Development of a self-calibrating routine for camera-based object pose estimation in the context of platooning
- Subject:Deep Learning, Object Detection, Perception
- Type:Bachelorarbeit
- Date:ab 03 / 2024
- Tutor:
Development of a self-calibrating routine for camera-based object pose estimation in the context of platooning
Context
Platooning is becoming increasingly important in local public transport. By linking several buses to form a platoon, it is possible to react flexibly to passenger fluctuations. In this context, the precise detection of the relative position of the bus in front using redundant sensors is of crucial importance. Detection by camera is marker-based and this leads to fast and robust detection. However, this assumes that the relative positioning of the markers to each other and relative to the bus is known. In practice, it can be assumed that the markers can slip or are not positioned correctly. Self-calibration is therefore necessary to determine the position of the markers relative to each other and relative to the bus.
Goals
- Conceptualization of a self-calibration routine for marker-based pose estimation
- Prototypical implementation and evaluation of the concept
Requirements
- Programming skills in Python and Ros2
- Basic knowledge in the field of machine learning
- Independent and solution-oriented way of working
- Business fluent in written and spoken English