Non-contact Breathing Measurement Method And Its Implementation For Sleep Monitoring

With the continuous development of society,people's living standards have been increasing,and more and more people have begun to pay attention to their own health conditions.The traditional mode of monitoring human physiological indicators is mainly hospital-centered,but now it is gradually focusing on mainly-preventable and the familyoriented,miniaturized development model.Therefore,a portable nursing instrument capable of real-time and dynamic monitoring of human physiological indicators will be the first choice for future prevention of family health and personal diseases.Relevant studies have shown that the body's frequency,intensity,rhythm and other sleep respiratory signals can largely reflect the health status of the body,revealing the emergence of the disease.Therefore,it has very important medical significance for the monitoring of sleep and respiratory activity in the human body.Most medical respiratory monitoring devices today have some drawbacks: they are relatively bulky and expensive,and they are all contact-type tests.Contact detection not only gives the patient a sense of restraint and affects the detection results,but it is even more inconvenient for special patients(infants and the elderly).Therefore,this paper proposes a noncontact sleep-breathing measurement program,which uses an infrared camera as a detection sensor and uses Gaussian mixture background differential algorithm and non-parametric histogram segmentation algorithm to accurately obtain the human sleep-breathing signal.The measuring device of this solution is not only inexpensive,practical,but also non-contact with the human body.The main research work of this paper is as follows:(1)Sleep-breath motion detection based on inter-frame difference.Through experiments,it is found that the simplest inter-frame difference algorithm is used to detect respiratory motion,the effect is not obvious,and there is a problem that the detection is not continuous.For this reason,this paper studies and implements several background modeling difference algorithms: based on KNN(K-Nearest Neighbor)background modeling difference,based on LBP(Local Binary Pattern)background modeling difference and based on Gaussian mixture background modeling difference.The background modeling differential algorithm principle: Establish a background model frame for a certain number of video frame images,and then use the difference between the current frame and the background frame to detect the moving object.The background modeling differential algorithm can well solve the problem of detecting discontinuous breathing target.(2)In this paper,the Gaussian mixed background difference algorithm and the nonparametric histogram analysis algorithm are combined to realize the measurement of sleep breathing.The traditional histogram segmentation algorithm uses threshold segmentation,parameter estimation,etc.It is difficult to make a good analysis for different populations(infants,elderly,disease patients)with significant differences in respiratory intensity and frequency.So the paper uses an improved non-parametric histogram segmentation algorithm to achieve respiratory measurements.(3)A non-contact sleep-breathing monitoring program was proposed in this paper.Through a large number of experimental analysis,this system was implemented using Gaussian mixture background differential,non-parametric histogram segmentation,sleep-breathing frequency mathematical modeling and interface.Through theoretical research on the proposed method for detection of human sleepbreathing signals,a large number of experiments have been validated and analyzed to achieve an efficient,convenient,and economical sleep-breathing signal detection system.Using this system to measure sleep breathing in humans,experiments show that it can effectively detect the body's sleep and breathing movements.In addition,this paper realizes the real-time display of the human sleep-breathing frequency line chart,which can display the respiratory status more intuitively.