Traffic Intersection Detection Using Floating Car Data

The autonomous vehicle navigation system and the intelligent traffic system need real-time routing planning in traffic engineering and management. Urban road intersections play bottlenecks of urban roads, contributing a great deal to the loss of travel time due to traffic interference, management and control. The detection of the traffic intersection is important for obtaining real-time road regulation information. This paper proposes a method to detect the traffic intersection based on floating car data that is typically timestamped geolocalization, direction and speed data directly collected by moving vehicles. The angle of the direction difference can be used as the features for detecting the traffic intersection. The identification of the traffic intersections algorithm is developed based on the data density clustering algorithm. The floating car data is collected by about 15,000 cars using an onboard with the Global Positioning System receiver and sensors that providing positions (latitude, longitude) instantaneous speed, drive direction and time etc. The assessment results show the approach has excellent performance with high accuracy for traffic intersection detection.