Illumination Invariant Non-Invasive Heart Rate And Blood Pressure Estimation From Facial Thermal Images using Deep Learning

Heart rate is usually measured using Electrocar-diogram(ECG) and Photoplethysmogram(PPG). Heart rate estimation from images and face videos are well studied in recent years. All methods need proper illumination and it must be able to operate under different lighting conditions in a restricted environment. Both ECG and PPG require direct skin contact with the person, hence it is not of much use in situations where establishing such a contact is difficult. PPG signal extraction from thermal images is light invariant. Also, it avoids the need for direct contact with the person and allows to monitor the vital signals which could be used for heart rate and blood pressure measurement. Here, using deep learning, heart rate and blood pressure is estimated from Infra Red facial images and the the blood pressure estimation by the model fulfilled the AAMI(Association for the Advancement of Medical Instrumentation) standard and BHS standard(British Hypertension Society).