Autonomous Vehicles

Abstract

Modern traffic safety systems, particularly autonomous vehicles, rely heavily on visual data to recognize and categorize surrounding items. Such as people, road signs, street signs and traffic lights, and other autos in order for the vehicle to function properly. However, in adverse weather, the performance of object detection algorithms may be impaired. The primary goal of this thesis is to provide research on the most recent technique for automated recognition of text and symbols on the road surface in the form of painted road markings. Our objective is to investigate and evaluate the performance of object recognition systems that have been trained and tested using visual data gathered in both good and bad weather, and trained object recognition model to test it.