Video-based Vehicle Signature Analysis and Tracking System Phase 2: Algorithm Development and Preliminary Testing

This report describes the results of the PATH/Caltrans-funded project Video-Base Signature Analysis and Tracking (V2SAT) System, Phase 2 Algorithm Development and Preliminary Testing. The V2SATSystem was conceived in 1995 by Loragen Systems, of San Luis Obispo, California, as a means for non-intrusively tracking individual vehicles on freeways for data collection purposes.  The concept involves the use of computer vision methods to make simple optical measurements on digitized real-time images of each vehicle on the freeway. A conventional color video camera serves as the primary sensor. Detection modules are placed directly above traffic lanes on an overcrossing or similar support structure, with one detector for each lane. For each passing vehicle, a numeric Video Signature Vector (VSV) is generated and transmitted by the detection module to a central correlation computer, via a low-power wireless network. The correlation computer continuously receives VSV’s asynchronously transmitted by all detection modules, and attempts to match VSV’s to re-identify vehicles at each detectorized site, in order to determine the progress of each vehicle through the freeway network.