Research On Non-contact Heart Rate Measurement Technology

Human heart rate information plays an irreplaceable role in the health field.It can directly reflect the physiological health of the human body,and medically consider heart rate information as one of the important reference indicators of human physiological status.In recent years,research has found that a measurement method called photoplethysmography can achieve non-invasive measurement,and achieve simple,non-invasive measurement of heart rate information by extracting information from the video.Photoplethysmography is realized according to the change principle of blood volume.Based on this principle,this method realizes a non-contact method for measuring heart rate.Using the method of detecting human face and directly tracking the region of interest,the information obtained by the camera is extracted.The heart rate measurement is finally achieved by noise filtering by means of filtering,independent component analysis and the like.Compared with the traditional contact method for measuring heart rate,the method has the advantages of comfort and simplicity,and the implementation cost is low,which is suitable for widespread promotion,and plays an important role in the development of human-computer interaction and medical monitoring in the future.The main work of this paper is as follows:On the one hand,the high-definition camera is used to obtain video continuous frames,and the image sequence is color-enhanced by using the Euler image enlargement algorithm.Considering the many factors that interfere with the target and the ambient light,this thesis considers the characteristics of facial texture and the degree of skin interference,and proposes a signal extraction method based on the bridge of the nose.Experiments show that the signal extraction based on the nasal bridge region can reduce the noise as much as possible.Interference is one of the ideal signal extraction areas.In the non-contact heart rate measurement,how to obtain the region of interest is very important.To solve this problem,this thesis proposes a KCF-based target tracking algorithm to directly track the region of interest.Experiments show that,under quasi-static,based on the positioning method of KCF target tracking algorithm,the continuous ROI region can be obtained quickly and stably,which plays an important role in improving the accuracy of heart rate measurement.According to the designed heart rate measurement system,the influence of measurement angle and distance on the actual heart rate measurement is studied.The experimental results show that the system measurement heart rate is highly correlated and consistent with the actual heart rate.On the other hand,in the actual measurement,the KCF-based target tracking algorithm will not be able to continue heart rate measurement when the target re-enters the video range.In view of this situation,the paper briefly analyzes the situation that may occur during the actual measurement and gives the corresponding solutions.For the problem of relocation after ROI loss,this thesis proposes the method of indirectly positioning the ROI part based on the extraction of the center point of the eye contour,using the Canny operator to perform edge detection on the eye image,calculating the extracted contour center point information,and finally solving the heart rate measurement.The problem of relocation in ROI.Experiments show that this method can relocate new ROI sites after ROI loss,thus achieving continuous uninterrupted heart rate monitoring.Finally,this thesis studies the effects of ROI and measurement position on heart rate measurement,and analyzes the results to illustrate the effectiveness of system heart rate measurement.