A Survey on Traffic Sign Classification using Artificial Intelligence Techniques

Recently, Intelligent Transportation Systems (ITS) have become a trend in scientific research, considered an essential and indispensable element to stimulate innovation and ensure safety on the road. In order to improve the reliability of autonomous vehicles, this survey presents current innovations in computer vision and Artificial Intelligence (AI) to classify traffic sign. This work examines both Machine Learning (ML) and Deep Learning (DL) techniques, addressing issues related to damaged panels, lighting changes, and atmospheric variations that impact the view of traffic sign. Furthermore, we have presented the evaluation parameters as well as some important datasets; in particular, German Traffic Sign Recognition Benchmark (GTSRB), Belgium Traffic Sign Classification (BTSC), and Chinese Traffic Sign Database (CTSD). Therefore, scientists and researchers in ITS will find great value in the information presented. After that, we have highlighted some existing limitations and made recommended approaches in this field.