CN105919568A - Gabor transformation based method and device for extracting and analyzing breathing and heartbeat signals - Google Patents

Abstract

The invention provides a Gabor transformation based method and device for extracting and analyzing breathing and heartbeat signals. The method comprises following steps: S1, radar echo signals are subjected to Gabor transformation, and time-frequency power distribution signals of the transformed signals are calculated; S2, the time-frequency power distribution signals are subjected to frequency spectrum weighted average operation, and a frequency spectrum centroid changing curve is obtained; S3, the frequency spectrum centroid changing curve is subjected to filtering operation, and breathing signals and heartbeat signals are obtained; S4, the breathing signals and the heartbeat signals are subjected to the Gabor transformation, and breathing frequency changing with time and heartbeat frequency changing with the time are obtained. With the adoption of the method and the device, the breathing frequency and heartbeat frequency changing with time, the signals and the frequency change of the signals can be effectively extracted.

Description

The extraction and analytical method of breathing and heartbeat signal of based on the conversion of gal cypress and device

Technical field

The present invention relates to processing of biomedical signals field, particularly relate to a kind of breathing based on the conversion of gal cypress Extraction and analytical method and device with heartbeat signal.

Background technology

Bioradar is a kind of by launching reception microwave or THz wave, to human body breathing and heartbeat etc. Life parameters carries out the medical apparatus and instruments of non-contact measurement, bioradar can replace traditional electrocardiogram, The contact type measurement equipment such as breathing zone, in addition to can be applicable to medical diagnosis on disease, also act as human health status Monitoring in real time, thus become the focus of research outside Present Domestic.

In reality, the human body echo received by bioradar causes with vital movements such as breathing, heartbeats Body surface fine motion is modulated, and is in particular in time dependent micro-doppler frequency, and traditional based on Fu In the signal analysis method of leaf transformation the Joint Distribution information of time-domain and frequency domain cannot be provided, thus nothing Method extracts time dependent breathing and palmic rate.

Summary of the invention

The technical problem to be solved is to provide one and can extract in radar echo signal respectively Breath signal and heartbeat signal, and time dependent respiratory rate and time dependent can be obtained The breathing based on the conversion of gal cypress of palmic rate information and the extraction and analytical method of heartbeat signal and device.

In order to solve above-mentioned technical problem, the invention provides following technical scheme:

A kind of breathing based on the conversion of gal cypress and the extraction and analytical method of heartbeat signal, described method is based on gal Bai Bianhuan obtains heartbeat signal and breath signal from the radar echo signal returned and is analyzed, described Method comprises the following steps:

S1: described radar echo signal is carried out gal cypress conversion, and calculates the time-frequency merit of the signal after conversion Rate distribution signal；

S2: described time frequency energy distribution signal is carried out frequency spectrum weighted average operation, it is thus achieved that spectral centroid becomes Change curve；

S3: described spectral centroid change curve is filtered operation, to obtain breath signal and heartbeat letter Number；

S4: described breath signal and heartbeat signal are carried out again gal cypress conversion, and acquisition changes over Respiratory rate and time dependent palmic rate.

As preferably, also include step S0 before step S1: the signal received is carried out Doppler and adopts Sample, and carry out quadrature demodulation, it is thus achieved that described radar echo signal.

As preferably, described step S2 may further comprise: and drops described spectral centroid change curve Sampling；Step S3 is that the spectral centroid change curve after down-sampled is filtered operation, to obtain breathing Signal and heartbeat signal.

As preferably, multiple down-sampled in described step S2 is, wherein f_sFor many General Le sample frequency,Represent and round operator.

As preferably, described step S3 farther includes: respectively by breath signal bandpass filter and the heart Jump signal bandpass filter described spectral centroid change curve to be filtered, to obtain breath signal respectively And heartbeat signal.

As preferably, described step S4 farther includes: enter described breath signal and heartbeat signal respectively Row gal cypress converts, and calculates the time frequency energy distribution signal of described breath signal and heartbeat signal, and leads to Cross detection power maximum and extract time dependent respiratory rate and time dependent palmic rate.

