3-D-environment reconstruction for mobile robots using fast-SLAM and feature extraction

This paper proposes an algorithm that can be used to reconstruct a 3-D environment on a mobile robot. As sensors, color and time-of-flight cameras are used. 2-D features are extracted from color images and assigned 3-D coordinates. Those are the input for a modified fastSLAM algorithm that is capable of rendering environment maps for small environments online in order to execute manipulation tasks. The method is evaluated on the service robot Care-O-bot® 3.