Laser-based People Tracking

— In this paper, we describe a method for real-time tracking of objects with multiple laser range-finders covering a workspace in a parallel and distributed fashion. Tracking people is a popular problem in machine vision. Here we adapt the methods used in vision to pla-nar laser scanners. We group range measurements into entities like blobs and objects. With the help of a Kalman Filter (KF), the tracker smooths object paths and estimates its path even when the underlying object is occluded from all the lasers. We finally evaluate the tracker's performance through four experiments.