Key points are not available for this paper at this time.
Measuring the heart-rate (HR) of people has multiple applications in telemedicine, Internet-of-Things (IoT), sports, security, etc. However, sometimes it is difficult to use a classic method for measuring HR or the classical method does not scale. This paper presents a solution that works on live video streams and can measure the HR of multiple people at the same time. Face detection combined with object tracking is used to produce a set of face rectangles, which are sampled in the later stages of the pipeline for color variations. The average of the color, in a region of interest (ROI) chosen on the face, represents a signal which corresponds to the heart rate. Using signal processing, a heart rate frequency can be extracted from this signal. The method is quite precise and stable. With this algorithm, four people's faces were detected and their heart rates were measured at the same time, and an error rate of 3-5% can be obtained for the HR. By utilizing object tracking combined with face detection, our method reduces the processing power needed and allows better scaling.
Nădrag et al. (Fri,) studied this question.