ABSTRACT
In the context of the COVID-19 health emergency, it is necessary to have devices that help identify symptoms that indicate whether a person has COVID-19. As one of the main symptoms is fever, which can be identified by measuring body temperature, different non-contact measurement methods are being widely used as an alternative to traditional contact thermometers. However, readings with thermographic cameras present limitations in terms of high dependence on the environment. For this reason, this article aims to validate the temperature measurements of a thermographic camera by comparing different models using as a reference standard the readings of a calibrated medical infrared thermometer. For this purpose, 463 measurements were analyzed using an infrared thermometer and a thermographic camera simultaneously. As significant differences were observed between the measurements made using statistical analyses with (p< 0.05 ), models were developed establishing weighting and compensation criteria to obtain similar readings between the values measured from the thermal images and the infrared thermometer. For uncertainty estimation, linear and non-linear regression models such as artificial neural networks were tested, selecting the best model that allows reducing the variation among the readings. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.