Research On Non-contact Measurement Method Of Human Heart Rate Based On Video

Heart rate is an important parameter that reflects the psychological and physiological state of human,the measurement of heart rate has a very important role for daily health care.Because of heartbeat cause blood volume changes,making the human skin changes the color of the reflected light,the magnitude of the color change which showed in the color of the reflected light is proportional to the intensity of the light source.The paper is based on the principle of photoplethysmography,this paper proposed a method that measures heart rate adaptively with non-contact camera.Firstly,face video is captured using an ordinary camera.Then,the face image sequence come pre-treated,improving image quality.Next,the region of interest is extracted from the denoised face image.Then,the heartrate signal is obtained from the region of interest used by adaptive remote photoplethysmography algorithm.Finally,we extract the effective peaks in the heart rate singal by using the peak detection method proposed in this paper and compute heart rate.The main research work of this paper has the following aspects:(1)Getting video with face images.Using an ordinary optical camera to shoot a video containing face images under natural light.Then,extract the appropriate face area from face position which detected by face-detection technology.The extracted appropriate face will be used as follow-up pulse signal extranction area.(2)Denoising images,The quality of the image has a decisive effect on the experimental results.This paper uses the wavelet soft threshold denoising method.(3)An adaptive non-contact heart rate signal algorithm based on photoplethysmography was proposed to solve the problem what is different people with different ranges of heart rate.,(4)The paper design algorithm,compared with the traditional PCA,Fast ICA algorithm,and using Bland-Altman consistency method for evaluation.The experimental results show that the adaptive non-contact heart rate signal algorithm which based on photoplethysmography in this paper can accurately measure the heart rate value of the human body.It can be seen from the measurement results and the Bland-Altman consistency analysis that all the measurement results are 95% within the confidence interval.And the absolute error of the measurement result is less than 5.