Motion-tolerant heart rate estimation from face videos using derivative filter
- 35 Downloads
Imaging photoplethysmography (IPPG) technique allows us to extract blood volume pulse (BVP) signals from face videos for measuring heart rate (HR), which is useful in applications such as neonatal monitoring, telemedicine and affective computing. Because the BVP signal is small, the HR estimation results are sensitive to face motion disturbance caused by spontaneous head movements and facial expressions of subjects. In this paper, we design a novel filtering method for refining the RGB signals with motion artifacts. Based on the observation that subtle color changes of face skin are smoother than large face motions at temporal scale, we use the three-order derivative of Gaussian filter to select subtle color changes under large motions. Our method is validated on both our self-collected dataset and public dataset MAHNOB-HCI containing face videos with head movements and facial expressions. By employing the proposed filtering method to pre-process the RGB signals before BVP signal extraction, a range of IPPG methods are improved to generate robust HR estimation results under realistic situations.
KeywordsPhotoplethysmography Imaging Derivative filter Motion interference Heart rate estimation
We acknowledge funding support from: Training Programme Foundation for Application of Scientific and Technological Achievements of Hefei University of Technology (JZ2018YYPY0289) and the Specialized Research Fund for the Doctoral Program of Higher Education of China (JZ2018HGBZ0186).
- 4.Flash T, Hogan N (1985) The coordination of arm movements: An experimentally confirmed mathematical model. J Neurosci 5(7):1688–1703. https://doi.org/10.1523/JNEUROSCI.05-07-01688.1985 CrossRefGoogle Scholar
- 15.Rohrer B, Fasoli S, Krebs HI, Hughes R, Volpe B, Frontera WR, Stein J, Hogan N (2002) Movement smoothness changes during stroke recovery. J Neurosci 22(18):8297–8304. https://doi.org/10.1523/JNEUROSCI.22-18-08297.2002 CrossRefGoogle Scholar
- 19.Tomasi C, Kanade T (1991) Detection and tracking of point features. Technical Report CMU-CS-91-132, Carnegie Mellon UniversityGoogle Scholar
- 20.Tulyakov S, Alameda-Pineda X, Ricci E, Yin L, Cohn JF, Sebe N (2016) Self-adaptive matrix completion for heart rate estimation from face videos under realistic conditions. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, pp 2396–2404. https://doi.org/10.1109/CVPR.2016.263 Google Scholar