A Fast Vehicle Counting and Traffic Volume Estimation Method Based on Convolutional Neural Network
A Fast Vehicle Counting and Traffic Volume Estimation Method Based on Convolutional Neural Network
复制标题
DOI：
10.1109/access.2021.3124675
复制
发表时间：
2021
期刊：
影响因子：
3.9
通讯作者：
Henglong Yang;Youmei Zhang;Yu Zhang;Hailong Meng;Shuang Li;Xianglin Dai
中科院分区：
文献类型：
--
作者：
Henglong Yang;Youmei Zhang;Yu Zhang;Hailong Meng;Shuang Li;Xianglin Dai
交通视频中的车辆计数和交通量估计得到了多媒体和计算机视觉领域的广泛关注。最近的车辆计数和流量估计方法，包括基于检测的方法和基于时间-空间图像(TSI)的方法都取得了显著的改进。然而，如何在精度和速度之间取得平衡仍然是这项任务面临的挑战。本文设计了一种快速准确的车辆计数和交通量估计方法。首先，将交通视频转换为TSIS，并在TSIS中人工标注车辆位置。然后，我们设计了一个简单的TSI密度图估计网络，该网络利用注意力机制来增强交通位置的特征，用于车辆统计。最后，我们利用车辆计数网络得到的参数来进一步估计交通量。在UA-DETRAC数据集上的实验表明，车辆计数网络不仅在计数精度和速度之间取得了平衡，而且在视频数据不足的情况下能够很好地估计交通量。
Vehicle counting and traffic volume estimation on traffic videos has gained extensive attention from multimedia and computer vision communities. Recent vehicle counting and volume estimation methods, including detection based and time-spatial image (TSI) based methods have achieved significant improvements. However, how to balance the accuracy and speed is still a challenge to this task. In this paper, we design a fast and accurate vehicle counting and traffic volume estimation method. Firstly, traffic videos are converted to TSIs and we annotate the vehicle locations in TSIs manually. Then, we design a simple TSI density map estimation network which utilizes attention mechanism to strengthen the features in the traffic locations for vehicle counting. Finally, we use the parameters obtained from the vehicle counting network to further estimate the traffic volume. Experiments on UA-DETRAC dataset demonstrate that the vehicle counting network not only takes a balance between counting accuracy and speed, but also well estimates the traffic volume when the video data is insufficient.