As preferably, the method extracting time dependent respiratory rate in described step S4 includes following step Rapid:

S41: the breath signal that step S3 is obtained carry out gal cypress conversion and calculate conversion after the time-frequency of signal Power distribution signal；

S42: the time frequency energy distribution signal obtaining step S41 is extracted by power maximum detection method Time dependent respiratory rate；

As preferably, the method extracting time dependent palmic rate in described step S4 includes following step Rapid:

S43: the heartbeat signal that step S3 is obtained carry out gal cypress conversion and calculate conversion after the time-frequency of signal Power distribution signal；

S44: the time frequency energy distribution signal obtaining step S43 is extracted by power maximum detection method Time dependent palmic rate.

Present invention also offers the extraction and analysis device of a kind of breathing based on the conversion of gal cypress and heartbeat signal, The application of described device breathing based on the conversion of gal cypress and the extraction and analytical method of heartbeat signal as mentioned above；And Described device at least includes filter unit, and it is configured to described spectral centroid change curve is filtered behaviour Make, to obtain breath signal and heartbeat signal.

As preferably, described filter unit includes the breath signal bandpass filter obtaining breath signal and obtains Heartbeat signal bandpass filter to heartbeat signal..

Compared with prior art, the beneficial effects of the present invention is:

1, analyzing method compared to traditional based on Fourier transformation, technical solution of the present invention uses gal cypress to become Change Time-Frequency Analysis Method, can effectively extract time dependent breathing and signal and frequency change thereof；

2, the present invention is after obtaining spectral centroid change curve, then carries out frequency division during second time gal cypress conversion Before analysis, signal data can be carried out down-sampled operation, be greatly improved signal transacting efficiency.

Accompanying drawing explanation

Fig. 1 is that the bioradar of the embodiment of the present invention measures human body respiration and heartbeat schematic diagram；

Fig. 2 is the extraction and analysis side of the breathing based on the conversion of gal cypress in the embodiment of the present invention and heartbeat signal The flow chart of method；

Fig. 3 is the breathing in the conversion of gal cypress in another embodiment of the present invention and the extraction and analysis of heartbeat signal The flow chart of method；

Fig. 4 is the principle flow chart obtaining time dependent respiratory rate in the embodiment of the present invention；

Fig. 5 is the principle flow chart obtaining time dependent palmic rate in the embodiment of the present invention；

Fig. 6 is the breathing based on the conversion of the gal cypress extraction and analysis device with heartbeat signal of the embodiment of the present invention Theory diagram.

Description of reference numerals

1-converter unit 2-weighted units

3-filter unit 4-pretreatment unit

The down-sampled unit of 5-

Detailed description of the invention

Below, in conjunction with accompanying drawing, embodiments of the invention are described in detail, but are not intended as the present invention Restriction.

It should be noted that in accompanying drawing or specification describe, similar or identical part all uses identical Figure number.The implementation not illustrated in accompanying drawing or describe, for those of ordinary skill in art Known form.It addition, although the demonstration of the parameter comprising particular value can be provided herein, it is to be understood that Parameter is worth equal to corresponding without definite, but can approximate in acceptable error margin or design constraint In corresponding value.

The invention provides a kind of breathing based on gal cypress conversion (Gabor Transform) and heartbeat signal Extraction and analytical method and device, therefore, before the details of embodiments of the present invention is discussed in detail, Some concepts and the principle of gal cypress conversion are first briefly described.

Jia Bai conversion is a kind of typical Time-Frequency Analysis Method, is to be divided into countless by a non-stationary signal Individual short time period, by the signal hypothesis in each time period for carrying out Fourier's change in the case of smoothly Obtain the frequency characteristic of signal in this time period.If with the time-frequency plane that T/F builds for axis being (t, f), further, samples with time shifting constant T and frequency shifting constant F to time-frequency plane, (m, n), wherein m is time shift method, and n is the frequency modulation(PFM) factor, then signal s (t) can to build two dimensional surface Can be with gal cypress expansion coefficient a_mnIt is expressed as

$s\left(t\right)=\underset{m=-\mathrm{\∞}}{\overset{\mathrm{\∞}}{\Σ}}\underset{n=-\mathrm{\∞}}{\overset{\mathrm{\∞}}{\Σ}}{a}_{mn}{g}_{mn}\left(t\right),$

Wherein, g_mnT () is gal cypress basic function, be to be obtained through time shift and frequency displacement by window function g (t), its expression formula For

$g_{mn}\left(t\right)=g(t-mT)e^{j2\mathrm{\π}\left(nF\right)t},FT\≤1,$

a_mnBeing gal cypress expansion coefficient, its expression formula is

$g_{mn}\left(t\right)={\∫}_{-\mathrm{\∞}}^{\mathrm{\∞}}s\left(t\right){\mathrm{\γ}}_{mn}^{*}\left(t\right)dt,$

Wherein, γ_mnT () is gal cypress basic function g_mnThe dual function of (t),Represent γ_mnThe conjugation of (t).Gal (m, n) any coefficient upper are (t, pass f) by it with time-frequency plane to cypress expansion coefficient i.e. two dimensional surface System is apparent from, and the time-frequency gal cypress conversion expression formula of signal s (t) is

S (mT, nF)=a_mn.

