Real-Time Face Mask Detection and Analysis System

ABSTRACT

Due to COVID-19 situation, we need to wear face masks in public places. Reports say that wearing face mask at public places and at workspace reduces the transmission of virus as the SARS-CoV-2 spreads through atmosphere among people, at gathering in any environment. In this paper, a real-time face mask detection system is presented which will detect mask presence on the face using TensorFlow. We are using MobileNetV2 model to provide a greater accuracy in determining the mask presence. Accuracy obtained is 99%. Older systems do not provide a proper working system. A face mask detector has been designed with computer vision using Python, OpenCV, Keras, and TensorFlow. Video surveillance input can be given directly, and our primary purpose is to identify to check people are wearing masks on daily basis or not wearing masks and prepare a weekly and monthly report based on this observation and display the data on an interactive web application. System provides option to see the historical records, thereby reducing transmission. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.