As a result of the increased density of automobiles with population growth in the modern world, several problems have arisen, including road congestion and air pollution due to emissions from vehicles, along with increased rates of accidents. This is very true for underdeveloped countries where activity monitoring is very difficult. This paper presents a smart traffic monitoring system utilizing some state-of-the-art computer vision methodologies with aerial drone footage. The high-resolution cameras on drones take pictures of the traffic flow, which after processing with YOLOv8 and ResNet50 algorithms for vehicle recognition, showed a 97% and 91% accuracy of identification rate, respectively. Besides, the application of a time series by LSTM and GRUs has performed well in predicting the flow of traffic. Among them, the LSTM model had an MSE of 20.98%, while RMSE accuracy in that produced by GRU was 0.943%. This integrated system will contribute to improving traffic monitoring, congestion identification, and real-time data provision for traffic management. Its future enhancement will enable it to recognize vehicle license plates to facilitate better law and order and surveillance action on the ground.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Traffic Monitoring System Using Drone Images




    Publication date :

    2024-12-06


    Size :

    2050590 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Smart Traffic Monitoring through Drone Images via Yolov5 and Kalman Filter

    Hanzla, Muhammad / Ali, Shuja / Jalal, Ahmad | IEEE | 2024


    TRAFFIC CONTROL SYSTEM USING DRONE

    SHIN SOO YOUNG / KIM JI HUN / WOO JUN HYUK et al. | European Patent Office | 2024

    Free access

    Traffic drone system

    GRAHAM RICHARD DANIEL / BILLMAN BRADLY JAY | European Patent Office | 2023

    Free access

    Traffic drone system

    GRAHAM RICHARD DANIEL / BILLMAN BRADLY JAY | European Patent Office | 2021

    Free access