Noncontact Blood Pressure Monitoring Technology using Facial Photoplethysmograms

We propose noncontact blood pressure estimation technology using photoplethysmograms (PPGs) extracted from camera images of only user's face and this estimation can be realized without calibration, which usually needs to be performed for each individual. Proposed technology contains the pulse wave detection method robust to lighting conditions. The influence of users' body motions was evaluated by our experimental system. We also propose the method to reduce the blood pressure prediction error by classifying users according to their vascular conditions which can be detected from facial PPGs.