Localization of a mobile robot by matching 3D-laser-range-images and predicted sensor images

Abstract

This article describes a system for localization of a mobile robot. Position and orientation of the vehicle are determined by extracting vertical surfaces from a 3D-laser-range-image, which are matched with surfaces predicted from an environmental model. The use of predicted sensor images accelerates the segmentation and matching processes by focussing on regions of interest. The three main components 3D-laser-range-camera, environmental model and the localization algorithm are presented in detail. Quantitative results for localization in laboratory and manufacturing environments proved accuracy, robustness and real-time capability of this localization system.