A Low Cost Prototype of A Real-Time Face-Mask Detection and Enforcement System Using CNN and Arduino for Developing Countries

Thanks to advancements in Artificial Intelligence (AI), particularly in Machine Learning (ML), public health safety can be maintained more efficiently. Rather than risking the health of security personnel who must ensure compliance with face-mask mandates, ML techniques can automate this process with high efficiency. This paper introduces a prototype for a comprehensive face-mask detection system that not only detects face-masks but also controls access to secured buildings and dispenses masks to those without one. Initially, a Convolutional Neural Network (CNN)-based system with the help of a webcam, identifies whether an individual is wearing a mask. Depending on the detection outcome, a hardware door lock controlled by an Arduino chip either grants or restricts entry. If an individual is not wearing a mask, the system activates a mask dispenser. Once the individual puts on a mask, the system verifies its presence and allows entry. This integrated approach ensures that only masked individuals can access the building, enhancing public health protection.