Automated system of recognition of road signs for ADAS systems

Abstract

Advanced Driver Assistance Systems (ADAS) require an environmental recognition function to inform the driver and for making management decisions. To solve the problems of pattern recognition, it is necessary to use one of the image classification methods. To date, there are quite a few similar algorithms that differ in the quality and speed of recognition. This paper presents a comparative analysis of classification algorithms in the problems of road signs recognition based on the GTSRB dataset. The results of the work showed that the most promising methods of classification in the problems of image recognition are the methods of reference vectors and k - nearest neighbors.