Is It Possible to Estimate Average Heart Rate from Facial Thermal Imaging?
David Perpetuini, Andrea Di Credico, Chiara Filippini, Pascal Izzicupo, Daniela Cardone, Piero Chiacchiaretta, Barbara Ghinassi, Angela Di Baldassarre, Arcangelo Merla
Abstract
The remote measurement of heart rate (HR) could have many applications, such as health and emotional conditions monitoring. Currently, methods based on visible cameras have been developed for HR estimation. However, the employment of such techniques with scarce illumination conditions could be challenging. Infrared Thermography (IRT) could be a valuable tool to overcome this limitation. This study investigated the possibility of estimating average HR with facial IRT through a cross-validated machine learning (ML) approach. The correlation coefficient between the estimated and the measured HR was 0.7. Although preliminary, these results demonstrate the feasibility of estimating HR with IRT.