Umfelderkennung und -modellierung von Baustellen für Fahrerassistenzsysteme
FacultiesFakultät für Ingenieurwissenschaften und Informatik
LicenseStandard (ohne Print-On-Demand)
Increasing traffic and freight volume on high- and freeways demands constant maintenance and repair of the infrastructure. Therefore, the number of road work zones are exptected to further increase. A future assistance system may support the drivers in guiding their vehicle through the road work zone. A precise knowledge of the environment is crucial for such new driver assistance systems. Therefore, this work aims at the perception of road work zones on high- and freeways. A laser scanner mounted on the front of the vehicle detects traffic pylons and beacons with high detection rate and precision. The recognition of the correct lane is a great challange for the perception system, as not only the road markings of the actual work zone are present, but also the original but now invalid markings. This work provides a system which enables the detection of the correct lane with data of a grey scale camera and probabilistic methods. Assistance systems, which interfere in lateral direction of the vehicle"s control, pose especially high demands on the robustness and reliability of the environmental perception. The work proposes the use of an accurate digital road work map which augments the information of the onboard sensors. The creation of the road work map is done with a special mapping vehicle. The exact assignment of the highly accurate road work map to the current environment model is enabled with the data of the front facing laser scanner. In case of poor measurment conditions due to weather or local occlusion, the information from the road work map helps to provide an entire model of the environment. The methods developed in this work allow the detection of the vehicle"s environment in the precision range of 10 cm. The association of the road work map is possible with an accuracy of approximately 25 cm. That way, a precise environmental model for the lateral control of a vehicle within the narrowed lanes at road work zones is achieved.
Subject HeadingsBaustelle [GND]
Driver assistance systems [LCSH]
Road work zones [LCSH]
Video cameras [LCSH]