城市交通监控与建模系统：一种提高道路安全的IoT解决方案
Urban Traffic Monitoring and Modeling System: An IoT Solution for Enhancing Road Safety
摘要：
卡塔尔预计在2022年世界杯期间将有超过100万游客，这将构成重大挑战。大量的人可能会导致道路交通拥堵、车辆碰撞、受伤和死亡的增加。为了解决这个问题，可以使用自然主义驾驶员行为，它将收集和分析数据，以评估当前的卡塔尔交通系统，包括交通数据基础设施，安全规划以及工程实践和标准。在本文中，提出了一种基于物联网的解决方案，以促进卡塔尔的此类研究。来自驾驶员的不同数据点以不显眼的方式被收集和记录，诸如行程数据、GPS坐标、罗盘航向、最小、平均和最大速度以及他的驾驶行为，包括驾驶员的困倦程度。对这些数据点的分析将有助于预测撞车事故和改善道路基础设施，以减少此类事件。它还将用于驾驶员风险评估和检测极端的道路使用者行为。还提出了一个有助于可视化和管理这些数据的框架，以及一个基于深度学习的应用程序，该应用程序可以检测到昏昏欲睡的驾驶行为，准确率达到82%。
英文摘要：
Qatar expects more than a million visitors during the 2022 World Cup, which will pose significant challenges. The high number of people will likely cause a rise in road traffic congestion, vehicle crashes, injuries and deaths. To tackle this problem, Naturalistic Driver Behavior can be utilised which will collect and analyze data to estimate the current Qatar traffic system, including traffic data infrastructure, safety planning, and engineering practices and standards. In this paper, an IoT based solution to facilitate such a study in Qatar is proposed. Different data points from a driver are collected and recorded in an unobtrusive manner, such as trip data, GPS coordinates, compass heading, minimum, average, and maximum speed and his driving behavior, including driver's drowsiness level. Analysis of these data points will help in prediction of crashes and road infrastructure improvements to reduce such events. It will also be used for drivers risk assessment and to detect extreme road user behaviors. A framework that will help to visualize and manage this data is also proposed, along with a Deep Learning-based application that detects drowsy driving behavior that netted an 82 percent accuracy.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
