City-scale vehicle tracking and traffic flow estimation using low frame-rate traffic cameras
City-scale vehicle tracking and traffic flow estimation using low frame-rate traffic cameras
复制标题
DOI：
10.1145/3341162.3349336
复制
发表时间：
2019-09
期刊：
影响因子：
--
通讯作者：
Peter Wei;Haocong Shi;Jiaying Yang;J. Qian;Yinan Ji;Xiaofan Jiang
中科院分区：
文献类型：
--
作者：
Peter Wei;Haocong Shi;Jiaying Yang;J. Qian;Yinan Ji;Xiaofan Jiang
车流估计在智能城市和交通领域有着广泛的应用前景。许多城市都有现有的摄像机网络来播放图像馈送；然而，分辨率和帧速率太低，以至于现有的计算机视觉算法无法准确估计流量。在这项工作中，我们提出了一种用于车辆跟踪的计算机视觉和深度学习框架。我们展示了一种新的跟踪流水线，该流水线能够在低分辨率和帧速率限制的一系列环境中实现准确的流量估计。我们证明了我们的系统能够以1赫兹或更低的帧速率跟踪纽约市交通摄像头视频中的车辆，并产生比流行的开源跟踪框架更高的交通流精度。
Vehicle flow estimation has many potential smart cities and transportation applications. Many cities have existing camera networks which broadcast image feeds; however, the resolution and frame-rate are too low for existing computer vision algorithms to accurately estimate flow. In this work, we present a computer vision and deep learning framework for vehicle tracking. We demonstrate a novel tracking pipeline which enables accurate flow estimates in a range of environments under low resolution and frame-rate constraints. We demonstrate that our system is able to track vehicles in New York City's traffic camera video feeds at 1 Hz or lower frame-rate, and produces higher traffic flow accuracy than popular open source tracking frameworks.