阅读说明：本技术 一种基于声学环境响应的与位置无关的呼吸监测方法 (Breathing monitoring method irrelevant to position based on acoustic environment response ) 是由 王天本 汪志胜 陈子毅 刘现涛 李张本 胡瑾 于 2020-12-14 设计创作，主要内容包括：本发明提供一种基于声学环境响应的与位置无关呼吸检测方法,包括以下步骤：在室内布设声波发射器和接收器,其中声波发射器循环发射激励信号,该激励信号为极短的宽频信号,由一段加窗的调频信号和0序列拼接而成,接收器实时接收与激励信号等长的信号,不断解算出设定频段内的CFR序列；再则,以此不断填入具有fifo功能的矩阵,当填满后对各频点在时间维度上去趋势并计算自相关,继而提取出自相关最强的几段CFR序列并归一化；最后,对这些序列进行连续波形的合成,以此轮询实现对象静息状态下的呼吸监测。本发明可解决传统基于声波测距和呼吸气流多普勒效应的呼吸检测对对象方位敏感的难题,能够实现与位置无关的呼吸检测。(The invention provides a position-independent respiration detection method based on acoustic environment response, which comprises the following steps: arranging an acoustic transmitter and a receiver indoors, wherein the acoustic transmitter transmits an excitation signal circularly, the excitation signal is an extremely short broadband signal and is formed by splicing a windowed frequency modulation signal and a 0 sequence, and the receiver receives a signal with the same length as the excitation signal in real time and continuously solves a CFR sequence in a set frequency band; then, continuously filling a matrix with fifo function, after filling, trending each frequency point in the time dimension and calculating autocorrelation, and then extracting and normalizing several CFR sequences with strongest autocorrelation; and finally, synthesizing continuous waveforms of the sequences, and polling to realize the respiration monitoring of the subject in a resting state. The invention can solve the problem that the traditional respiratory detection based on the acoustic ranging and the respiratory airflow Doppler effect is sensitive to the position of a target, and can realize respiratory detection independent of the position.)

1. A method for location-independent respiratory monitoring based on acoustic environmental responses, comprising the steps of:

step 1, arranging an acoustic wave transmitter and an acoustic wave receiver indoors, wherein the acoustic wave transmitter transmits an excitation signal x circularly_t(n) the sonic receiver receives the excitation signal x in real time without blocking_t(n) equal-length signal data x_r(n)；

Step 2, adopting fast Fourier transform algorithm to carry out processing on the current received signal x_r(n) with a known transmitted excitation signal x_t(n) solving the frequency spectrum, solving the channel frequency response sequence H (k) of the current time in the indoor environment,n is the power of 2 closest to the size of N, according to the excitation signal x_t(n) frequency range extracting H from H (k)_r(k) Representing a useful sequence of channel frequency responses;

step 3, setting a matrix Buffer with fifo function, and comparing the current time H_r(k) Writing from the end of Buffer, wherein the column direction of the obtained matrix represents the time dimension, and the row direction represents [ f [ ]_c,f_c+B]Frequency in the range, f_cThe initial frequency of frequency modulation is B, and the frequency modulation bandwidth is B;

step 4, when the Buffer is in a full state, entering step 5, otherwise, resuming the step 2 and the step 3;

step 5, performing detrending on each line of data in the Buffer, namely a channel frequency response sequence;

step 6, quickly calculating autocorrelation of each row of sequences in the Buffer by using a time domain convolution theorem, and acquiring a maximum value R (k);

step 7, according to the set autocorrelation threshold value R and the parameter j, selecting the front j-column channel frequency response sequence with the strongest autocorrelation from the Buffer, writing the front j-column channel frequency response sequence into the set Final FRs matrix cache, and normalizing the front j-column channel frequency response sequence;

step 8, synthesizing continuous waveforms of data in FinalFRs to obtain a respiratory wave sequence currBreathwave;

and 9, smoothing the synthesized respiratory wave sequence currBreathwave at the current moment, namely the respiratory wave monitored in real time, realizing the visualization of the respiratory wave, continuously skipping to the step 2 for circulation, and realizing the respiratory monitoring of the indoor object in a resting state.

