DESIGNING OF COMPUTER VISION BASED SAFETY GEAR DETECTION USING YOLO V8 AND JETSON NANO

DESIGNING OF COMPUTER VISION BASED SAFETY GEAR DETECTION USING YOLO V8 AND JETSON NANO

K.Gurucharan, S.S. Kiran, M. Rajan Babu, K.DHARMATEJA, K.AVINASH BALAJI, K.UDAY KUMAR

Electronics and Communication Engineering, Lendi Institute of Engineering and Technology,

Visakhapatnam

Abstract—Enhancing worker safety on construction sites is pivotal, as lapses in wearing safety gear such as helmets, masks, and vests can lead to potential hazards. To address this issue, we introduce a novel solution leveraging the YOLO V8 device to detect helmet usage among workers. However, challenges arise due to security unawareness and discomfort, causing helmet removal. To mitigate this risk, we propose an advanced system employing the Jetson Nano board and . This system not only detects helmet adherence but also monitors the use of safety masks and vests. Unlike the existing YOLO V5-based approach, our proposed system adopts YOLO V8, ensuring heightened accuracy. The integration of a CSI Camera Module further enhances real-time monitoring capabilities. Through the synergy of Jetson Nano, YOLO V8 and CSI Camera Module, we pioneer an innovative solution to enforce safety compliance by preventing unauthorized entry without proper safety measures.

Index Terms— NVIDIA Jetson Nano, Object detection, Deep Learning, YOLO V8, CSI Camera module.