Traffic analysis problem has been deeply studied in the last years, in particular different techniques has been investigated to extract information concerning traffic flow. The problem needs 24/7 working algorithms with different weather conditions and different scenarios. Non invasive technologies have a lead role in this field since their installations requirements are very low, the system presented in this paper is vision based, some possible installation patterns will be detailed. The algorithm uses a features tracking approach that can manage traffic jam, queue, day/night light conditions. The system proposed is currently installed on a real site and the results and the statistics are up to 2,400 hours of video data.
Vision-based traffic analyzer
Verkehrsanalyse mit Hilfe von optischen Sensoren
2010
6 Seiten, 9 Bilder, 7 Quellen
(nicht paginiert)
Aufsatz (Konferenz)
Datenträger
Englisch
German traffic analyzer uses doppler principle
Engineering Index Backfile | 1960
|Traffic Accidents Analyzer Using Big Data
IEEE | 2015
|