An online PPGI approach for camera based heart rate monitoring using beat-to-beat detection

In this work an online method for camera-based heart rate detection, also known as Photoplethysmography Imaging (PPGI), is presented. The pulse related signal is obtained from RGB videos of the human face, using an off the shelf camera under ambient light conditions. The algorithm for heart rate estimation is based on the beat-to-beat analysis of the PPGI signal, allowing the estimation of psychophysiological parameters like the heart rate variability. The system is tested indoors under laboratory conditions, and in field trials integrated in a car. Field tests are done as a step towards robust appraisal of psycho-physiological driver states. In measurements with eight subjects, a sensitivity of ~93% for pulse peak detection is reached under resting and constant light conditions. First test drives show good results for heart rate estimation while driving on highways outside of city traffic. The results indicate that PPGI can be a suitable solution for providing vital sign information of drivers in cars.