Highway Data Collection & Incident Management: Implementing a Video Image Processing System

Traffic Managers worldwide are faced with an increasing demand for state-of-the-art intelligent traffic systems: both for statistics purposes as for safety issues. Fast-developing urban regions have a need for information on traffic streams, to take well-founded decisions regarding new road infrastructure and changes to the existing infrastructure. Also on highways, traffic congestion and secondary accidents are now costing thousands of lives and billions of dollars every year. Therefore, traffic managers need an effective incident management system. Traditionally, loops and CCTV cameras provide ample information to direct traffic flows and assemble statistics. But their information is limited, and increasing traffic volume and complexity has created a need for more optimized systems; highly automatic incident management systems in particular. Today, video image processing systems handle both traffic data collection and automatic incident detection. This ITS technology has proven to be very reliable. Its incident detection shows a high detection rate, a short time to detect, fast incident verification and a low false alarm rate. This paper discusses the wide range of capabilities and some of the limitations of video image processing for highway incident detection as Traficon has experienced it over the past 20 years. Two main items are focused: nowadays video detection has proven to be a very reliable incident management tool. This ITS technology is the fastest system to detect incidents. Next to incident management, this paper will also focus on traffic data collection via Video Image Processing (VIP). A definition is given of traffic data quality and some new insights will demonstrate that video detection, when it is used correctly, offers great potential for highway data acquisition. A detailed case study of the highway incident detection and traffic data collection project in Atlanta will serve as an illustration.