Heart Rate Variability Parameters Extraction Based on Facial Video

Heart rate variability can reflect the activity of the autonomic nervous system, and can quantitatively assess the tension and balance of the sympathetic and vagus nerves of the heart. It is also the most valuable independent predictor of sudden cardiac death. This paper proposes a novel method of heart rate variability parameter extraction based on facial video. It is a non-contact and convenient measurement method. This method first separate 2D facial video information from RGB three channels, normalizing, detrending and bandpass filtering to obtain one-dimensional time series signals. Then, the systolic peak value of one dimension signal is obtained and IBI information is obtained based on the systolic peak values. In this paper, a simple and fast method of heart rate extraction is given based on IBI signal, and the time domain parameters of heart rate variability (mean of IBI, SDNN, RMSSD) and frequency domain parameters (LF, HF, LF /HF) are extracted. Finally, the feasibility of the proposed method is verified by collecting multiple groups of facial video data.