| Title |
Mask, Hairnet, and Handwash Detection System for Food Manufacturing Health and Safety Monitoring |
|
Authors |
CP Pepito, ER Aleluya, FJ Alagon, S Clar, JJ Abayan, CJ Salaan, MF Bahinting |
|
Publication date |
2023 |
|
Conference |
2023 IEEE 15th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM) |
| Pages |
1-6 |
| Publisher |
IEEE |
| Abstract |
Food safety is a top concern within food manufacturing facilities, with strict measures in place to ensure the quality and integrity of products. The existing food safety measures typically rely on manual inspections and human vigilance to ensure compliance with safety protocols. These measures, although essential, are prone to human error and limitations in terms of real-time monitoring. Moreover, the global pandemic has underscored the importance of adherence to health and safety guidelines, including the wearing of masks, hairnets, and proper handwashing practices to ensure the well-being of workers and the integrity of food production processes. This research introduces the development of a mask, hairnet, and handwash detection system employing pre-trained deep learning models, specifically YOLOv5 and YOLOv8 (You Only Look Once). Each model was evaluated on various metrics to assess its accuracy and overall performance. The YOLOv8 model showcases outstanding performance in mask and hairnet detection, along with handwash detection. For mask and hairnet detection, it achieved a precision, recall, and F1 score of 100%, and mAP of 99.40%. While handwash detection attained a precision, recall, and F1 score of 100%, and mAP of 92.70%. To provide a comparative analysis between the YOLO models used in the study, the YOLOv5 model was also trained using the same dataset as the YOLOv8 and was evaluated accordingly. Moreover, this study aims to impart significant contributions, particularly in the real-time application of health and safety monitoring within food manufacturing settings. |
| Index terms / Keywords |
deep learning , industrial smart monitoring , object detection , YOLOv5 , YOLOv8 |
| DOI |
10.1109/HNICEM60674.2023.10589183 |
| URL |
|
Mindanao State
University - Iligan Institute of Technology