A System for the Automat~c Sumeillance of TrafRcSituations

Knowledge of the actual traffic condition in a road section is the basis for an efficient performance of traffic management and control systems, automatic incident detection, route guidance systems, and so forth. Present traffic data acquisition systems are usually based on the collection of local data, such as passing vehicles’ speed or volume, which are picked up at certain points along the road. Using these local data it is possible to estimate the traffic conditions in the road section between the measuring points. A high accuracy of such estimations requires a relatively high density of measuring points. A more reliable basis for the abovementioned aims can be obtained, however, if it is possible to measure sectionrelated data such as journey time and traffic density, thus allowing a deeper insight into the actual traffic condition. In this article such a system is introduced; the system uses inductive loop detectors and pattern recognition principles tore-identify vehicles after they have passed a specific road section. Correlation of platoons of vehicles from the entry of a road section with those from the exit delivers the above-mentioned traffic data. This can be achieved with a relatively low number of data acquisition points in comparison with common systems. The result may be displayed in a map of a road network that is dynamically updated with information of the actual traffic situation. This map can be stored in a traffic monitoring center, where it can be accessed by different users (individuals, authorities, etc.) for their specific purposes.