Facial Recognition and Thermal Imaging: A Cost-Effective Solution for Covid-19 Detection

Abstract

The global pandemic induced by the 2019 Novel Coronavirus Disease (Covid-19) has posed significant challenges for nations across the globe. Given the pandemic's pervasive nature, there is an emerging demand for a dependable tool capable of identifying individuals exhibiting fever, a primary symptom of Covid-19 infection. To address this, utilizing facial recognition technology in conjunction with temperature measurement has been widely embraced within various infrastructures such as residential buildings and office spaces. This research proposes the adoption of a system capable of recognizing human faces while simultaneously monitoring individual temperatures. This is achieved through the utilization of Python and open-source libraries such as OpenCV and NumPy to develop an effective facial identification system. Furthermore, this research suggests leveraging the capabilities of a cost-effective AMG8833 thermal imaging camera to measure human body temperature. The thermal image, reflecting the individual's body temperature, is displayed on the Node-RED dashboard, a platform based on Internet of Things (IoT) technology. Should the temperature reading of an individual exceed 37.5 degrees Celsius, the system is designed to activate an alarm and dispatch notifications via an administrative mobile application. All pertinent information regarding the individual is securely stored within a MySQL webserver database. A comparative analysis reveals that the proposed system provides nearly 95% cost reduction when compared against commercial alternatives such as the Flir C3, with the added advantage of image recognition capabilities.