Traffic Sign Classification Using Deep Neural Network
Merin Annie Vincent, K R Vidya, Santhosh P Mathew
Abstract
Technology has transformed the way we live. Artificial Intelligence has positively impacted every field and is making life easier for us. Technology aided traffic system is a remarkable breakthrough in the automotive industry. Automotive Industry is making giant leaps with the advancement of autonomous cars. Autonomous cars represent a certitude for the near future and it is expected to be safe and hassle-free. Traffic Sign Classification is one of the most important tasks of autonomous cars. Our research aims to do this classification efficiently to avoid accidents and to foster the credibility of autonomous cars. Otherwise, this could jeopardize traffic ecology. This research led to the implementation of a Traffic Sign (TS) CNN model for traffic sign classification, with the aid of the GTSRB dataset. An accuracy of 98.44% was achieved on test data.