A Machine Learning Model For Detection and Recognition Of Traffic Signs

Abstract

Road accidents and deaths has been a major concern all over the world. There is never a day where there is no news about road accidents and deaths. The main cause for the frequency of road accidents to drastically increase is supposed to be the changing environmental conditions and non-realization of traffic sign recognition system. Now the world is moving towards an era of driver-less or autonomous vehicles, the paper proposes a novel approach to help vehicle drivers for safety driving concentrating mainly towards road traffic sign detection mechanism. The proposed system is a part of Intelligent Transport System(ITS) which can pave way as one of the dimension of a Smart City. The work here mainly focus on the challenges that could be tackled to avoid road accidents by introducing an alert system to detect the traffic signs while driving and also assist drivers with low vision. The work-done describes the use of Convolution Neural Network(CNN) classification model with inbuilt OpenCV functions to pre-process the images for traffic sign detection and recognition using Binarization and Region Of Interest(ROI). Overall, the proposed model has resulted an accuracy of 96.9% for training set and 98.8% for the validation set.