Performance Assessment of a People Tracker for Social Robots

Service robots are becoming more and more pervasive in natural environments where human beings actually carry out daily duties. One of the main enabler in these scenario is the capability of such a system to understand the positions and the motions of people moving in its surroundings. In this paper, we will investigate the effectiveness of a people tracking system that uses a stereo camera as the sensor and then use the gathered data in an estimation algorithm to continuously track their position in the environment. Experiments in a laboratory are carried out to characterise and evaluate the performance of such a system and establish its suitability for an application on an actual service robot.