An Application of Omnidirectional Vision to Grid-based SLAM in Indoor Environments

Abstract

Abstract — In this work we study the use of an omnidirec-tional camera for the estimation of a consistent metric represen-tation of an indoor scenario. The proposed approach is based on Rao-Blackwellized Particle Filters and allows the robot to use a single vision sensor for estimating an occupancy grid map of the environment. The prediction phase of the filter is performed by means of an accurate visual odometry algorithm, whereas the update phase is based on floor segmentation, allowing to treat the omnidirectional camera in full similarity with laser-based approaches to SLAM. The technique is validated in simulation and through real experiments and it is shown to perform consistent map estimation while reducing the costs and the equipment of the robotic system. I.