ISBN 978-3-8439-0637-1

978-3-8439-0637-1, Reihe Ingenieurwissenschaften

Katharina Quast
Automatic Object Detection and Tracking for Video-Based Traffic Surveillance

190 Seiten, Dissertation Universität Erlangen-Nürnberg (2012), Softcover, A5

Zusammenfassung / Abstract

Road and air traffic develops and increases rapidly. To manage the increasing traffic volume on streets and airport aprons, systems for video-based automatic traffic control are needed. Such systems require automatic detection and tracking of moving objects of interest. Since both tasks have to deal with similar challenges synergy effects can be exploited by detecting and tracking objects simultaneously.

In this dissertation, an overview over the tasks and challenges of automatic object detection and object tracking for video surveillance of traffic scenarios is given and several object detection and tracking algorithms are introduced and discussed. It is shown that using the proposed methods not only the tracking of the object position, but also a tracking of the object's size and rotation, respectively a tracking of the object contour is possible. Furthermore, a detection and a tracking method are combined in such a way that the detection method serves as an initialization for the tracking method. In this way, a system for automatic object detection and object tracking for video surveillance of traffic scenarios is achieved.