Field Implementation and Testing of a Machine Vision Based Incident Detection System

by Panos G. Michalopoulos, Univ of Minnesota, Minneapolis, United States,
Richard D. Jacobson, Univ of Minnesota, Minneapolis, United States,
Craig A. Anderson, Univ of Minnesota, Minneapolis, United States,

Document Type: Proceeding Paper

Part of: Pacific Rim TransTech Conference: Volume I: Advanced Technologies


This paper contains test results of a video detection device installed in the field for over two years and the development and field deployment of an incident detection sysytems called IDEAS which is based on traffic information extracted by the video detection. In spite of the fact that for the time being the incident detection system must rely on single camera input, preliminary test results suggest an 80% detection rate and 3% false detection rate. The results also indicate detection of incidents within almost 2 miles from the detection zone even when the incidents occur beyond the field of view of the camera as well as in adjacent freeways.

Subject Headings: Cameras | Traffic accidents | Highways and roads | Urban development | Traffic surveillance | Field tests | Automatic identification systems | Computing in civil engineering

Services: Buy this book/Buy this article


Return to search