Relocalisation by Partial Map Matching

The autonomous operation of an intelligent service robot in practical applications requires that the robot builds up a map of the environment by itself. A prerequisite for building large scale consistent maps is that the robot is able to recognise previously mapped areas and relocalise within these areas. The recognition is based on constructing partial maps of geometric landmarks which are then compared to yield the optimal correspondence between these landmarks. For each landmark a signature is constructed which contains additional information about its immediate environment and its non-geometric properties. It is ensured that the signatures are robust with respect to missing landmarks, rotation and translation of landmarks and varying landmark lengths. Both simulation and experiments on real robots have shown that the approach is capable of recognising previously mapped areas robustly in real-timel.