Its power spectrum is

P (mT, nF)=| S (mT, nF) |².

Jia Bai conversion belongs to the Time-Frequency Analysis Method of Moving Window, and wherein window function is chosen as Gaussian window, its table Reaching formula is

$g\left(t\right)=\sqrt[4]{2}\exp (-{\mathrm{\πt}}^2).$

Fig. 1 is that the bioradar according to the embodiment of the present invention measures human body respiration and heartbeat schematic diagram.Due to Human body respiration and heartbeat fine motion are all the changes in nearly cycle, breathe and are completed by contraction and the expansion in thoracic cavity, Can be approximated to be sine wave oscillations model, it is the same that heartbeat signal is similar to signal in electrocardiogram, can be approximately arteries and veins Punch die type.Note radar operating frequency is f_c(corresponding wavelength λ), human body target is R, then radar away from radar The echo that transmitted waveform is returned through human body target scattering is after quadrature demodulation

$S_r\left(t\right)={\mathrm{Ae}}^{-j\frac{4\mathrm{\π}R}{\mathrm{\λ}}}{e}^{-j\frac{4\mathrm{\π}\[acos\left(2{\mathrm{\πf}}_{b}t\right)\]}{\mathrm{\λ}}}{e}^{-j\frac{4\mathrm{\π}\[b\mathrm{\δ}\left(2{\mathrm{\πf}}_{h}t\right)\]}{\mathrm{\λ}}}+N\left(t\right)$

Wherein A is radar echo signal amplitude, a and b is respectively breathing fluctuating and the heartbeat vibration width of human body target Degree, f_bAnd f_hIt is respectively breathing of human body target to rise and fall and the frequency of heartbeat vibration, a cos (2 π f_bT) represent Based on Sinusoid Model with a as amplitude, f_bFor the respiratory movement of frequency, b δ (2 π f_hT) represent based on arteries and veins Rush model of vibration with b as amplitude, f_hHeartbeat for frequency is vibrated, and N (t) represents noise signal.Then exhale Inhaling with heartbeat signal analysis is to analyze time dependent breathing and heartbeat signal in real time based on above-mentioned echo, Extract and breathe and heartbeat signal frequency.

Below, the specific embodiment of the present invention is described in detail.As described in Figure 2, for the present invention A kind of based on breathing and the heartbeat signal of based on the conversion of gal cypress in the embodiment of the present invention in embodiment carries Take the flow chart of analysis method.Wherein, radar antenna sends radar signal to human body target, and reception is returned The echo-signal returned, embodiment of the present invention purpose is for process described echo-signal, and extracts wherein About breath signal and the information of heartbeat signal, such as time dependent respiratory rate and time dependent Palmic rate, described method may comprise steps of:

S1: radar echo signal is carried out gal cypress conversion, and the time-frequency power calculating the signal after conversion divides Cloth signal P_r(t,f)；Preferably, before step S1, also include step S0: radar antenna irradiates people Body target with pulse recurrence frequency f_sLaunch and receive echo-signal, after quadrature demodulation, obtaining Doppler Sample rate is f_s, time span be the radar echo signal s of T_r(t)；Wherein, pulse recurrence frequency f_sI.e. Doppler sample frequency is much larger than 3Hz, such as, f_sCan be 300Hz；

S2: time frequency energy distribution signal is carried out frequency spectrum weighted average operation, it is thus achieved that spectral centroid change song Line g (t)；Wherein, as can according to the following formula to time frequency energy distribution signal along frequency to carrying out frequency spectrum weighting Average operation:

$g\left(t\right)=\frac{{\∫}_{F_1}^{F_2}f\×P_r(t,f)df}{{\∫}_{F_1}^{F_2}{P}_r(t,f)df},$

