This thesis addresses the problem of provident vehicle detection at night: The detection of oncoming vehicles by detecting its light artifacts. In particular, this means detecting a vehicle is oncoming even if it is not directly visible. For this purpose, a field study is conducted to understand human provident detection abilities and to record data. With the recorded data, a dataset is derived that can be used to analyze the performance of computer vision algorithms and to train machine learning approaches to detect vehicles providently. Furthermore, a complete computer vision algorithm is designed and trained on the derived dataset to detect light artifacts caused by the headlights of oncoming vehicles. The performance of this algorithm is compared with the measured performance of humans and an in-production computer vision system to demonstrate that the proposed system detects oncoming vehicles faster than humans and even faster than an in-production computer vision system. In addition, to show the usefulness of the developed algorithm, it is deployed in a test vehicle to use the detected light artifacts to control the glare-free high beam system proactively.
Schlagworte
Publikation durchsuchen
Bibliographische Angaben
Auflage
1/2023
Copyrightjahr
2023
ISBN-Online
978-3-95900-821-1
Verlag
TEWISS, Garbsen
Reihe
Berichte aus dem iPeG
Sprache
Deutsch
Seiten
141
Produkttyp
Monographie
Inhaltsverzeichnis
KapitelSeiten
Acknowledgments Kein Zugriff
Abstract Kein Zugriff
Kurzfassung Kein Zugriff
Contents Kein Zugriff
1.1 Scope and goal Kein Zugriff
1.2 Terminology Kein Zugriff
1.3 Overview Kein Zugriff
2.1 Driver reaction times at night Kein Zugriff
2.2 Object annotation methods Kein Zugriff
2.3 Datasets for vehicle detection at night Kein Zugriff
2.4 Camera-based vehicle detection at night Kein Zugriff
2.5 Motivation of the research questions Kein Zugriff
3.1 Design of the field study Kein Zugriff
3.2 Conducted field study Kein Zugriff
3.3 Results Kein Zugriff
3.4 Discussion Kein Zugriff
4.1 Proof-of-concept study: Detection of indirect light instances Kein Zugriff
4.2 Comparison of object annotation methods Kein Zugriff
4.3 Annotation methodology Kein Zugriff
4.4 The PVDN dataset Kein Zugriff
4.5 Generation of bounding boxes for light artifacts Kein Zugriff
4.6 Discussion Kein Zugriff
5.1 General design decisions Kein Zugriff
5.2 Light artifact detector Kein Zugriff
5.3 Distance estimator Kein Zugriff
5.4 Detection time analysis Kein Zugriff
5.5 Assessment of the algorithm in a test car: Provident glare-free high beam Kein Zugriff
5.6 Discussion Kein Zugriff
6 Summary and Outlook Kein Zugriff Seiten 123 - 126
Publications, Supervised Theses, and Online Material Kein Zugriff Seiten 127 - 130
