The reliability and accuracy of estimating heart-rates from RGB video recorded on a consumer grade camera

Adam Eaton; Vinoin Vincely; Paige Lloyd; Kurt Hugenberg; Karthik Vishwanath

doi:10.1117/12.2252629

3 March 2017 The reliability and accuracy of estimating heart-rates from RGB video recorded on a consumer grade camera

Adam Eaton, Vinoin Vincely, Paige Lloyd, Kurt Hugenberg, Karthik Vishwanath

Proceedings Volume 10055, Optics and Biophotonics in Low-Resource Settings III; 100550Y (2017) https://doi.org/10.1117/12.2252629
Event: SPIE BiOS, 2017, San Francisco, California, United States

Abstract

Video Photoplethysmography (VPPG) is a numerical technique to process standard RGB video data of exposed human skin and extracting the heart-rate (HR) from the skin areas. Being a non-contact technique, VPPG has the potential to provide estimates of subject’s heart-rate, respiratory rate, and even the heart rate variability of human subjects with potential applications ranging from infant monitors, remote healthcare and psychological experiments, particularly given the non-contact and sensor-free nature of the technique. Though several previous studies have reported successful correlations in HR obtained using VPPG algorithms to HR measured using the gold-standard electrocardiograph, others have reported that these correlations are dependent on controlling for duration of the video-data analyzed, subject motion, and ambient lighting. Here, we investigate the ability of two commonly used VPPG-algorithms in extraction of human heart-rates under three different laboratory conditions. We compare the VPPG HR values extracted across these three sets of experiments to the gold-standard values acquired by using an electrocardiogram or a commercially available pulseoximeter. The two VPPG-algorithms were applied with and without KLT-facial feature tracking and detection algorithms from the Computer Vision MATLAB® toolbox. Results indicate that VPPG based numerical approaches have the ability to provide robust estimates of subject HR values and are relatively insensitive to the devices used to record the video data. However, they are highly sensitive to conditions of video acquisition including subject motion, the location, size and averaging techniques applied to regions-of-interest as well as to the number of video frames used for data processing.

Citation Download Citation

Adam Eaton, Vinoin Vincely, Paige Lloyd, Kurt Hugenberg, and Karthik Vishwanath "The reliability and accuracy of estimating heart-rates from RGB video recorded on a consumer grade camera", Proc. SPIE 10055, Optics and Biophotonics in Low-Resource Settings III, 100550Y (3 March 2017); https://doi.org/10.1117/12.2252629

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available