Research On No-contact Measurement Of Heart Rate Based On Image Processing

Heart rate,i.e. frequency of heart beating, is a barometer of physiological health assessment. Tachycardia and bradycardia are both abnormal phenomena which are caused by cardiovascular diseases. High resting heart rate is also universally acknowledged as one of four independent risk factors, as well as smoking, hyperlipidemia and hyperviscosity. Keeping better track of heart rate is of great significance to early diagnosis, treatment and prognosis of cardiovascular diseases.Electrocardiogram(ECG) requires patients to wear adhesive gel patches or chest straps that bring inconvenience to patients in psychology or physiology. Furthermore,ECG devices are highly expensive and hard to be installed, so that ECG is not applicable to mobile or continuous heart rate monitoring. All of those disadvantages may hinder diseases from early detection, treatment and being controlled. Thus, the ability to monitor a patient’s heart rate by a remote, simple, low-cost and no-contact means is a tantalizing prospect to promote and extend basic medical services.In each systole and diastole, heart pumps blood to each part of body, meanwhile artery blood volumn periodically changes with it, which result in changes in displacement and color of the human’s skin. Photo Plethysmo Graphy(PPG) takes characteristics of those changes to measure heart rate. This paper records videos of human faces by means of ordinary cameras, and then extracts heart rate from PPG signals by face detection, blind source separation, filtering and discrete Fourier transform.The main works and results of this paper are as follows:(1)It introduces importance in measurement of heart rate, and summarizes existing contact heart rate monitoring and their disadvantages. That leads to necessity of non-contact technology of heart rate monitoring.(2)It carries out the new non-contact technology of heart rate monitoring that combines PPG, image processing and digital signal processing. The experiment’s result shows that the method is feasible and effective.(3)To eliminate robustly noise induced by motion artifacts and improve the accuracy of detection, this paper makes use of independent component analysis(ICA)method to separate independent signals from original mixed PPG signals. On the other hand, because the disorder of ICA components, this paper conducts correlation analysis to find out the component that is most correlated to the green signal.(4)Fast Fourier Transform(FFT) helps to extract heart rate more quickly. And it does not reduce the accuracy and effectiveness.