A symbol identifier based recognition and relative positioning approach suitable for multi-robot systems

Abstract

In this paper, a symbol identifier based recognition and relative positioning approach suitable for multi-robot systems is proposed. The symbol identifier is composed of central area and peripheral area, and there exists radial spokes in central area. The recognition approach utilizes some features including luminance feature of center point, luminance difference between center point and its ambient region, the number and distribution of spokes as well as the shape of peripheral area to filter the points in the image. Finally, the resulting pixel points set to characterize the center points of symbol identifiers is generated. On this basis, the positions of these center points relative to the camera are then calculated. The proposed approach is verified by the experiments. © Springer-Verlag Berlin Heidelberg 2012.