2. The method of claim 1, wherein the excitation signal x is based on an acoustic environment response_t(n) is formed by splicing a section of windowed linear or sinusoidal frequency modulation signal and a full 0 sequence, wherein the frequency modulation signal period is T, the full 0 sequence signal duration is T ', the total signal duration is T ═ T + T', namely, the duty ratio of the signal isWherein the frequency of the chirp signal varies with time f₁(n) and phase variation with time u₁(n) are as follows:

frequency variation over time f of sinusoidal FM signals₂(n) and phase variation with time u₂(n) are as follows:

the expression for the frequency modulated signal is as follows:

x_t1(n)＝cos(u(n))

u (n) is u₁(n) or u₂(n)；

The expression for the all 0 sequence signal is as follows:

x_t2(n)＝[0,2*0,…n*0]

the transmitted excitation signal is expressed as follows:

x_t(n)＝x_t1(Tf_s)+x_t2(T′f_s)

where n is the number of sampling points, i.e. n is 1,2, …, tf_s，f_sFor the system sampling frequency, T_sIn order to be the sampling period of the system,

3. the method of claim 1, wherein the current time channel frequency response sequence h (k) is calculated as follows:

wherein H (k) represents the current time and the frequency isThe sequence of channel frequency responses of the first and second channels,n is the power of 2 closest to the size of N, X_t(k) Representing the frequency spectrum of the excitation signal, X_r(k) Representing the frequency spectrum of the received signal, due to the frequency range of the excitation signal being f_c,f_c+B]So f_c<f<f_c+ B, the final value range of k is as follows:

thereby extracting H_r(k)。

4. The method of claim 1, wherein the matrix Buffer size is a x b and is expressed as follows:

wherein the content of the first and second substances,H_{r_a}(k) h indicating the moment of the stored row a_r(k) And (4) sequencing.

5. The method of claim 4, wherein the Buffer is a full state indicator matrix Buffer full H_r(k) The number of rows of the sequence is equal to a; the detrending of each line of data in the Buffer is to subtract an optimal fitting curve from each line of data.

6. The method for location-independent respiratory monitoring based on acoustic environment response of claim 1, wherein the maximum value r (k) is calculated as follows:

R(k)＝max(IFFT(X(w)X^*(w)))

wherein X (w) represents the frequency spectrum of each column in Buffer, X^*(w) represents a conjugate spectrum.

7. The method of claim 1, wherein the written FinalFRs matrix buffer is initialized to 0 matrix with a size of a x j, and directly assigned, and the normalization is a mode of maximum normalization.

8. The method for location-independent respiratory monitoring based on acoustic environment response of claim 1, wherein in step 8, the following traversal operation is performed on j columns of data in FinalFRs:

adding corresponding elements of the sequence, and recording as Wave;

step two, after the absolute value of the Wave sequence is taken, all elements are added and summed to obtain a value which is recorded as Sum;

step three, if Sum is less than LastSum, subtracting the current l-th row sequence amplified by 2 times from Wave, and executing step two again to avoid the waveform with the phase difference being nearly a half cycle, wherein LastSum is the value obtained by the last traversal, and is initially 0, and l is 1,2, … and j;

fourthly, assigning the current Sum to LastSum;

and after traversing, averaging the values of the Wave sequence except j to obtain the current synthesized respiratory Wave sequence which is marked as currBreath Wave.

9. The method for monitoring the respiration based on the acoustic environment response and being independent of the position according to the claim 8, is characterized in that in order to enable the continuous refreshing respiration waveform to be continuous, the waveforms which are at the previous moment and have the phase difference of nearly half cycle are avoided, the following steps are adopted:

adding two sequences of currBreath wave and LastBreath wave separately and corresponding elements, operating according to the step II, and respectively recording the obtained values as Sum2, Sum1 and tempSum, wherein LastBreath wave is a respiratory wave sequence synthesized at the last moment;

and (2) if the tempSum is less than or equal to Sum2 or the tempSum is less than or equal to Sum1, negating each numerical value of the currBreath wave sequence.

10. The method of claim 1, wherein the acoustic transmitter and the acoustic receiver are disposed in a room, the room being relatively enclosed, in a single-shot single-receiver or multiple-shot single-receiver format.

Technical Field

The invention belongs to the technical field of respiration monitoring, and particularly relates to a sleep respiration monitoring method which utilizes sound wave signals in a non-contact mode indoors and does not need a transceiver to face an object.

