Abstract:Efficient non-contact heart rate detection has received much attention in the field of health testing. In previous studies on factors affecting the accuracy of remote photoplethysmography (rPPG) detection, light intensity, facial reflection angle, and motion artifacts have been investigated, but interference from specularly reflected light has not been studied. In this paper, a total of 84 videos were first acquired from four different angles and used to study the effect of specularly reflected light on the accuracy of heart rate estimation. Then we proposed a reflected light angle estimation method which firstly segments the face into 71 regions, and then extracts the angle and illumination of each region as possible reflected light directions and weights respectively. Secondly, all sample points are fitted by sample interpolation, and the angle value corresponding to the maximum value of the weight in the fitted curve is selected as the estimated angle of reflected light direction. Finally, the reflected light angle value is used as a reference for (region of interest) ROI selection to improve the quality of the obtained (image photoplethysmography) iPPG signal. We find that the accuracy of the estimated heart rate and the SNR of the signal are higher when the reflected light angle is larger. The overall root mean squared error (RMSE), standard deviation (SD) and mean error (MAE) of our proposed reflected light angle estimation method are 2.085, 1.745 and 0.7225, which not only have high accuracy but also have good stability. Our proposed method has low error and is superior to existing methods by comparison in publicly available databases.

Ghost imaging-based Non-contact Heart Rate Detection

Remote Heart Rate Measurement from Face Videos under Realistic Situations

Non-contact Measurement of Heart Rate Based on Facial Video

Non-contact Real-time Heart Rate Measurements Based on High Speed Circuit Technology Research

Heart rate measurement based on face video sequence

Remote Heart Rate Estimation in Low-Light Environments Based on Eulerian Video Magnification

Noncontact Heart Rate Measurement Using a Webcam, Based on Joint Blind Source Separation and a Skin Reflection Model: for a Wide Range of Imaging Conditions.

New Advances in Remote Heart Rate Estimation and Its Application to DeepFake Detection

Non-contact Heart Rate Monitoring by Combining Convolutional Neural Network Skin Detection and Remote Photoplethysmography Via a Low-Cost Camera

Remote Heart Rate Measurement From Near-Infrared Videos Based on Joint Blind Source Separation With Delay-Coordinate Transformation

Non-Contact Heart Rate Estimation from Photoplethysmography Using EEMD and Convolution-Transformer Network

Video-Based Heart Rate Measurement: Recent Advances and Future Prospects.

Heart rate estimation based on face video under unstable illumination

High-accuracy heart rate detection using multispectral IPPG technology combined with a deep learning algorithm

Resolution-independent liveness detection via computational ghost imaging

An Efficient Non-Contact Heart Rate Measurement Based on CIELab Color Space

Video-based Human Heart Rate Measurement Using Joint Blind Source Separation

Non-Contact Detection and Estimation of Heart Rate by Improved Chrominance Model

Motion Resistant Facial Video based Heart Rate Estimation Method using Head-Mounted Camera

Remote photoplethysmography based on reflected light angle estimation

Heart rate estimation by leveraging static and dynamic region weights