The invention relates to the field of road traffic management, and discloses a road traffic abnormal condition detection system and method, and the method comprises the steps: extracting a road monitoring region through semantic segmentation in a road monitoring video image, and obtaining a continuous traffic observation frame sequence; performing cross-frame multi-target tracking on the vehicle targets in the frame sequence, and constructing a corresponding vehicle motion state time sequence vector; based on the time sequence vector, traffic flow evolution characteristics are extracted by adopting a trajectory aggregation algorithm, and an abnormal behavior mode is identified; aiming at the detected abnormal behavior area, constructing a graph neural network model by combining a time-space relationship between vehicles, and analyzing local mutation nodes of a traffic interaction structure; judging whether a macroscopic traffic abnormal event exists or not according to an output result of the graph neural network; and if the traffic abnormal event exists, generating a traffic abnormal distribution thermodynamic diagram on a map according to the type, the duration and the coverage range of the identified traffic abnormal event. The method has the advantage that the intelligent level and the reliability of the detection system are improved.

    本发明涉及公路交通管理领域,且公开了一种公路交通的异常状况检测系统及方法,包括通过道路监控视频图像中的语义分割,提取出道路监测区域,获得连续的交通观测帧序列;对帧序列中的车辆目标进行跨帧多目标跟踪,构建对应的车辆运动状态时序向量;基于时序向量,采用轨迹聚合算法提取交通流演变特征,识别异常行为模式;针对检测到的异常行为区域,结合车辆之间的时空关系构建图神经网络模型,分析交通交互结构的局部突变节点;根据图神经网络的输出结果,判断是否存在宏观交通异常事件;若存在,根据识别出的交通异常事件的类型、时长和覆盖范围,在地图上生成交通异常分布热力图。本发明具备提高检测系统的智能化水平和可靠性的优点。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Road traffic abnormal condition detection system and method


    Weitere Titelangaben:

    一种公路交通的异常状况检测系统及方法


    Beteiligte:
    ZHU YE (Autor:in) / ZHOU GUANGQI (Autor:in) / SHEN ZIMIAN (Autor:in) / KONG XIANGGUO (Autor:in) / ZHAO XUE (Autor:in) / LI YING (Autor:in) / ZHAO YIHE (Autor:in) / LIU FUZHONG (Autor:in) / HU HAIXIAO (Autor:in) / SHEN CONG (Autor:in)

    Erscheinungsdatum :

    30.05.2025


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

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




    Traffic road condition intelligent lightning protection detection system

    WANG ENSHI / YANG TAO / YU ZHIHUA et al. | Europäisches Patentamt | 2024

    Freier Zugriff

    Traffic abnormal road section probability identification method

    WANG HAO / LI JIANYUAN / ZHAO BEIBEI et al. | Europäisches Patentamt | 2015

    Freier Zugriff

    Traffic road condition display sign and road condition display system

    CHEN LIHUA / SUN JIGANG / QIAN YUANYUAN | Europäisches Patentamt | 2023

    Freier Zugriff

    ROAD TRAFFIC CONDITION MONITORING SYSTEM

    YOKOTA KENSUKE | Europäisches Patentamt | 2023

    Freier Zugriff