EmerT – Supporting traffic parameter estimation from low cost and low resolution uncalibrated web cameras

Abstract

This paper describes the EmerT portal (Emergency mobility of rescue forces and regular Traffic), an extended version of the DELPHI web tool. EmerT is a web-based decision support application for real time traffic situation, prognosis and traffic simulation appropriate for exceptional situation like disasters, major events, incidents or emergencies. One of the main features in this system is to provide traffic state estimation and forecasting with the use of the mesoscopic traffic simulation. Current traffic data are needed for accurate simulation results. Typical inputs data are from ground-based sensor like induction loops. One further possibility represents the use of traffic web cameras, in which images are most freely available. Web cameras have mostly a low resolution and a low frame rate. These underlying limitations as well as the integration in the simulation are the big challenges for an efficient use as data input for traffic simulation. This paper aims to present the process for extraction of traffic data from traffic web cameras and the integration of this approach in the EmerT portal.