Contactless heart rate and respiratory measurement for sleep monitoring

Abstract

Heart rate and respiration, which are physiological indicators closely related to cardiopulmonary function, are widely used in clinical diagnosis and daily monitoring. This paper proposes a heart rate measurement based on imaging photovoltaic plethysmography (iPPG) and a respiratory measurement method of sift algorithm. We mark the region of interest by dlib. Signal is obtained by color channels separation of ROI and independent component analysis. Next we filter signal with bandpass filter and extract heart rate from the blood volume pulse via Fourier transform. Meanwhile, the trajectory is decomposed into a set of component motions by calculating the feature points at the chest and performing principal component analysis. The composition of the respiration that is most suitable for the calculation is then selected based on its time spectrum. Finally, the peak of the trajectory corresponding to the respiration is determined. This method could be used for contactless sleep monitoring .