A real-time system for monitoring pedestrians

In this research, we propose a novel system of monitoring pedestrians and tracking their trajectories in a wide and open area, such as shopping mall, exhibition hall, railway station and so on. A network of single-row laser range scanners is exploited, which are set on the floor doing horizontal scan at an elevation of about 15cm above the floor. A pedestrian's walking model on the horizontal cross section of ground level is defined and a tracking algorithm is developed to detect and trace the rhythmic swing feet. On the other hand, video cameras are integrated with the network of laser scanners, which focus on some spots inside laser scanner's measurement coverage, and get image data for pedestrian's trajectories. An experiment for real-timely monitoring visitor's flow was conducted in a large exhibition, which had totally 30,000 visitors during three exhibition days. Four LMS200 (by SICK) and a video camera were located near a crossing point, covering a floor area of about 30m*30m and 5m*6m respectively. Visitors inside and nearby video camera's vision field were tracked in a real-time mode. Their trajectories as well as the laser points on their feet were back projected onto the video images on the air. On the other hand, visitor's flow nearby the crossing point, and its change along with time is examined by analyzing the motion behavior of each visitor in an off-line mode.