Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Advantages and limitations of a dual approach in video-based traffic data acquisition

: Grimm, Jan

Albrecht, Thomas (Ed.):
3rd International Conference on Models and Technologies for Intelligent Transportation Systems, MT-ITS 2013. Proceedings : Dresden, December 2nd-4th, 2013
Dresden: TUDpress, 2013 (Verkehrstelematik 3)
ISBN: 978-3-944331-34-8
International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS) <3, 2013, Dresden>
Fraunhofer IVI ()
video detection; Videodetektion; traffic data acquisition; incident detection; image processing; Bildverarbeitung; quality evaluation

For traffic data acquisition and automatic incident detection, video detection has become a reasonable alternative to established detection technologies, especially due to a potential for more flexible application and lower installation costs. However, achieving a high level of quality and reliability in video-based traffic data acquisition is still a challenge in research. The objective of this paper is to present the general approach as well as advantages, practical limitations and implementation issues of the video-based algorithms developed by the Fraunhofer Institute for Transportation and Infrastructure Systems (Fraunhofer IVI). The traffic data acquisition algorithms are based on a dual approach, combining the tripwire method for vehicle counting and classification, and the tracking method for speed measurement. At first, the relevant use cases and requirements for video-based traffic data acquisition are identified. Then, the Fraunhofer video detection algorithms are described, and advantages as well as limitations found during field operational tests and in practical implementations are summarised. In addition, approaches for an online and offline quality evaluation are discussed, considering both the aspect of data quality and system performance. Finally, an outlook is given on what still needs to be done to ensure a sufficient level of quality even under adverse conditions and to convince traffic engineers and decision makers of the benefits of video detection compared to conventional traffic data acquisition techniques.