Pulse Wave And Cardiac Rate Detection Method Research Based On Visible Light Imaging Device With The Motion Artifact Restrained

A method to detect pulse wave and cardiac rate of people in moving states basedon the imaging on forehead with visible light imaging device is proposed in this paper.The fundamental principle used in this paper is PhotoPlethysmoGraphy (PPG). Thechange of forehead region’s gray average reflects the change of light absorption, whichshows the change of blood perfusion of arteriole. Through that, the pulse wave andcardiac rate could be obtained from the gray average wave, and the signal to noise ratiocould be improved through algorithm.A common camera as video-capture device is used in this paper. The camera isdesigned to be mounted in a light shield, and the whole light shield is mounted on theforehead to decrease the motion artifact as much as possible while the video is beingcaptured. And methods about the object region trace matching are also proposed inorder to stabilize the swaying video in the earlier stage of extracting pulse wave andcardiac rate. And then, pulse wave and cardiac rate’s optimizing algorithms aredesigned to improve the signal to noise ratio of pulse wave and the accuracy of cardiacrate in moving states.The main research in this paper contains4parts.1. To solve the problem that the swaying of the detected region affects theextraction of pulse wave and cardiac rate, the crossing center positioning method withtwo-dimensional Gaussion distribution to match the object region trace is proposed inthe paper.2. To solve the problem that the swaying of the detected region affects theextraction of pulse wave and cardiac rate, the template matching method to match theobject region trace is proposed in the paper.3. To solve the problem that the swaying of the detected region lowers the qualityof pulse wave signal, the independent component analysis combining with Part1and2is used to decrease the noise influence and the related results analysis is also conducted.4. To solve the problem that the swaying of the detected region lowers the qualityof pulse wave signal, the wavelet denoising combining with Part1and2is used todecrease the noise influence and the related results analysis is also conducted.While the above research is conducted, finger-clamping pulse wave sensor is usedto obtain contrast signal. By comparing the cardiac rate from above algorithms and from contrast signal, it could be found that the above research’s accuracy rate could reach toa high level. The results show the research in this paper could obtain the pulse waveand cardiac rate accurately when people are in moving states. The further study mayget more physiological signals and bring good news to the routine health monitoring.