Abstract

This review presents a comprehensive survey of recent developments in road safety enforcement systems, focusing on the integration of deep learning (DL) and computer vision (CV) technologies. Road safety is a major concern in urban environments, and this paper examines how DL and CV have been used to improve enforcement activities. The review includes a detailed analysis of state-of-the-art applications including real-time vehicle emission monitoring, vehicle identification and helmet detection. Various DL models such as Convolutional Neural Networks (CNNs) have been investigated for their performance in predicting air pollution levels and extracting valuable information such as vehicle number plates. Additionally, the paper explores the emerging area of helmet detection, shedding light on how computer vision is contributing in strengthening safety regulations. Challenges and opportunities in dataset acquisition, model training, and real-time processing are discussed. By integrating insights from a spectrum of research efforts, this review aims to provide a comprehensive understanding of the current landscape, inform researchers of key challenges, and stimulate future directions for the development of intelligent and effective road safety enforcement systems.