交通运输系统工程与信息

交通运输系统工程与信息 ›› 2020, Vol. 20 ›› Issue (3): 39-46.

• 智能交通系统与信息技术 • 上一篇    下一篇

基于3DCNN-DNN 的高空视频交通状态预测

彭博1,2，唐聚*2，蔡晓禹1,2，谢济铭2，张媛媛2，王玉婷2   

  1. 1. 山地城市交通系统与安全重庆市重点实验室，重庆 400074；2. 重庆交通大学交通运输学院，重庆 400074
  • 收稿日期:2019-12-12 修回日期:2020-02-07 出版日期:2020-06-25 发布日期:2020-06-28
  • 作者简介:彭博(1986-)，男，四川南充人，副教授.
  • 基金资助:

    国家自然科学基金/National Natural Science Foundation of China(61703064)；重庆市基础前沿与技术创新项目/ Chongqing Research Program of Basic Research and Frontier Technology Innovation(cstc2017jcyjAX0473)；山地城市交通系统安全实验室开放基金/Scientific Research Project of Key Laboratory of Traffic System & Safety in Mountain Cities(2018TSSMC05).

3DCNN-DNN Based Traffic Status Prediction from Aerial Videos

PENG Bo1,2, TANG Ju2, CAI Xiao-yu1,2, XIE Ji-ming2, ZHANG Yuan-yuan2,WANG Yu-ting2   

  1. 1. Chongqing Key Lab of Traffic System & Safety in Mountain Cities, Chongqing 400074, China; 2. College of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, China
  • Received:2019-12-12 Revised:2020-02-07 Online:2020-06-25 Published:2020-06-28

PDF

摘要：

为使用高空视频识别和预测道路交通状态，提出基于三维卷积神经网络—深度神经网络(3DCNN-DNN)的交通状态预测方法. 将道路切分为D 个路段，每个路段视频片段时长 m s，基于典型3DCNN结构C3D识别路段视频片段交通状态；建立道路? 个历史时段、D 个路段的交通状态矩阵Φ ，将道路交通状态预测问题转化为以Φ 为输入，有限交通状态为输出的分类问题，构建基于DNN的短时交通状态预测模型原型；建立交通视频数据集，对DNN预测模型原型的隐藏层数量、神经元数量及训练批大小进行测试优化，提出有4 隐藏层，各层神经元数量为64/128/128/64，训练批大小为64 的优化模型DNN*.测试结果表明：C3D视频交通态识别平均F1 值为95.71%，DNN*道路交通状态预测准确率为91.18%，比DNN线性分类、KMeans 、KNN、SVM和线性分类分别高6.86%、57.85%、62.26%、26.47%、43.14%；C3D能提供准确的交通状态矩阵，3DCNN-DNN可有效识别和预测道路视频交通状态.

关键词: 智能交通, 交通状态预测, DNN, 高空视频, 3DCNN, 深度学习

Abstract:

This study proposes a 3D Convolutional Neural Network- Deep Neural Network (3DCNN- DNN) method to recognize and predict traffic status from aerial videos. First, the roadway was divided into D sections and each section had an m - seconds video clip. The traffic state was recognized based on a typical 3DCNN structure named C3D (Convolutional 3D). Then, traffic state matrix Φ was established containing D road sections and ? historic time periods, and the traffic state prediction problem was transformed into a classification task with the input of Φ and output of limited number of traffic states. A model prototype for short time traffic state prediction was developed based on DNN. Traffic video sets were then assembled, and the DNN prediction prototype was tested and optimized through the number of hidden layers, neurons amount and training batch sizes. As a result, an optimal model named DNN* was proposed, which included 4 hidden layers with 64/128/128/64 neurons and training batch size of 64. The test results indicate that: C3D reaches an average F1 score of 95.71% to recognize traffic states from aerial videos. The prediction precision of DNN* is 91.18%, which has been improved by 6.86%, 57.85%, 62.26%, 26.47% and 43.14% compared to the DNN-Linear classification, K-Means, KNN (Knearest Neighbor), SVM (Support- vector Machines) and Linear classification respectively. The C3D is able to provide accurate traffic state matrix, and 3DCNN-DNN could effectively recognize and predict traffic state from road aerial videos.

Key words: intelligent transportation, traffic status prediction, Deep Neural Network(DNN), traffic aerial video, 3D Convolutional Neural Network (3DCNN), deep learning

中图分类号: