The detection and tracking of vehicle speed are crucial for ensuring the safety of civilians and preventing accidents. Speed tracking plays a pivotal role in traffic surveillance, where effective management and the well-being of citizens take precedence. The approach leverages computer vision-based models, specifically Multi-scale Detector for object detection, and a fusion of the DeepSORT and Centroid Sort algorithms for vehicle tracking from streaming video. To cater to the requirements of these algorithms, a custom dataset was developed within a controlled setting. Dataset augmentation was employed to prevent overfitting, thereby enhancing the dataset's diversity and the model's ability to generalize. The method demonstrates the capacity to accurately detect and track vehicles, maintaining performance even in high-traffic conditions. Cross-validation at every epoch during the training phase was implemented to increase the model's accuracy. In addition to the technical aspects, an LED display board was integrated to provide real-time data to oncoming vehicles at road curves. This display showcases information such as vehicle types, passage time, and the number of vehicles within the curvy road. The proposed Multi-Scale Detector model, a customized version of the You Only Look Once (YOLO) architecture, achieved a mean average precision (mAP) of 0.954 at a confidence interval of 0.5. This overall research provides an effective and accurate solution for optimizing vehicle tracking, enhancing transportation safety, and enabling a variety of practical applications in traffic management and monitoring.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Revolutionizing Road Safety for Advanced Vehicular Accident Prediction


    Contributors:


    Publication date :

    2024-01-04


    Size :

    982081 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Revolutionizing Safety: the Smart Accident Response System

    Nandal, Poonam / Jindal, Srishty / Singh, Meeta et al. | IEEE | 2024


    Accident Prediction Models for Winter Road Safety

    Usman, Taimur / Fu, Liping / Miranda-Moreno, Luis F. | Transportation Research Record | 2011


    Road Network Safety Ranking Using Accident Prediction Models

    Jasiūnienė, Vilma / Ratkevičiūtė, Kornelija / Peltola, Harri | Springer Verlag | 2019



    SAFETY - French road accident

    Online Contents | 1994