Laser-based people tracking by multiple mobile robots

This paper presents laser-based people tracking by a group of mobile robots located near each other. Each robot finds moving people in its own laser scan images using an occupancy-grid-based method. It then tracks the detected people via Kalman filter and Global-nearest-neighbor-based data association. Tracking data are broadcasted to multiple robots through intercommunication and are combined using the Covariance intersection method. In our tracking method, all the robots share tracking data with each other; hence, they can always recognize people who are invisible to any other robot. The method is validated by the results of an experiment in which two walking people are tracked by three mobile robots.