Traffic Surveillance and Detection Technology Development: New Traffic Sensor Technology

Problem

Transportation management agencies are faced with having to manage ever increasing volumes of traffic on roadways. New developments in traffic surveillance and detection technology are aimed at improving the accuracy and reliability of real-time traffic data so that traffic conditions can be observed more accurately.

Method

A video system was used to obtain visual traffic data. A computer hardware and software system was developed to process the video data. The goal of the processing was to identify, group, and track vehicles using various

Findings

The method of vehicle detection used is known as "feature based tracking." This method uses a computer to process the video data to identify features of the vehicles, such as corners, the arrangement of lights (at night), height, width, and so on.

This method was found to be very accurate in determining traffic speed because even false detections will move with the general traffic speed. A vehicle detection that doesn't move within a percentage of the other detected vehicles is eliminated from the data set, so false detections don't significantly alter the speed data.