A series of arterial roads and major intersections in most major cities in the world often fall into traffic congestion, slowing socioeconomic growth and negatively impacting the environment. Therefore, this article presents a solution to identify traffic congestion by leveraging the view of many cameras at intersections in the city to recognize the amount and volume of vehicles moving and whether or not traffic congestion will occur. Total Vehicles and Traffic Velocity are two essential features considered as an evaluation index of traffic conditions. Our proposed solution encompasses three main stages: vehicle detection, motion estimation, and identification. To begin with, we introduce the improved YOLOv4-CSP model and OC-SORT tracking algorithm to detect and count total vehicles with the training dataset derived from the HCM AI Challenge competition. For the motion estimation, the feature-matching method extracts the keypoints of detected vehicles. Subsequently, we calculate the Euclidean distance of the corresponding keypoints in the next frame to estimate traffic velocity. Finally, we combine the two preceding stages and abide by the following rule to identify the traffic state. A traffic state will be normal if the total number of vehicles is dense but moving steadily. On the contrary, if the vehicle density is prolonged but the vehicles move very slowly, this traffic state is identified as traffic congestion. Experimental results demonstrate that our study obtained satisfactory results and suggest that it is prospective for practical applications.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Identifying Traffic Congestion Through Vehicle Counting and Motion Estimation


    Additional title:

    Lect. Notes in Networks, Syst.


    Contributors:

    Conference:

    International Conference on Green Technology and Sustainable Development ; 2024 ; Ho Chi Minh, Vietnam July 25, 2024 - July 26, 2024



    Publication date :

    2024-12-24


    Size :

    14 pages





    Type of media :

    Article/Chapter (Book)


    Type of material :

    Electronic Resource


    Language :

    English




    TRAFFIC CONGESTION ESTIMATION

    YAMAMOTO GAKU / MIZUTA HIDEYUKI / WATANABE SHOICHIRO | European Patent Office | 2020

    Free access

    Identifying and quantifying congestion within traffic flow

    GORDON RAPHAEL / GRIMM DONALD K / BAI FAN | European Patent Office | 2024

    Free access


    Method of traffic congestion estimation

    KOSHIZEN TAKAMASA / EGUCHI JUNJI / NISHINARI KATSUHIRO | European Patent Office | 2016

    Free access

    IDENTIFYING AND QUANTIFYING CONGESTION WITHIN A TRAFFIC STREAM

    GORDON RICHARD / GRIMM DONALD K / BAI FAN | European Patent Office | 2024

    Free access