Real-time architecture for a highway vehicle detection system

Abstract

A three layer real-time architecture for a highway vehicle detection system is presented in the article. The Laser-Based Detection System (LBDS), focused on helping the Intelligent Transportation System (ITS), measures a key quantitative parameter of vehicles moving across a link of highway, namely, travel time. Travel time is based upon identifying and re-identifying vehicles at various points on the highway. This article provides a method to collect hard real-time signals from an active laser source in the LBDS and calculate real-time vehicle parameters using a standard computer. A method of seamlessness message exchange between a hard real-time kernel process, for real-time data acquisition, and a soft real-time application process, for computing and displaying, is given under the RTLinux environment. Experimental results from field tests show that the application of the real-time architecture to the LBDS provides vehicle parameter calculations deterministically. Experimental data are analyzed using the Video Sync program with both video images and graphical display.