Robot Map Similarity Evaluation for Non-identical Maps

When several robots explore the same environment, it is at some point necessary to merge their local maps into a global map to exploit the full potential of the robot team. Efficient multi-robot task coordination is virtually impossible without a common interpretation of the environment. If there is no information available about the relative positioning of the robots or if this information is uncertain, the evaluation of the map merging result is a vital step in the creation of a global map. The proposed map similarity evaluation offers a way to successfully evaluate the similarity of occupancy grid maps that are not locally identical.