Road Sign Recognition System using Radial Basis Function Neural Network Architecture

Abstract

In this paper, road sign recognition system (RSR) is implemented with radial basis function (RBF) neural network architecture. The system consists of image pre-processing and algorithm of RBF function to recognition that road sign data set.RSR can be performed with dynamic and static road sign images. Road sign consists of various shapes according to their command. In this paper, the mostly use road signs are applied for training and testing data in the implementation of the system.RBF neural network is used in the recognition system because it is a supervised learning neural network and can have better approximation in pattern recognition than other systems. Road signs for warning for danger, forbidden and restriction, and command signs with various geometrical shapes are trained and tested. The result of the system is described by applying the recognition rate described from training and testing phase in RBF neural network.