Wherein, P_r(t, is f) the time frequency energy distribution signal that obtains of described step S1, and t represents that time, f represent Integrating range [F in frequency, molecule and denominator₁,F₂] represent that the frequency of this time-frequency distributions signal is initial and whole Only scope, g (t) is the spectral centroid change curve obtained；

S3: spectral centroid change curve g (t) is filtered operation, to obtain breath signal e_b(t) and the heart Jump signal e_h(t)；This filtering operation can be by spectral centroid change curve g (t) respectively by breath signal band Bandpass filter and heartbeat signal bandpass filter are filtered operation respectively, to obtain breath signal respectively e_b(t) and heartbeat signal e_h(t)；

S4: to breath signal e_b(t) and heartbeat signal e_hT () carries out gal cypress conversion again, and obtain in time Respiratory rate F of change_b(t) and time dependent palmic rate F_h(t).In the present embodiment, can be right Breath signal e_b(t) and heartbeat signal e_hT () carries out gal cypress conversion respectively, and calculate breath signal and heartbeat The time frequency energy distribution signal of signal, and extract time dependent breathing by detection power maximum Frequency and time dependent palmic rate.

Based on above-mentioned steps, embodiments of the invention i.e. can be by carrying out twice gal to radar echo signal Bai Bianhuan, extracts breath signal and heartbeat signal respectively, and time dependent breath signal frequency and Heartbeat signal frequency.

In another embodiment as shown in Figure 3, step S2 can further include: to step Spectral centroid change curve g (t) obtained in S1 carries out N times of down-sampled operation, obtain down-sampled after frequency Spectrum barycenter change curve g_d(t)；

Described down-sampled multiple, whereinRepresent and round operator, such as f_sFor 300Hz Time, the most down-sampled multiple N takes 20.

Correspondingly, step S3 is then to the spectral centroid change curve g after down-sampled_dT (), uses respectively Breath signal bandpass filter and heartbeat signal bandpass filter are filtered operation, obtain breath signal e_b(t) and heartbeat signal e_h(t)；Owing to having carried out down-sampled operation, can be greatly improved at signal Reason efficiency.

Wherein, the effect of the breath signal bandpass filter in the present embodiment is to extract breath signal composition, It is have upper cut-off frequecy of passband be 0.6Hz, upper stopband cut-off frequency be 0.9Hz, lower passband cutoff frequency Rate is 0.15Hz, lower stopband cut-off frequency is 0.02Hz, passband ripple is 0.1dB, stopband attenuation is 80dB 6 rank infinite-duration impulse response bandpass digital filters；The effect of heartbeat signal bandpass filter is to extract the heart Jump signal component, its be have upper cut-off frequecy of passband be 2.2Hz, upper stopband cut-off frequency be 2.4Hz, Lower cut-off frequecy of passband is 0.7Hz, lower stopband cut-off frequency is 0.5Hz, passband ripple is 0.1dB, resistance Band decays to the 8 rank infinite-duration impulse response bandpass digital filters of 80dB.

Additionally, the breath signal bandpass filter used of step S3 described in above-described embodiment and heartbeat letter Number bandpass filter is not limited in the various concrete forms mentioned in embodiment, the common skill of this area It can be carried out replacing, such as with knowing simply by art personnel:

(1) breath signal bandpass filter is in addition to can using infinite-duration impulse response bandpass digital filter, also Can to be the form of finite impulse response bandpass digital filter, if other in addition to filter order Parameter keeps constant；

(2) heartbeat signal bandpass filter is in addition to can using infinite-duration impulse response bandpass digital filter, also Can to be the form of finite impulse response bandpass digital filter, if other in addition to filter order Parameter keeps constant.

It addition, be illustrated in figure 4 in the embodiment of the present invention principle obtaining time dependent respiratory rate Flow chart.Including following steps:

S41: the breath signal e that step S3 is obtained_b(t) carry out gal cypress conversion and calculate conversion after signal Time frequency energy distribution signal P_b(t,f)；

S42: the time frequency energy distribution signal P that step S41 is obtained_b(t f) is detected by power maximum Method extracts time dependent respiratory rate F_b(t), wherein

$F_b\left(t\right)=\underset{f}{\mathrm{argmax}}{P}_b(t,f)$

It is illustrated in figure 5 in the embodiment of the present invention principle flow chart obtaining time dependent palmic rate.

