Non-contact Heart Rate Measurements Based On Video

The heart rate, as one of the important physiological parameters, can effectively and directly reflect the health status and plays an irreplaceable role in human health monitoring. At present, the main method of heart rate measurement needs to contact with human skin, if users keep contacting with the instruments for a long time, it will lead the users feel uncomfortable, and the process of measurement is very complex,it needs users to cooperate,so it has certain limitations.The project is based on the theory of PPG technology, combines with the computer recognition technology to deal with color video frames of human face, extracts the heart rate information based on video, and then extracts the heart rate. The research achieves a comfortable, noninvasive, convenient, and easy to operate heart rate measurements, but no need to change people's daily habits.The main work of this paper is as follows:(1) Extract the heart rate information based on video. Improve the strength of the videobased heart rate information through the Eulerian Video Magnification technology. Analysis face detection and tracking algorithm, and then propose the improved face detection and tracking algorithm, locate the region of interest, and extract the continuous gray value variation of region of interest.(2) Extract the heart rate. Analysis the currently method of extracting heart rate, the standard contact heart rate signal feature and the corresponding power spectrum feature. Propose a new heart rate calculation method combined with the time domain and the frequency domain to deal with the heart rate information to extract heart rate from the video, and then get the heart rate.(3) Achieve non-contact heart rate measurement based on video. According to system requirements to select the hardware equipment and light source, and then achieve the heart rate measurements on Matlab platform.(4) Assesse the agreement. Assessing the agreement between the proposed heart rate measurements method and standard contact heart rate measurements method through the Bland-Altman analysis method.(5) Analysis system optimization results. The improved face detection and tracking method could inhibit the motion article and improve the accuracy of heart rate measurements. Video-based heart rate information is very weak, the project can solve it through the Eulerian Video Magnification technology, and can improve the accuracy of measurements.Experimental results showed that in motion and at rest state the video-based non-contact heart rate measurements method and the standard contact heart rate measurements method had high agreement, it can meet the daily life needs. Through the experimental analysis, the proposed improved face detection and tracking method and Eulerian Video Magnification method could improve the accuracy of the system.