Research On Non-contact Vital Signs Detection Technology Using Radar Signals

Non-contact vital signs detection technology based on radar signals is a new type of vital detection technology which combines the knowledge of telecommunication?biomedical science and computer science.This technology uses radar signals to extract vital signs such as respiratory frequency and heartbeat frequency without touching the detected targets,which could be widely used for searching and rescuing at earthquake disaster,family medicine,military medicine,etc.Ultra-wideband radar(UWB),due to its lower power consumption,stronger anti-interference and penetration ability,lower cost,has shown its increasing importance on the field of non-contact vital signs detection.Firstly,aiming at reducing the severe noise and clutter interference consisting in the received UWB signals,we use the moving target detection algorithm to remove clutter interference and apply time domain autocorrelation,frequency filter and fast-time mean filter methods to eliminating the noise and eventually improve the signal to noise and clutter ratio.Secondly,due to the fact that the detection targets may move slowly(without much body shakes)during detection process,we propose tracks catching method which can extract the moving tracks of the detection target,then we can obtain the body surface vibration signals using these tracks.After that,we propose the improved parameter estimation algorithm which can accurately obtain the hidden periodic parameters from the body surface vibration signals and these parameters contain the respiratory and heartbeat information.By means of simulation and experiment,the proposed parameters estimation method shows superior performance.Thirdly,we propose two different multi-target non-contact vital signs detection methods which are based on clustering,one uses the distance energy distribution difference of targets,and another uses the distance and frequency domains.The former preprocesses the distance-gate energy signals and clusters the processed data.Then,we can obtain the positions of the multiple targets,take out the body surface vibration signals,and extract vital signs by signal target vital signs detection.The latter method expands the frequency domain of the original signals,processes the two-dimensional signals,and then obtain the positions as well as the frequencies of the detected multiple targets by clustering the processed signals.Both methods show excellent detection accuracy.Comparing these two methods,the former shows higher detection accuracy,but cannot perform well when the distances between the targets and the UWB radar are the same.The latter may get lower accuracy but it can detect the vital signs of the targets with same far from UWB radar.Finally,according to the technologies proposed this paper,a simple and practical real-time monitoring system for non-contact vital signs is built.This system can detect the frequency of respiratory and heartbeat and display the respiratory and heartbeat signal waveform instantly.