The invention discloses a vehicle track detection method based on deep learning, and relates to the technical field of track detection, and the method comprises the steps: collecting a physical address of an intelligent central controller in advance, and generating a virtual address for the intelligent central controller in real time when a vehicle moves; the driving condition of a road vehicle is obtained in real time through a monitoring camera; pre-training a CNN neural network model capable of identifying whether the vehicle has a turning intention or not; recognizing a vehicle with a lane turning intention in real time according to the model; according to the distance between the vehicle and the front vehicle, judging whether the vehicle needs to turn a lane or overtake; if yes, the speed needed by overtaking is calculated according to the vehicle running speed, the front vehicle running speed, the nearest opposite vehicle running speed, the distance to the front vehicle and the distance to the nearest opposite vehicle; judging whether the vehicle has an overtaking condition or not according to the overtaking speed; when no overtaking condition exists or careful overtaking is needed, an alarm is sent to the intelligent center controller; when the vehicle overtakes, the overtaking condition is analyzed in advance, and vehicle collision is avoided in advance.

    本发明公开了一种基于深度学习的车辆轨迹检测方法,涉及轨迹检测技术领域,预先收集智能中控的物理地址,并在车辆移动时实时为智能中控生成虚拟地址;通过监控摄像头实时获取道路车辆的行驶情况;并预先训练可识别车辆是否有转道意图的CNN神经网络模型;根据模型实时识别具有转道意图的车辆;根据车辆与前车的距离判断该车是转道或超车的需求;若为超车需求,根据车辆行驶速度、前车行驶速度、最近对向车辆行驶速度、与前车距离以及与最近对向车辆距离计算超车需要的速度;并根据超车速度判断车辆是否具有超车条件;在无超车条件或需谨慎超车时,向智能中控发送警告;在车辆超车时预先分析超车条件,提前避免了车辆碰撞的发生。


    Access

    Download


    Export, share and cite



    Title :

    Vehicle track detection method based on deep learning


    Additional title:

    一种基于深度学习的车辆轨迹检测方法


    Contributors:
    ZHANG ZHONG (author) / SUN GUANLONG (author) / HUANG JUNJIE (author)

    Publication date :

    2022-11-18


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G06V / G08B SIGNALLING OR CALLING SYSTEMS , Signalisier- oder Rufsysteme



    Abnormal vehicle track anomaly detection method and system based on deep learning

    WANG LEI / LIU YUHANG / ZHAO XIAOYONG et al. | European Patent Office | 2023

    Free access

    Vehicle track deep learning prediction method considering physical constraint

    XU FANG / HAN SHUAI / GUO ZHONGYI et al. | European Patent Office | 2023

    Free access

    Railway Track Fasteners Fault Detection using Deep Learning

    Lin, Ya-Wen / Hsieh, Chen-Chiung / Huang, Wei-Hsin et al. | IEEE | 2019


    TrackNet - A Deep Learning Based Fault Detection for Railway Track Inspection

    James, Ashish / Jie, Wang / Xulei, Yang et al. | IEEE | 2018


    Communication unmanned aerial vehicle track optimization control method based on deep intensity learning

    YIN RUI / PENG JINGWEI / YUAN JIANTAO et al. | European Patent Office | 2024

    Free access