The majority of road accidents occur as a result of driver irresponsibility. Utilizing deep learning models to analyze traffic CCTV can effectively mitigate traffic accidents. This paper aims to identify and monitor vehicles and assess traffic conditions by comparing and categorizing the similarities between vehicle trajectories. Three similarity measurement methods are employed: cosine similarity, Jensen-Shannon divergence, and Euclidean distance similarity. The results show that lanes and traffic patterns can be effectively identified using the presented approach. Through our approach, one can further develop traffic monitoring and analysis advancements, thereby enhancing road safety and traffic management


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Real-Time Traffic Analysis Using Vehicle Trajectory Similarity in Edge Computing


    Contributors:


    Publication date :

    2025-02-18


    Size :

    5215310 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Self‐Similarity of Real Time Traffic

    Sheluhin, Oleg I. / Smolskiy, Sergey M. / Osin, Andrey V. | Wiley | 2007



    Vehicle trajectory calculation method based on space-time similarity

    JING GUOSHENG / ZHOU ZHIHUA / HU JINGSONG et al. | European Patent Office | 2021

    Free access

    Real-Time Vehicle Trajectory Prediction for Traffic Conflict Detection at Unsignalized Intersections

    Qianxia Cao / Zhongxing Zhao / Qiaoqiong Zeng et al. | DOAJ | 2021

    Free access