There has been a tremendous growth in the transportation sector over the last decade. Importantly, this has also resulted in the need for communication between vehicles. However, the biggest disadvantage has been the high-speed nature of the vehicles which inhibits the stability of communication. A cluster-based technique wherein all vehicles within a region are grouped together with different levels of the hierarchy have been well investigated. However, such technique fails with dynamic variation in the traffic environment. In this paper, an adaptive clustering technique is proposed based on image-based real-time traffic identification. The novelty lies in proposing an inception v3 network with deep CNN (IP-DNN) for image-based traffic identification which then serves as the basis for selecting the best clustering technique. It has been found that such an adaptive mechanism results in up to 25% faster cluster selection as compared to static techniques. This is a significant result that provides the necessary impetus for developing real-time V2V (vehicle to vehicle) communication across varying traffic environments.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    An Adaptive Clustering Technique Based on Image-Based Traffic Identification for Real-Time V2V Communication


    Beteiligte:


    Erscheinungsdatum :

    01.06.2018


    Format / Umfang :

    611942 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Adaptive Traffic Management System Based on Real-Time Traffic Conditions

    Gherghinescu, Mihai / Ivașcu, Todor / Ștefănigă, Sebastian | Springer Verlag | 2024


    Clustering-based real-time traffic jam prediction method and equipment

    LI XIAOCHUN / HUANG DANYANG / YI SHU et al. | Europäisches Patentamt | 2024

    Freier Zugriff

    Traffic monitoring based on real-time image tracking

    Lin, Ching-Po / Tai, Jen-Chao / Song, Kai-Tai | Tema Archiv | 2003