Over the last decades, a special attention has been given to Intelligent Transportation Systems. The increase in the number of vehicles circulating on the roads and the subsequent increase of the number of roads led to the necessity of creating active systems of roadway safety. In this project, a specific roadway safety system will be addressed, focusing on the monitoring of drivers' behaviors and the traffic information extraction through the analysis of the CCTV signal.

Some vision-based solutions are proposed to help human operators in the hard task of detecting abnormal situations in highways and manage the traffic flow. This ATS system is divided into two main parts:

  • Automatic Incident Detection system
  • The aim of the AID system is to detect and track potentially anomalous traffic events in highways. By anomalous events it is meant the detection of vehicles that stop on highways or on hard shoulders, vehicles driving on the wrong direction, vehicles constantly switching between lanes, low speed vehicles, and also high speed vehicles.

  • Automatic Traffic Monitoring system
  • The aim of the ATM system is to provide some useful statistical information about the traffic. Some of the statistical features to be retained are: the number of vehicles circulating on each of the lanes, vehicles' average speed, road occupancy rate, etc.

The Intelligent Traffic Surveillance Project is being developed at Computer And Robot Vision Lab, Institute of Systems and Robotics [+], University of Coimbra.

The presented solutions are part of an Automatic Traffic Surveillance system that is being developed to be applied in the surveillance system of Brisa's highways [+].

Wrong Way Driver Detection

Vehicle circulating in the wrong direction detected by AID System on a Portuguese highway.