Surveillance system of mask detection with infrared temperature sensor on Jetson Nano Kit

Abstract

Coronavirus desease-19 (COVID-19) has made it mandatory for people to cover their faces in public areas since the right use of a mask is effective at protecting people from viral infection. It has been also shown that body temperature may indicate an individual's health state. Deep learning is being used in this work to construct an actual strategy to meet the current demand for mask-wearing status and facial temperature detection before entering a public venue. For the mask detection service provider, a surveillance system is constructed utilizing a deep learning technique employing a Jetson Nano. The alert is triggered by an infrared temperature sensor and a buzzer. AMG8833 and C920e camera are used to take input images and measure a person's body temperature at the same time. A warning sound is produced when the temperature of a person's face exceeds the normal range for human beings during these tests, which result in a live video showing the right information on whether the individual is wearing a mask correctly and how hot his or her face is. The model is light and fast, with a 99% accuracy rate for training and a 100% accuracy rate while testing.