Personal Protective Equipment Wearing Detection Using Deep Learning
Article Sidebar
Main Article Content
Abstract
This research aimed to develop a personal protective equipment (PPE) wearing detection system by employing deep learning techniques in conjunction with the YOLOv7 algorithm for real-time image analysis. The system was capable of detecting safety helmets, reflective vests, safety glasses, and ear protection. It was trained on a dataset constructed from images of correctly worn PPE and evaluated using performance metrics such as Precision, Recall, mAP@0.5, and mAP@0.5:0.95. The results demonstrated that the model achieved high accuracy in PPE detection, with an overall Precision of 0.956, Recall of 0.928, mAP@0.5 of 0.966, and mAP@0.5:0.95 of 0.744. It could also effectively detect the correct wearing of PPE based on the defined compliance conditions. In addition, the model was also able to perform real-time alerts upon detecting violations and to record the alert data for later review. In the real-world simulations, the model could detect PPE wearing in real-time manner while the users were working. However, there were still limitations in identifying certain incorrect wearing behaviors. The developed system can be applied to improve safety in the workplace and serve as a foundation for the development of better safety monitoring systems in the future.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
บทความที่ได้รับการตีพิมพ์เป็นลิขสิทธิ์ของ วารสารวิทยาศาสตร์และเทคโนโลยี มหาวิทยาลัยอุบลราชธานี
ข้อความที่ปรากฏในบทความแต่ละเรื่องในวารสารวิชาการเล่มนี้เป็นความคิดเห็นส่วนตัวของผู้เขียนแต่ละท่านไม่เกี่ยวข้องกับมหาวิทยาลัยอุบลราชธานี และคณาจารย์ท่านอื่นๆในมหาวิทยาลัยฯ แต่อย่างใด ความรับผิดชอบองค์ประกอบทั้งหมดของบทความแต่ละเรื่องเป็นของผู้เขียนแต่ละท่าน หากมีความผิดพลาดใดๆ ผู้เขียนแต่ละท่านจะรับผิดชอบบทความของตนเองแต่ผู้เดียว
References
Social Security Office. 2024. Occupational Injury or Illness Situation Due to Work 2019-2023. https://www.sso.go.th/wpr/assets/upload/files_storage/sso_th/1675d2a95c38687dd649989003beb08a.pdf. Accessed 11 December 2024. (in Thai)
Thai RSC. 2025. Comparison of Road Accident Statistics in Thailand. https://www.thairsc.com/data-compare. Accessed 11 December 2024. (in Thai)
The Bureau of Registration Administration. 2023. Announcement of the Central Registration Office Regarding the Number of People throughout the Kingdom of Thailand According to the Population Registration Evidence. https://stat.bora.dopa.go.th/stat/pk/pk_66.pdf. Accessed 24 January 2025. (in Thai)
Office of the National Economic and Social Development Council. 2023. Thai Social Conditions, Fourth Quarter and Overall Picture, Year 2023. https://www.nesdc.go.th/the-story-of-thai-social-conditions-in-the-fourth-quarter-and-the-overall-picture-of-2023/. Accessed 24 January 2025. (in Thai)
Ferdous, M. and Ahsan, S.M.M. 2022. PPE detector: A YOLO-based architecture to detect personal protective equipment (PPE) for construction sites. PeerJ Computer Science. 8: e999.
Lo, J.H., Lin, L.K. and Hung, C.C. 2023. Real-time personal protective equipment compliance detection based on deep learning algorithm. Sustainability. 15(1): 391.
Wang, C.Y., Bochkovskiy, A. and Liao, H.Y.M. 2022. YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors. https://arxiv.org/pdf/2207.02696. Accessed 22 December 2024.
Dwyer, B. and Gallagher, J. 2023. Getting Started with Roboflow. https://blog.roboflow.com/getting-started-with-roboflow/. Accessed 29 January 2025.
Bochkovskiy, A., Wang, C.Y. and Liao, H.Y.M. 2020. YOLOv4: Optimal Speed and Accuracy of Object Detection. https://arxiv.org/pdf/2004.10934. Accessed 22 December 2024.
Redmon, J. and et al. 2016. You only look once: Unified, real-time object detection. In: Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, 27-30 June 2016. Las Vegas, Nevada, USA.
Redmon, J. and Farhadi, A. 2017. YOLO9000: Better, faster, stronger. In: Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, 21-26 July 2017. Honolulu, Hawaii, USA.
Redmon, J. and Farhadi, A. 2018. YOLOv3: An Incremental Improvement. https://arxiv.org/pdf/1804.02767. Accessed 15 December 2024.
Department of Labour Protection and Welfare. 2018. Announcement of the Department of Welfare and Labor Protection Regarding the standards of light intensity. https://greenoffice.rsu.ac.th/law/law4_2.pdf. Accessed 15 December 2024. (in Thai)
Kaur, P., Khehra, B.S. and Mavi, B.S. 2021. Data augmentation for object detection: A review. In: Proceedings of the 2021 IEEE International Midwest Symposium on Circuits and Systems, 09-11 August 2021. Online Conference.
Lin, T.Y. and et al. 2014. Microsoft COCO: Common objects in context. In: Proceedings of the 13th European Conference on Computer Vision, 6-12 September 2014. Zurich, Switzerland.
Wong, K.Y. 2023. yolov7. https://github.com/WongKinYiu/yolov7. Accessed 24 January 2025.
Cheng, Y. and et al. 2017. A Survey of Model Compression and Acceleration for Deep Neural Networks. https://arxiv.org/pdf/1710.09282v1. Accessed 12 January 2025.
Gou, J. and et al. 2021. Knowledge distillation: A survey. International Journal of Computer Vision. 129: 1789-1819.
