maps can be built online and that they are extremely useful in plan-based control of autonomous robots as well as for robot-human interaction.