Recognition of people's positioning by cooperative mobile robots for human groups steering

In this research we attempt to build a multi-cooperative mobile robot system able to perform the process of steering people in indoors. When a target-group is integrated by several persons, the steering task is more difficult to accomplish, specially when only a robot is intended to be used. However, the problem can be overcome if we consider that a cooperative mobile robot system's performance is higher for steering multiple humans than only one, and will greatly improve the performance of the tasks. In this paper, we detail a novel method for multiple human localization which has been developed by using multi homogeneous mobile robots, equipped with trinocular stereo vision. As a first step of this research the method includes spatial noise data filtering, multi sensor data fusion and clustering based segmentation. In addition, some experimental results are shown to verify the feasibility of the method.