The Novel Estimation Algorithm of Heart Rate Variability and Stress Using Facial Video Analysis

Heart Rate Variability (HRV) can provide extensive information about human health. Calculating HRV requires careful measurement (in milliseconds) of the time interval between each heartbeat. HRV differs among individuals due to varying physiological conditions. We can identify certain pathologies and health issues by accurately detecting changes in these time intervals. HRV is typically detected through ECG signals, however, using video-based Remote Photoplethysmography (rPPG) offers a more convenient and contactless method for calculating HRV. In this study, we propose a system for monitoring HRV using a low-cost camera. We use the Eulerian Video Magnification (EVM) technique. It can magnify subtle changes in the face that cannot be seen by the naked human eye. To improve the accuracy of EVM, a series of signal-processing steps were implemented to enhance the precision of HRV estimations. We evaluate our method on a self-recorded custom dataset and Pulse Rate Detection (PURE) datasets. Compared with other methods, the results show that the proposed method has better performance in HRV analysis.