ABSTRACT

Over the past few years, real-time visualisation of pedestrian dynamics has become more crucial to successfully organise and monitor open-crowded events. However, the process of collecting, efficiently handling and visualising a large volume of pedestrians' dynamic data in real time is challenging. This challenge becomes even more pronounced when pedestrians move in large-size, high-density, open and complex environments. In this article, we propose an efficient and accurate approach to acquire, process and visualise pedestrians' dynamic behaviour in real time. Our goal in this context is to produce GPS-based heat maps that assist event organisers as well as visitors in dynamically finding crowded spots using their smartphone devices. To validate our proposal, we have developed a prototype system for experimentally evaluating the quality of the proposed solution using real-world and simulation-based experimental datasets. The first phase of experiments was conducted in an open area with 37,000 square meters in Palestine. In the second phase, we have carried out a simulation for 5000 pedestrians to quantify the level of efficiency of the proposed system. We have utilised PHP scripting language to generate a larger-scale sample of randomly moving pedestrians across the same open area. A comparison with two well-known Web-based spatial data visualisation systems was conducted in the third phase. Findings indicate that the proposed approach can collect pedestrian's GPS-based trajectory information within 4 m horizontal accuracy in real time. The system demonstrated high efficiency in processing, storing, retrieving and visualising pedestrians' motion data (in the form of heat maps) in real time.

摘要

在过去的几年里，行人动态的实时可视化对于成功组织和监控开放拥挤的事件变得更加重要。然而，实时收集、有效处理和可视化大量行人动态数据的过程具有挑战性。当行人在大尺寸、高密度、开放和复杂的环境中移动时，这一挑战变得更加明显。在本文中，我们提出了一种有效且准确的方法来实时获取、处理和可视化行人的动态行为。在这种情况下，我们的目标是制作基于 GPS 的热图，帮助活动组织者和游客使用他们的智能手机设备动态地寻找拥挤的地点。为了验证我们的提议，我们开发了一个原型系统，用于使用真实世界和基于模拟的实验数据集对所提出的解决方案的质量进行实验评估。第一阶段实验在巴勒斯坦3.7万平方米的空地上进行。在第二阶段，我们对 5000 名行人进行了模拟，以量化所提出系统的效率水平。我们利用 PHP 脚本语言生成了一个更大规模的随机移动行人穿过同一开放区域的样本。在第三阶段与两个著名的基于 Web 的空间数据可视化系统进行了比较。结果表明，所提出的方法可以实时收集行人基于 GPS 的轨迹信息，水平精度在 4 m 以内。该系统在处理过程中表现出高效率，