Recognition for multiple robots linear formation

In the movement of multiple robots, the robot formation is one of the important attributes of colony activities. It has an important significance to recognize robot formations for evaluating the spot situation and making decisions. A new algorithm is proposed in this paper for recognition of robot linear formations. On the basis of pre-processing, (that is, each robot is segmented from the spot and expressed with center of mass which is binarized), the property is extracted by Hough transform. In the end, formation recognition is realized by dot matching according to the affine transform and particle swarm optimization (PSO). The simulation shows that the algorithm is effective.