Detection of personal protective equipment through automated systems based on computer vision and machine learning

Adjiski, Vancho (2024) Detection of personal protective equipment through automated systems based on computer vision and machine learning. In: Technology of underground and surface mining of mineral raw materials, Podeks - Poveks ’24, 18-20 Oct 2024, Struga.

Text Adjiski_Podeks-2024.pdf - Accepted Version Download (1MB)

Abstract

This scientific paper presents the development and implementation of a system for automatic detection of personal protective equipment (PPE) in the mining industry, using computer vision and machine learning. The model is based on the YOLO v8 object detection architecture and is trained on a dataset of images featuring various types of personal protective equipment. The primary goal of this research is to provide efficient and real-time detection of helmets, protective glasses, ear protectors, safety vests, boots, and gloves in mining environments, where worker safety is of critical importance. The model results demonstrate high accuracy and efficiency, with a mean
Average Precision (mAP) of 76.9%, precision of 78.3%, and recall of 71.1%. This system has the potential to be integrated with existing surveillance cameras in mines, thereby enabling automatic monitoring of compliance with safety regulations.

Item Type:	Conference or Workshop Item (Speech)
Uncontrolled Keywords:	Computer Vision, Machine Learning, YOLO, Object Detection, Personal Protective Equipment, Mining, Roboflow.
Subjects:	Engineering and Technology > Other engineering and technologies
Divisions:	Faculty of Natural and Technical Sciences
Depositing User:	Vanco Adziski
Date Deposited:	07 Nov 2024 08:45
Last Modified:	07 Nov 2024 08:45
URI:	https://eprints.ugd.edu.mk/id/eprint/34958

Actions (login required)

View Item