Topological map learning for a mobile robot in indoor environments

Abstract

A system that builds burrow-like topological maps and solves the localization of a mobile robot for indoor environments is presented. The approach uses visual features extracted from a pair of stereo images as landmarks. New landmarks are merged into the map and transient landmarks are removed from the map over time. A learning rule associated to each landmark is used to compute the landmark’s existence state. The position of the robot in the map is estimated by combining sensor readings, motion commands, and the current map state by means of an Extended Kalman Filter. The combination of neural network principles for map updating and Kalman filtering for position estimation allows for robust robot localization in indoor dynamic environments.