Title :
Accurate camera-based traffic sign localization
Author :
Welzel, Andre ; Auerswald, Andreas ; Wanielik, Gerd
Author_Institution :
Dept. of Commun. Eng., Tech. Univ. Chemnitz, Chemnitz, Germany
Abstract :
Traffic signs are the most important elements to control the right of way of intersections. In unknown areas, in dense city traffic, or under poor visibility conditions traffic signals are easily overlooked and ignored unintentionally. This paper provides two methods for absolute traffic sign localization for driver assistance systems. The system consists of a single color camera for the sign recognition, the sign localization algorithm and a GNSS receiver to obtain the vehicle position and heading. One localization method is a bearing-based approach, the second method is a relative localization approach. The accuracy of the two localization methods is evaluated on several road junctions using our concept vehicle Carai.
Keywords :
cameras; driver information systems; image recognition; road vehicles; satellite navigation; Carai; GNSS receiver; camera-based traffic sign localization; driver assistance systems; road junctions; sign recognition; vehicle heading; vehicle position; Accuracy; Cameras; Global Positioning System; Image color analysis; Receivers; Roads; Vehicles;
Conference_Titel :
Intelligent Transportation Systems (ITSC), 2014 IEEE 17th International Conference on
Conference_Location :
Qingdao
DOI :
10.1109/ITSC.2014.6957730