Research On Noncontact Physiological Parameter Monitoring System

In recent years,cardiovascular disease has become the biggest killer of human life and health.It is very important to prevent and monitor cardiovascular disease.The traditional measurement technology has become mature,but its disadvantage is that it needs contact measurement,which greatly constrains the patients,and it is only suitable for a single measurement,and can not carry out 24-hour continuous monitoring.However,non-contact measurement can overcome these pain points.At present,there are many ways to realize non-contact measurement,and imaging photoelectric plethysmography is one of the typical methods.In this method,the tiny changes of the reflected light of human skin with the heartbeat are captured by a color camera,and then the physiological parameters such as respiration,heart rate,blood pressure and oxygen saturation are obtained by denoising,amplification and other algorithms.At present,the non-contact physiological parameter detection system using this technology still has the disadvantages of high power consumption,not portable,high cost,inaccurate measurement,single measurement signal and so on.In order to improve the accuracy of heart rate,pulse wave variability,blood pressure and other physiological parameters measurement,reduce the cost and power consumption,and improve the portability,this paper focuses on how to obtain higher signal-to-noise ratio of pulse wave signal,and build an embedded software and hardware platform to achieve physiological parameters monitoring.The main work of this paper is as follows1.This paper compares the current mainstream face detection algorithms,and selects AdaBoost algorithm for face detection.By comparing the original pulse wave signals of different parts of the face,the cheek part is selected as the ROI area.Through a large number of comparative experiments,finally use RGB three channel signal for independent component analysis denoising and seven layer wavelet packet denoising to obtain high signal-to-noise ratio pulse wave signal.2.This paper uses hkg-07b pulse measurement sensor to build photoelectric plethysmography equipment as reference equipment.The cubic spline interpolation method is used to process the signal of 30 frame camera system.The pulse wave variations measured by 120 frame camera system and 30 frame camera system were compared.The results show that the camera frame rate has little influence on the measurement results,and the measurement accuracy can reach 89% compared with PPG system.Low frame rate camera can be used to measure heart rate variability.The embedded measurement system uses a 30 frame camera,which can double the speed of calculation.3.In order to get more accurate blood pressure measurement,this paper uses phase based method to calculate the pulse wave time difference between face and hand.Pulse wave conduction time,heart rate and other parameters are used as input signals of neural network to predict blood pressure.The results show that the accuracy of blood pressure prediction is 92.6% under the condition of sufficient light and no large face shaking,which can be used to measure blood pressure more accurately.4.A non-contact physiological parameter monitoring system is designed.Including schematic design,PCB design,high-speed signal simulation,embedded system transplantation,development environment construction.The a311 d SOC of Jingchen electronics is used as the main control core of the system,QT is used as the development platform,window programming is carried out,and heart rate,pulse wave variation and blood pressure can be calculated.At the same time,it can connect to the cloud server and upload the physiological parameter test data of the tested personnel.