International Journal of Advanced Research in Computer and Communication Engineering

Abstract: In industries such as construction, manufacturing, and chemical processing, Personal Protective Equipment (PPE) plays a critical role in protecting workers from serious injuries. Even with strict safety rules in place, many workplaces struggle to ensure consistent PPE use, often due to negligence or lack of constant supervision. Relying on manual checks is time-consuming, error-prone, and impractical for large-scale monitoring. This study presents a real-time PPE detection system that combines computer vision with deep learning to address these challenges. The system uses the YOLOv8 object detection model to identify key safety items—helmets, safety vests, and face masks—directly from live video streams. A diverse and annotated dataset of industrial scenarios was used for training, enabling the model to reach a mean Average Precision (mAP) of 96%. The results show that the system can accurately and quickly detect PPE usage, offering a practical, scalable, and cost-effective alternative to manual oversight. By reducing reliance on human monitoring, this approach can improve compliance, enhance workplace safety, and help prevent avoidable accidents.

Keywords: PPE detection, YOLOv8, deep learning, computer vision, workplace safety, real-time monitoring.

Downloads: | DOI: 10.17148/IJARCCE.2025.14832