Reconstruction of the electronic horizon of a navigation system on camera for improvement of a driver assistance system

OData support
Dr. Harmati István
Department of Control Engineering and Information Technology

In Europe 35000 people die in car accidents and 1.9 million people get injured in each year. Using driver assistance systems helps to decrease the number of accidents. One of these systems, the traffic sign recognition displays the speed limit applicable to the driven road segment as well as further information based on the recognized signs and road markings, e.g. lanes. It is a problem that camera-based road sign recognition systems do not recognize some of the speed limit related signs and in many places the signs are missing. The fusion of the video and navigation signals significantly improves the performance of the system. The ADASIS protocol sends the actual map data (electronic horizon) calculated by the navigation control unit to the road sign recognition system running on the camera via CAN communication, which is standard in automotive industry. The goal of this thesis is the reconstruction of the electronic horizon on the camera.

While working on this thesis I got familiar with the communication system of vehicles (CAN), and the ADASIS protocol, which is used in driver assistance systems. The most important tasks are the design of the appropriate data structure, and the implementation of navigation database using information available in ADASIS protocol. The implemented software updates the horizon-database as the vehicle moves ahead. The algorithm deletes passed and adds approaching sections of the road. I tested the operation of the software during implementation with test messages. I integrated the complete software into a simulated version of an existing automotive project, in order to test the program simulating real environment using real data. I implemented the software in C++ language.


Please sign in to download the files of this thesis.