Advancements in Deep Learning for Automatic Moving Vehicle Detection in Intelligent Traffic Surveillance: A Comprehensive Review and Statistical Analysis

The traffic monitoring system plays an important role in Intelligent Transportation Systems (ITS), representing a substantial investment by transportation agencies. It serves to gather and analyze traffic data, optimizing road usage, enhancing transportation safety and shaping future transportation strategies. Recent strides in MEMS, deep learning and wireless communication have spurred the development of innovative traffic monitoring systems. This paper investigates state-of-the-art schemes, focusing primarily on vehicle categorization, discussing their organization, research challenges, technical complexities, hardware/software frameworks, deployment insights and performance. Additionally, crucial open issues and future research avenues are explored, aiming to offer valuable insights to academia, industry and government entities for effectively choosing the technologies suited to the traffic monitoring needs.