Computer Vision (CV) and Machine Learning (ML) have transformed manufacturing by enabling real-time monitoring and optimization. This study introduces a novel CV-based system employing multiple RGB 2D cameras for the localization of human workers on the shop floor. The system utilizes the SCRFD pre-trained 2D person detection neural network, leveraging existing surveillance and common video cameras to monitor worker positions accurately. By tracking workers in real time, the system enhances safety by detecting hazardous situations, thereby preventing accidents. The proposed methodology was validated using videos from an industrial setting in the production of wooden house modules, demonstrating robust performance with a detection rate of 67.37% and a mean absolute error of 0.5 m. This approach provides a cost-effective and precise solution to improve worker safety and operational efficiency in manufacturing environments, advancing the integration of advanced CV techniques in industry.


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Advanced Computer Vision for Industrial Safety: Indoor Human Worker Localization Using Deep Learning


    Weitere Titelangaben:

    Lect.Notes Mechanical Engineering


    Beteiligte:
    Alexopoulos, Kosmas (Herausgeber:in) / Makris, Sotiris (Herausgeber:in) / Stavropoulos, Panagiotis (Herausgeber:in) / Berardinucci, Francesco (Autor:in) / Urgo, Marcello (Autor:in)

    Kongress:

    European Symposium on Artificial Intelligence in Manufacturing ; 2024 ; Athens, Greece October 16, 2024 - October 16, 2024



    Erscheinungsdatum :

    22.03.2025


    Format / Umfang :

    10 pages





    Medientyp :

    Aufsatz/Kapitel (Buch)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch







    Railroad Worker Safety and Railroad Worker Safety Method using the System

    LEE WON HEE / CHUN JUNG HWAN / KIM HEE MAN | Europäisches Patentamt | 2024

    Freier Zugriff

    Vision-Based Indoor UAV Navigation with Deep Reinforcement Learning

    Guo, Yuxuan / Tang, Lei / Zhou, Xin et al. | Springer Verlag | 2025