Traffic Monitoring using a Three-Dimensional Object Tracking Approach

An orthographic projections-based algorithm was investigated for the traffic monitoring application of tracking a truck moving across a bridge through a sequence of video image frames. The two-dimensional position of a specific marker on the truck for each video frame is found using color thresholding-based segmentation. A coordinate transformation process provides the real-world, three-dimensional location of the marker position for tracking the truck. The procedure is experimentally illustrated with a field bridge. This paper presents the experimental bridge setup, the marker identification technique, the orthographic coordinate transformation algorithm, and web-accessible laboratory exercises developed for image processing and related curricula