The invention discloses a tunnel side-mounted video traffic flow detection method and system based on deep learning, and relates to the field of road traffic. The method includes steps of: collecting tunnel side-mounted video data; carrying out real-time detection of tunnel traffic parameters on the video data through a trained deep learning vehicle detection model; and obtaining vehicle information passing through the camera and tunnel traffic parameters. Herein, the traffic flow parameters of the tunnel side-mounted video are detected in real time through the deep learning vehicle detection model of the scheme, the tunnel traffic parameters including the flow, the vehicle speed and the vehicle distance can be extracted in real time, and the method has practical value in tunnel traffic operation monitoring, prediction, early warning, management and control and the like. Meanwhile, vehicle types can be identified, traffic monitoring images can be provided, wherein the composite performance is better than that of traditional detection methods such as an annular coil and an ultrasonic detector.
本发明公开了一种基于深度学习的隧道侧装视频交通流检测方法和系统,涉及道路交通领域。采集隧道侧装视频数据;通过训练好的深度学习车辆检测模型,对所述视频数据进行隧道交通参数的实时检测;获得经过摄像头的车辆信息,以及隧道交通参数。通过本方案的深度学习车辆检测模型对隧道侧装视频的交通流参数进行实时检测,可实时提取隧道交通参数包括流量、车速、车间距,对隧道交通运行监测、预测、预警、管控等具有实用价值。同时还能识别车辆类型和提供交通监控图像,复合性能超过环形线圈、超声波检测器等传统检测方法。
Tunnel side-mounted video traffic flow detection method and system based on deep learning
一种基于深度学习的隧道侧装视频交通流检测方法和系统
2021-07-20
Patent
Elektronische Ressource
Chinesisch
Traffic violation real-time flow monitoring method and system based on bus-mounted video
Europäisches Patentamt | 2023
|Traffic flow detection method based on traffic video network
Europäisches Patentamt | 2020
|Deep Learning based Traffic Flow Analysis Method and System
Europäisches Patentamt | 2021
|Europäisches Patentamt | 2020
|