3-D MAP GENERATION BY A MOBILE ROBOT EQUIPPED WITH A LASER RANGE FINDER

Abstract

When an autonomous mobile robot acts in constructed environments, a map that has the information about obstacles is necessary for the robot. If a map is given, however, the mobile robot cannot start its action without the information of its present location on the map. Therefore, techniques of map generation and self-localization are important for mobile robots. In this paper, we propose map generation and self-localization methods by a mobile robot equipped with a laser range finder. The robot measures the distance between obstacles around it with a laser range finder at multiple positions, and generates a map by integrating these range information. The integration is realized by using a dead-reckoning (odometry) method and ICP (Iterative Closest Point) algorithm together. We also propose a method for estimating the present location of the robot on the generated map. Experimental results have shown the validity of the proposed method. 1.