• search hit 1 of 6
Back to Result List

Automotive Camera Modeling and Integration with Standardized Interfaces

  • Various technological challenges still prevail when it comes to the development and validation process of advanced driver-assistance systems and autonomous functions. This dissertation aims to introduce new approaches and methods that still adhere to the process flow of the V-cycle, and to address some of the existing development challenges. To satisfy these approaches, a physical sensor model for an advanced driver assistance camera has been created, and in the context of the proposed modeling approach, pertinent levels of abstraction and relevant optical and image sensor effects are identified. Additionally, applied concepts and employed methods for the camera model parametrization are presented as well. Moreover, this dissertation illustrates the effectiveness of hybrid-development and test strategies. For the development part, the usage of a camera model in training image-based neural network algorithms is illustrated. Besides, concerning the test part, several classical computer vision algorithms and image-based neural networkVarious technological challenges still prevail when it comes to the development and validation process of advanced driver-assistance systems and autonomous functions. This dissertation aims to introduce new approaches and methods that still adhere to the process flow of the V-cycle, and to address some of the existing development challenges. To satisfy these approaches, a physical sensor model for an advanced driver assistance camera has been created, and in the context of the proposed modeling approach, pertinent levels of abstraction and relevant optical and image sensor effects are identified. Additionally, applied concepts and employed methods for the camera model parametrization are presented as well. Moreover, this dissertation illustrates the effectiveness of hybrid-development and test strategies. For the development part, the usage of a camera model in training image-based neural network algorithms is illustrated. Besides, concerning the test part, several classical computer vision algorithms and image-based neural network algorithms are tested and evaluated in a dedicated but generic/modular framework. Finally, co-simulation frameworks are presented for the integration and coupling process of various software components (sensor models, highly automated functions and simulation software) with standardized interfaces like the functional mockup interface and the open simulation interface.show moreshow less

Export metadata

Additional Services

Search Google Scholar
Metadaten
Author:Kmeid Saad
URL / DOI:https://diglib.tugraz.at/download.php?id=5f90184e706db&location=browse
Referee:Martin Horn, Daniel Watzenig, Kazuo Ohzeki
Advisor:Martin Horn, Stefan-Alexander Schneider
Document Type:Doctoral Thesis
Language:German
Date of Publication (online):2019/10/24
Year of first Publication:2019
Volume:2019
Number of pages:161
Institutes:Fakultät Elektrotechnik
Publication reviewed:begutachtet
Release Date:2021/03/01
Verstanden ✔
Diese Webseite verwendet technisch erforderliche Session-Cookies. Durch die weitere Nutzung der Webseite stimmen Sie diesem zu. Unsere Datenschutzerklärung finden Sie hier.