Deep Learning Approaches for Classification of Vehicles in Intelligent Transportation Systems

Abstract

Intelligent Transportation Systems (ITS) play a critical role in improving road safety, traffic flow, and efficiency, highlighting the significance of precisely categorizing vehicles for applications such as congestion management, traffic monitoring, and law enforcement. While traditional vehicle classification approaches have challenges and limitations, deep learning techniques have shown promising results in this area over the recent past. Deep learning approaches have been more precise and accurate in the classification of vehicles, which is a critical part of ITS. This work provides an in-depth analysis of recent literature utilizing deep-learning techniques for vehicle identification. Furthermore, deep learning techniques were expressed, in terms of taxonomy, performance matrices, future direction, and identification of research gaps in literature. The primary aim of this study is to present a comprehensive review of deep learning approaches in detecting vehicle types for ITS.  To the best of our knowledge, this is the only comprehensive study on the classification of vehicle identification using deep learning models in recent literature.  This study demonstrates the state-of-the-art deep learning-based ITS techniques for vehicle classification. It also provides valuable insight for future research in this domain.