Integrated System for Stereoscopic Cognitive Vision, Localization, Mapping, and Communication with a Mobile Service Robot

Abstract

This paper describes a stereo-vision-based mobile robot that can navigate and explore its environment autonomously and safely and simultaneously building a tridimensional virtual map of the environment. The control strategy is rule-based and the interaction with robot is done via Bluetooth. The stereoscopic vision allows the robot to recognize objects and to determine the distance to the analyzed objects. The robot is able to generate and simultaneously update a full colour 3D map of the environment that is being explored. The position and type of each detected and recognized object is marked in this 3D map. Furthermore, the robot will be able to use a gripper in order to collect detected objects and carry them to dedicated collecting bins, and so will be able to work in commercial waste cleanup applications. This application represents a successful integration of computers, control and communication techniques in mobile service robot control.