S43: the heartbeat signal e that step S3 is obtained_h(t) carry out gal cypress conversion and calculate conversion after signal Time frequency energy distribution signal P_h(t,f)；

S44: the time frequency energy distribution signal P that step S43 is obtained_h(t, f) by power maximum detection side Method extracts time dependent palmic rate F_h(t), wherein

$F_h\left(t\right)=\underset{f}{\mathrm{argmax}}{P}_h(t,f).$

It addition, present invention also offers the extraction and analysis of a kind of breathing based on the conversion of gal cypress and heartbeat signal Device, as shown in Figure 6, for the breathing based on the conversion of gal cypress in the embodiment of the present invention and heartbeat signal The theory diagram of extraction and analysis device.Wherein may include that converter unit 1, weighted units 2, filtering list Unit 3, wherein converter unit 1 is configured to radar echo signal and carries out gal cypress conversion, and after calculating conversion The time frequency energy distribution signal of signal；Weighted units 2 is configured to time frequency energy distribution signal is carried out frequency spectrum Weighted average operation, it is thus achieved that spectral centroid change curve；Filter unit 3 is configured to change spectral centroid Curve is filtered operation, to obtain breath signal and heartbeat signal respectively；Wherein converter unit 1 also may be used Again carry out gal cypress conversion being further configured to breath signal that filter unit is obtained and heartbeat signal, And obtain time dependent respiratory rate and time dependent palmic rate.

Additionally, it is preferred that, the present embodiment can also include a pretreatment unit 4, it is to radar antenna The signal received carries out Doppler sample, and carries out quadrature demodulation, it is thus achieved that Doppler sample rate is f_s, time Between echo-signal s of a length of T_r(t)；Its medium-PRF f_sMuch larger than 3Hz, such as, f_sCan Think 300Hz.Wherein radar signal receives and the frequency of transmission signal is f_s。

By weighted units 2, time frequency energy distribution signal is carried out frequency spectrum weighted average operation again and obtain frequency After spectrum barycenter change curve, it is also possible to by down-sampled unit 5, frequency spectrum intelligence letter change curve is carried out fall and adopt Sample, to strengthen calculation process speed.Filter unit 3 can be filtered operation to the data after down-sampled, To obtain breathing and heartbeat signal respectively.Wherein, down-sampled multiple, whereinTable Show and round operator, such as f_sDuring for 300Hz, the most down-sampled multiple N takes 20.

Filter unit 3 in the present embodiment may further include the breath signal band of acquisition breath signal and leads to Wave filter and heartbeat signal bandpass filter, wherein, the effect of breath signal bandpass filter is to extract to exhale Inhale signal component, be that there is upper cut-off frequecy of passband 0.6Hz, upper stopband cut-off frequency 0.9Hz, lower passband Cut-off frequency 0.15Hz, lower stopband cut-off frequency 0.02Hz, passband ripple 0.1dB, stopband attenuation 80dB 6 rank infinite-duration impulse response bandpass digital filters.The effect of heartbeat signal bandpass filter is to extract the heart Jump signal component, be that there is upper cut-off frequecy of passband 2.2Hz, upper stopband cut-off frequency 2.4Hz, lower passband Cut-off frequency 0.7Hz, lower stopband cut-off frequency 0.5Hz, passband ripple 0.1dB, stopband attenuation 80dB 8 rank infinite-duration impulse response bandpass digital filters.

But, the breath signal bandpass filter used in above-described embodiment and heartbeat signal bandpass filter Being not limited in the various concrete forms mentioned in embodiment, those of ordinary skill in the art can be to it Carry out replacing, such as with knowing simply:

(1) breath signal bandpass filter is in addition to can using infinite-duration impulse response bandpass digital filter, also Can to be the form of finite impulse response bandpass digital filter, if other in addition to filter order Parameter keeps constant；

(2) heartbeat signal bandpass filter is in addition to can using infinite-duration impulse response bandpass digital filter, also Can to be the form of finite impulse response bandpass digital filter, if other in addition to filter order Parameter keeps constant.

Breath signal for obtaining filter unit 3 and heartbeat are believed by converter unit 1 in the present embodiment again When number again carrying out the conversion of gal cypress, can extract time dependent by the method for detection power maximum Respiratory rate and time dependent palmic rate.Specifically may include that

The breath signal e that step filter unit 3 is obtained_b(t) carry out gal cypress conversion and calculate conversion after signal Time frequency energy distribution signal P_b(t,f)；And to the time frequency energy distribution signal P obtained_b(t f) passes through power Maximum detection method extracts time dependent respiratory rate F_b(t), i.e.

