Most vision-based robot positioning techniques rely on analytical formulations of the relationship between the robot pose and the projected image coordinates of several geometric features of the observed scene. This usually requires that several simple features such as points, lines or circles be visible in the image and be properly extracted.

In this paper, we present a method to compare images (scenes that the robot has learned) based on a fast and exact distance between histograms. In contrast to the methods described before, our method is faster and with less storage space do to the images do not need to be segmented and only a lossless description of the histograms are stored in the data base.