Design and Implementation of Low-Cost Smart Contactless Thermometer with Polynomial Regression Model

Abstract

Body temperature is an important measure of a person's health. It is used to diagnose a variety of disorders. Core temperature measurement takes some time and puts patients under strain. To measure temperature in real-time and without contact. We proposed a contactless thermometer that can be calibrated by deriving the ambient and body temperatures from thermometer readings. We analyze this accuracy using second-order polynomial linear regression. After getting the actual body temperature of 124 subjects, we achieved 99% prediction temperature accuracy and linearity with gold standard reference temperature sensors. Additionally, the application has been developed to save the readings. This application can be connected through either OTG or Bluetooth depending on the user's preference. They will have the option to save multiple people's temperature data in separate excel sheets. Following that, the user can access the data and share the link to the data using any of the shareable platforms.