| Cardiovascular disease has become the most fatal disease.Therefore,it is of great significance to measure the physiological parameters closely related to cardiovascular diseases.The current physiological parameter detection equipment mainly uses contact measurement,which is not suitable for special groups such as newborns and burn patients.In recent years,imaging photoplethysmography(IPPG)technology developed on the basis of traditional photoplethysmography technology has become a research hotspot.This technology uses a common camera to capture video of human skin tissue and extracts physiological parameters from the video.It has the advantages of non-contact,convenient operation,and low cost.Although there have been some testing equipment based on IPPG technology,the computing platform of these equipment is usually a computer or a mobile phone,which has a high cost and is not suitable for long-term monitoring.The few testing equipment that use lower-cost embedded systems as computing platforms have problems such as fewer types of measurement parameters,low real-time performance,and inconvenient data storage.In this paper,researches on these issues are carried out,and a non-contact multi-physiological parameter monitoring system running on an embedded computing platform is designed and implemented,which realizes real-time monitoring of respiratory rate,heart rate and blood oxygen saturation.The work of this paper mainly includes the following aspects:1.Starting from physiology and optics,the principle of IPPG technology is deeply studied,the source of respiratory signal and heart rate signal is explained,and the calculation model of blood oxygen saturation is deduced in detail.In addition,various noises that affect the quality of physiological signals and the calculation accuracy of physiological parameters are analyzed,and the existing denoising techniques are summarized.2.From a physiological point of view,the parts of the human face suitable for physiological parameter detection are analyzed,and finally the forehead and both cheeks are selected as the detection areas.At the same time,a color camera is used to capture human facial video,and the automatic positioning of the detection area is realized through facial landmark detection.On this basis,detection algorithms are designed for respiratory rate,heart rate,and blood oxygen saturation,and the algorithms are ported to the embedded system.Experiments show that algorithms designed in this paper can run on the embedded system at a detection rate of once per second,and has high accuracy.3.The hardware components of the monitoring system are selected,and the embedded software development environment is built.On this basis,the software running in multithreading mode is designed for the monitoring system to make full use of the computing resources.Furthermore,a cloud platform is built,and the remote storage and access of data is realized by uploading the monitoring data to the cloud platform.4.In order to verify the performance of the monitoring system,reference instruments are selected for various physiological parameters,and the hardware and software environment of the experiment is built.And 15 volunteers are recruited to carry out the comparative experiment in multiple scenarios.Bland Altman analysis,root mean square error and average accuracy are used to analyze the consistency between the measured value and the reference value.The results show that the monitoring system can meet the need for real-time monitoring in daily situations. |
PDF Full Text Request