Abstract
This paper introduces a map building method using a human-following robot. The map built in this study is a simplified hybrid map based on both of metric and topological information. In this paper, important points in the environments are detected by monitoring human movements while following by the robot. Especially, intersections are detected as important points and topological relationships among intersections are derived. The proposed system is constructed by RT components of following a target human, obtaining information for map building and etc. In this paper, details of the developed components and some experimental results are shown.