$F_b\left(t\right)=\underset{f}{\mathrm{argmax}}{P}_b(t,f);$

And the heartbeat signal e that filter unit 3 obtained_h(t) carry out gal cypress conversion and calculate conversion after signal Time frequency energy distribution signal P_h(t,f)；And to the time frequency energy distribution signal P obtained_h(t f) passes through power Maximum detection method extracts time dependent palmic rate F_b(t), i.e.

$F_h\left(t\right)=\underset{f}{\mathrm{argmax}}{P}_h(t,f).$

By above-mentioned embodiment, i.e. can effectively realize the present invention and extract time dependent breathing frequently Rate and the purpose of palmic rate.

Particular embodiments described above, is carried out the purpose of the present invention, technical scheme and beneficial effect Further describe, be it should be understood that the specific embodiment that the foregoing is only the present invention, Be not limited to the present invention, all within the spirit and principles in the present invention, any amendment of being made, etc. With replacement, improvement etc., should be included within the scope of the present invention.

Claims (10)

1. a breathing based on the conversion of gal cypress and the extraction and analytical method of heartbeat signal, it is characterised in that described method obtains heartbeat signal and breath signal from the radar echo signal returned based on the conversion of gal cypress and is analyzed, and said method comprising the steps of:

S1: described radar echo signal is carried out gal cypress conversion, and calculates the time frequency energy distribution signal of the signal after conversion；

S2: described time frequency energy distribution signal is carried out frequency spectrum weighted average operation, it is thus achieved that spectral centroid change curve；

S3: described spectral centroid change curve is filtered operation, to obtain breath signal and heartbeat signal；

S4: described breath signal and heartbeat signal are carried out gal cypress conversion again, and obtains time dependent respiratory rate and time dependent palmic rate.

Method the most according to claim 1, it is characterised in that also include step S0 before step S1: the signal received is carried out Doppler sample, and carries out quadrature demodulation, it is thus achieved that described radar echo signal.

Method the most according to claim 2, it is characterised in that described step S2 may further comprise: and carries out down-sampled to described spectral centroid change curve；Step S3 is that the spectral centroid change curve after down-sampled is filtered operation, to obtain breath signal and heartbeat signal.

Method the most according to claim 3, it is characterised in that multiple down-sampled in described step S2 isWherein f_sFor Doppler sample frequency,Represent and round operator.

Method the most according to claim 3, it is characterized in that, described step S3 farther includes: be filtered described spectral centroid change curve by breath signal bandpass filter and heartbeat signal bandpass filter respectively, to obtain breath signal and heartbeat signal respectively.

Method the most according to claim 1, it is characterized in that, described step S4 farther includes: described breath signal and heartbeat signal are carried out gal cypress conversion respectively, and calculate the time frequency energy distribution signal of described breath signal and heartbeat signal, and extract time dependent respiratory rate and time dependent palmic rate by detection power maximum.

Method the most according to claim 6, it is characterised in that the method extracting time dependent respiratory rate in described step S4 comprises the following steps:

S41: the breath signal that step S3 is obtained carry out gal cypress conversion and calculate conversion after the time frequency energy distribution signal of signal；

S42: the time frequency energy distribution signal obtaining step S41 extracts time dependent respiratory rate by power maximum detection method.

Method the most according to claim 6, it is characterised in that the method extracting time dependent palmic rate in described step S4 comprises the following steps:

S43: the heartbeat signal that step S3 is obtained carry out gal cypress conversion and calculate conversion after the time frequency energy distribution signal of signal；

S44: the time frequency energy distribution signal obtaining step S43 extracts time dependent palmic rate by power maximum detection method.

9. a breathing based on the conversion of gal cypress and the extraction and analysis device of heartbeat signal, it is characterised in that the application of described device is breathing based on the conversion of gal cypress and the extraction and analytical method of heartbeat signal as described in any one in claim 1-8；And described device at least includes filter unit, it is configured to described spectral centroid change curve be filtered operation, to obtain breath signal and heartbeat signal.

Device the most according to claim 9, it is characterised in that described filter unit includes obtaining the breath signal bandpass filter of breath signal and obtaining the heartbeat signal bandpass filter of heartbeat signal.