The invention discloses a traffic accident prediction method, which comprises the following steps of: acquiring a vehicle-mounted video and preprocessing the vehicle-mounted video to obtain continuous image frames, and extracting time features and space features from the continuous image frames; carrying out medium-term fusion on the spatio-temporal characteristics; performing position weighting based on a graph structure on the object-level features; performing spatial attention on the weighted object-level features; splicing the object-level features and the full-frame features, carrying out frame-level aggregation and time period-level aggregation to fully learn traffic scene context information, and carrying out space-time relation reasoning on obtained double-layer aggregation information to obtain hidden state representation of a current time step; and inputting the hidden state representation of the current time step into the fully connected network to obtain an accident score, and giving an early warning signal when the accident score exceeds a predetermined threshold. According to the method, the occurrence of the traffic accident can be successfully predicted 2-4 seconds ahead of time, so that enough reaction time is provided for a driving system or a driver, and the possibility of occurrence of the traffic accident is greatly reduced.

    本发明公开了一种交通事故预测方法,包括:获取车载视频并进行预处理,得到连续的图像帧,对连续的图像帧提取时间特征与空间特征;并将时空特征进行中期融合;对对象级特征进行基于图结构的位置加权;对加权后的对象级特征进行空间注意;将对象级特征与全帧特征拼接并进行帧级聚合和时段级聚合,以充分学习交通场景上下文信息,对获得的双层聚合信息进行时空关系推理,以获得当前时间步的隐藏状态表示;将当前时间步的隐藏状态表示输入至全连接网络,得到事故评分,并在事故评分超过预定阈值时给出预警信号。该方法可以提前2~4秒成功的预测交通事故的发生,从而给与驾驶系统或驾驶员足够的反应时间,因此大大减少交通事故发生的可能性。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Traffic accident prediction method and device, electronic equipment and storage medium


    Weitere Titelangaben:

    交通事故预测方法、装置、电子设备及存储介质


    Beteiligte:
    LIU WEI (Autor:in) / ZHANG TAO (Autor:in) / LU YISHENG (Autor:in) / CHEN JUN (Autor:in) / WEI LONGSHENG (Autor:in)

    Erscheinungsdatum :

    2023-06-27


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G06Q Datenverarbeitungssysteme oder -verfahren, besonders angepasst an verwaltungstechnische, geschäftliche, finanzielle oder betriebswirtschaftliche Zwecke, sowie an geschäftsbezogene Überwachungs- oder Voraussagezwecke , DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G06V / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS



    Traffic accident prediction method and device, equipment and storage medium

    HOU CHEN | Europäisches Patentamt | 2020

    Freier Zugriff

    Traffic accident detection method and device, electronic equipment and storage medium

    ZHANG SHANGXIN | Europäisches Patentamt | 2022

    Freier Zugriff

    Traffic accident prediction method and device, terminal and storage medium

    XU XIANJIE / WU RUI / HU MINZHI | Europäisches Patentamt | 2023

    Freier Zugriff

    Traffic accident severity prediction method and device and storage medium

    LI YUAN / ZHANG HUALEI | Europäisches Patentamt | 2023

    Freier Zugriff

    Traffic accident analysis method and system, electronic equipment and storage medium

    ZHOU WENHUI / LU PING / WANG YIFAN et al. | Europäisches Patentamt | 2022

    Freier Zugriff