Options
Fahrzeugklassifikation und Schätzung streckenbezogener Verkehrsdaten mit Methoden der Mustererkennung
Citation Link: https://doi.org/10.15480/882.725
Other Titles
Vehicle Classification and Estimation of Section-Related Traffic Data Using Pattern Recognition Methods
Publikationstyp
Doctoral Thesis
Date Issued
1997
Sprache
German
Author(s)
Advisor
Referee
Title Granting Institution
Technische Universität Hamburg
Place of Title Granting Institution
Hamburg
Examination Date
1997-08-29
Institut
TORE-DOI
TORE-URI
Citation
Wohlers, Jens (1997): Fahrzeugklassifikation und Schätzung streckenbezogener Verkehrsdaten mit Methoden der Mustererkennung. Düsseldorf: VDI-Verl (Fortschritt-Berichte VDI : Reihe 12, Verkehrstechnik/Fahrzeugtechnik, 339).
Publisher
VDI-Verlag
Within the scope of this thesis efficient methods with pattern recognition systems are specified for the classification and recognition of vehicles with signals from induction loops. Different approaches to vehicle classification, including neural networks, are investigated and evaluated. Vehicle traffic recognition can be used to estimate route-related traffic data. Two different methods are developed and analyzed in comparison to a reference method. The first method uses cost functions on the basis of vehicle characteristics and local route forecasts, the second method takes into account the statistical links of the vehicles in the route section by modeling with Hidden Markov models. Finally, the use of the route-related traffic data in traffic management systems is described using the state-of-the-art estimation method using a model-based estimation method.
Subjects
Pattern Recognition
Traffic Monitoring
Traffic Control
Classification
DDC Class
620: Ingenieurwissenschaften