The real-time road traffic flow optimization system aims to enhance traffic management at intersections by employing deep learning techniques for dynamic signal adjustments. Traditional traffic control methods that rely on vehicle count often fail to account for the varying sizes of vehicles, which can result in inefficient signal timing. In contrast, the proposed system calculates the total area occupied by different vehicle types, including cars, buses, trucks, two-wheelers, and auto-rickshaws, within each lane. By fine-tuning a YOLO model specifically for these categories, it enables accurate detection and classification of vehicles from live traffic video feeds. The comprehensive measure of congestion, based on the total occupied area, ensures that traffic light durations are adjusted according to the actual traffic conditions in each lane, rather than just the number of vehicles. The proposed traffic system introduces a dynamic and intelligent framework for traffic signal management that outperforms traditional methods relying on fixed-timer signals or basic vehicle counting. Unlike conventional systems, the traffic signal control algorithm adapts signal durations in real-time, considering traffic density to reduce congestion and waiting times effectively. The mode-finding algorithm further enhances functionality by detecting ambulances or priority vehicles in specific lanes, ensuring immediate passage through adaptive signal adjustments. Additionally, the duration-finding algorithm calculates the optimal green light duration based on vehicle density and waiting time, enabling balanced traffic flow. Compared to static or less responsive traditional systems, the proposed method significantly improves traffic efficiency, minimizes delays offering a smarter and more responsive solution for urban traffic challenges.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Real-Time Road Traffic Flow Optimization using Deep learning Techniques


    Beteiligte:
    Abinaya, S. (Autor:in) / Cheran, D. (Autor:in) / Kumar, C. Ranjith (Autor:in) / Kumar, P. Saravana (Autor:in) / Gandhi, K. Indra (Autor:in)


    Erscheinungsdatum :

    09.04.2025


    Format / Umfang :

    382954 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Real-Time Traffic Intensity Estimation and Management using Deep Learning Techniques

    Murhekar, Suryakant / Kamble, Yash / Dongre, Snehlata et al. | IEEE | 2024


    Real-Time Traffic Flow Optimization

    Black, B.C. / Gazis, D.C. | Tema Archiv | 1971


    Real Time Traffic Sign Classification Using Deep Learning

    Gurupriya, M. / Veluru, Abhilash / Venkat, Hruday et al. | Springer Verlag | 2025

    Freier Zugriff

    Road Traffic Condition Monitoring using Deep Learning

    SASI PRIYA, S. / S, RAJARAJESHWARI. / K, SOWMIYA. et al. | IEEE | 2020


    Predicting the Risk of Citywide Traffic Road Accidents Using Deep Learning Techniques

    Poonkuzhali, S. / Rajesh Kannan, G. | Springer Verlag | 2025