TY -的A2 -谢,库恩AU - Skrabanek,帕维尔AU - Dolezel,切赫盟——Nemec格里格拉盟——Stursa杜米尼克PY - 2020 DA - 2020/10/14 TI -人检测垂直放置的单眼相机SP - 8843113六世- 2020 AB -计数的乘客进入和退出的交通工具是客流监测系统的基本功能。乘客的确切数字是重要的在公共交通等领域监测、客流预测、交通规划和交通车辆负载监控。允许大规模客流监测系统,利用他们的成本要低。总体价格主要是由价格的传感器和处理单元使用,我们建议利用可见光谱相机和数据处理算法时间复杂度低,以确保最终产品的低价格。保证乘客的匿名性,我们建议正交扫描的一个场景。的精确计算是贴切地受到乘客识别的精度,我们专注于一个适当的识别方法的发展。我们提出两种相反的方法,可用于交通工具的乘客识别,没有入口的步骤,或者分裂地板水平。第一种方法是利用一个适当的卷积神经网络(事先),目前在计算机视觉的方法。第二种方法是利用直方图的梯度(猪)的特性结合支持向量机分类器。这种方法是一种经典方法的代表。 We study both approaches in terms of practical applications, where real-time processing of data is one of the basic assumptions. Specifically, we examine classification performance and time complexity of the approaches for various topologies and settings, respectively. For this purpose, we form and make publicly available a large-scale, class-balanced dataset of labelled RGB images. We demonstrate that, compared to ConvNets, the HOG-based passenger recognition is more suitable for practical applications. For an appropriate setting, it defeats the ConvNets in terms of time complexity while keeping excellent classification performance. To allow verification of theoretical findings, we construct an engineering prototype of the system. SN - 0197-6729 UR - https://doi.org/10.1155/2020/8843113 DO - 10.1155/2020/8843113 JF - Journal of Advanced Transportation PB - Hindawi KW - ER -