Location-Based Tracking of Moving Obstacles from a Mobile Robot

Tracking moving obstacles from a moving platform is a useful skill for the coming generation of mobile robot. The methods used in existing moving objects tracking that operated from fixed platform and fixed background, are not applicable. Instead, we propose a system that detects unexpected moving obstacles that appear in the path of mobile robot. The system is designed to track and determine the location of the obstacles with the use of calibrated stereo camera mounted on the top of robot. This method is implemented in the domain of moving objects tracking from a mobile platform using a novel combination of motion segmentation, using spatiotemporal segmentation from the left image sequence, and the 3-D information which is calculated from a stereo image sequence. Position estimation of the moving objects is performed by using Kalman filter on the plan-view map. The system was implemented and tested on an Omni-directional robot at our laboratory. The results show that we are able to reliably detect and track moving objects in natural environments approximately 3×8 meters in front of mobile robot. Our system is able to detect the moving objects at 13 Hz for a 320×240 pixels stereo image sequence using a standard laptop computer.