Non-contact Measurement of Heart Rate Based on Facial Video

Heart rate refers to the number of heartbeats per minute, which is one of the most important indicators to estimate human health. Traditionally, the heart rate can be measured by using a stethoscope or electrocardiogram, or even by feeling the pulse. In addition, a growing number of wearable devices are invented to detect human heart rate. However, these methods rely on physical contact, which makes the measurement inconvenient to some extent. This paper discusses the method of non-contact heart rate measurement based on Photoplethysmography (PPG) by capturing human face from the video under ordinary light. With this method, the change of light is utilized to perceive the volume pulse wave. In this paper, a webcam embedded in laptop is used to shoot facial videos under ordinary light conditions. With the Region of Interest (ROI) selected, we can first extract the initial data of three channels of RGB and then convert the color space of the image from RGB to YCbCr. Eulerian Video Magnification (EVM) is used to magnify the motion. In EVM, the converted results are decomposed in the spatial domain and filtered in temporal domain. By collapsing the amplified signals to the original ones, we can obtain the final signal waveform to estimate the heart rate. After the consistency analysis with the traditional method, it is found out that the error of the results obtained by our method is relatively small. All the experimental results indicated that the method to measure heart rate proposed in this paper is effective.