Multi sensor data fusion algorithm development for driver assistance systems for commercial vehicles

OData support
Supervisor:
Kovács Viktor
Department of Automation and Applied Informatics

The number of modern automotive vehicles with automated functions is increasing. In the near future, the autonomous driving is a realizable target. The sensing of surroundings of a vehicle is an essential part of these systems. If the perception layer fails, it occurs serious mistakes in the decision logics. Therefore, an accurate processing layer of input sensors is an important part of automotive developments.

The purpose of this thesis is the presentation of a sensor data fusion method. The fusion system’s goal is the generating of a stable, credible list of the relevant objects in the surrounding area from the information of input sensors.

The selection of a method, software design, prototype implementation and the installation to a truck will be presented.

In this thesis, the modularity of inputs and the internal structure plays an important role. The flexibility and the source efficiency is a very important factor too with taking into account the safety and robustness.

The experiences, which given by the implementation and testing, provide an image from the requirement and possibilities of future development

Downloads

Please sign in to download the files of this thesis.