Research On Non-contact Heart Rate Measurement Method Based On Embedded Platfor

As an important indicator of human health,heart rate is of great significance for daily physiological health monitoring and preventive diagnosis of cardiovascular diseases.In view of the shortcomings of traditional contact measuring instruments that the operation is complicated,the price is expensive,and the long-term measurement is likely to cause the patient’s skin discomfort.This paper uses the combination of Remote Photoplethysmo Graphy(rPPG)and embedded image processing technology to realize a low-cost,comfortable,non-invasive,simple and convenient non-contact heart rate monitoring system.In this paper,a research on non-contact heart rate monitoring based on rPPG technology is carried out,and a set of embedded heart rate monitoring equipment is developed based on the main noise sources of rPPG technology and specific application scenarios.First,analyze the basic principles of rPPG in depth,take the forehead and one cheek area as the region of interest(ROI),and use the blind source separation method of face recognition and multi-channel data fusion on the basic technical framework of rPPG to improve non-compliance.Contact the accuracy and robustness of the heart rate monitoring system,improve the portability in the embedded platform,and evaluate the performance of the algorithm in this paper by constructing simulation experiments for daily use scenarios.Analyze the basic principles of rPPG technology,build a skin backscatter model based on physiological and optical principles,summarize the formation mechanism and characteristics of blood volume pulse wave(Blood Volume Pulse,BVP),and propose indicators to evaluate the pros and cons of the algorithm.Aiming at the difficulties and challenges faced by rPPG technology,this article improves the performance of the entire heart rate measurement system by improving key technical links.First,the Ada Boost cascade classifier with Haar features is used to recognize the face and select the ROI area to separate the face area from the background area.The gray-level average method is used to extract the rPPG signal in the ROI,and then the initial signal of the BVP is extracted according to the chromaticity characteristics obtained by the multi-channel data fusion.The Fast ICA algorithm is used to separate the BVP signal.Analyze the signal-to-noise ratio of the power spectrum of the separated signal,determine the effective separation component,and select the frequency corresponding to the point with the largest peak in the effective separation component as the heartbeat frequency.The results show that this method can effectively extract and process pulse signals,while suppressing non-periodic interference.Experiments have proved that the algorithm in this paper has good consistency with the heart rate measured by the PPG method.Based on the completion of the algorithm simulation,this paper presents the hardware and software design of the rPPG heart rate monitoring system.The hardware part includes Raspberry Pi 4B development board,USB interface color camera and display screen.The system software is composed of program modules such as video acquisition and video heart rate signal extraction.In order to improve the problem of low frame rate in the signal acquisition process,a multi-threaded acquisition scheme is proposed,and a user interaction interface is designed to realize the user’s independent measurement function.It has been verified that the performance of the system is good,and the system has the same heart rate detection accuracy and robustness as that of the PC.The research results in this article can provide a strong guarantee for remote monitoring and treatment in smart medicine.