Background

Respiration is one of the most basic and important physiological sign information of animals, and human bodies or cultured livestock are easy to suffer from respiratory diseases, which can cause respiratory processes with different degrees of morbidity, such as: some diseases result in breathing exhibiting different symptoms of shortness of breath, disordered breathing, dyspnea, etc., compared to normal breathing. Therefore, continuous respiratory monitoring is an important means for determining the health status and further preventing and controlling diseases. With the rapid development of sensing technology and information technology, the non-destructive and automatic respiration monitoring technology gradually becomes a development trend and a research hotspot. The current automatic respiration monitoring technology can be roughly divided into two categories, namely contact type and non-contact type. The contact method utilizes specific sensors in wearable equipment, such as pressure-sensitive, gas-sensitive or heat-sensitive sensors, and realizes respiratory monitoring by measuring thoracic and abdominal movements, respiratory sounds, respiratory airflows and the like, but has the difficulties of high price, intrusiveness, close-fitting carrying at any time and the like.

At present, a non-contact respiration monitoring technology, particularly a respiration monitoring technology based on sound waves, is gradually paid attention to by people, compared with contact monitoring, the non-contact respiration monitoring device does not need to be carried by a human body or attached to any equipment, sound wave signals are widely available, and the non-contact respiration monitoring device has the advantages of non-invasive property, convenience, low cost and the like. However, the current research is Based primarily on direct thoracic and abdominal distance measurement (Nandakuar R, Golakota S, Watson N.Contactless Sleep Apnea Detection on smartphone [ C ]// International Conference on Mobile Systems, Applications, and services ACM,2015: 45-57; Wang, T.T., D.Zhang, Y.Zheng, T.Gu, X.Zhou and B.Dorizzi.2018.C-FMCW Based Detection Using Signal. proceedings of the ACM on Interactive, Mobile, Werable and UbitoTechnics 1(4):1-20.) and respiratory airflow (Doppler emission P, depth P.Q.S. J. Pat. No. 5. J.F.: sample No. 5. J.S.: 1, 4) and Doppler shift frequency of respiratory flow of gases, sample P.S. 4. Q.S. 4. and Q.S. 4. and Q.S. 4. sample D.S. 4. sample D. 4. Q.S. 4. sample D. 4. and Q.S. 3. noise No. 4. Q.S. 3. Q.S. 4. sample D. 3. Q.S. 4. A. and D. 3. A. These studies are extremely sensitive to the location of the monitored object, and require that the transceiver and the monitored object be in a fixed area and be directed towards the object.

Disclosure of Invention

In order to overcome the disadvantages of the prior art, the present invention provides a method for monitoring respiration independent of position based on acoustic environment response, which mainly utilizes the phenomenon that fluctuation of the thorax and abdomen of a person or a livestock during respiration causes indoor acoustic channel change, i.e. the principle that fluctuation of the thorax and abdomen can cause Channel Frequency Response (CFR) change in an indoor environment, and then realizes respiration monitoring independent of the position of a monitored object by a design method.

In order to achieve the purpose, the invention adopts the technical scheme that:

a method of location-independent respiratory monitoring based on acoustic environmental responses, comprising the steps of:

step 1, arranging an acoustic wave transmitter and an acoustic wave receiver indoors, wherein the acoustic wave transmitter transmits an excitation signal x circularly_t(n) the sonic receiver receives the excitation signal x in real time without blocking_t(n) equal-length signal data x_r(n)；

Step 2, adopting fast Fourier transform algorithm to carry out processing on the current received signal x_r(n) with a known transmitted excitation signal x_t(n) solving the frequency spectrum, solving the channel frequency response sequence H (k) of the current time in the indoor environment,n is the power of 2 closest to the size of N, according to the excitation signal x_t(n) ofExtracting H from H (k) in the frequency range_r(k) Representing a useful sequence of channel frequency responses;

step 3, setting a matrix Buffer with fifo function, and comparing the current time H_r(k) Writing from the end of Buffer, wherein the column direction of the obtained matrix represents the time dimension, and the row direction represents [ f [ ]_c,f_c+B]Frequency in the range, f_cThe initial frequency of frequency modulation is B, and the frequency modulation bandwidth is B;

step 4, when the Buffer is in a full state, entering step 5, otherwise, resuming the step 2 and the step 3;

step 5, performing detrending on each line of data in the Buffer, namely a channel frequency response sequence;

step 6, quickly calculating autocorrelation of each row of sequences in the Buffer by using a time domain convolution theorem, and acquiring a maximum value R (k);

step 7, according to the set autocorrelation threshold value R and the parameter j, selecting the front j-column channel frequency response sequence with the strongest autocorrelation from the Buffer, writing the front j-column channel frequency response sequence into the set Final FRs matrix cache, and normalizing the front j-column channel frequency response sequence;

step 8, synthesizing continuous waveforms of data in FinalFRs to obtain a respiratory wave sequence currBreathwave;

and 9, smoothing the synthesized respiratory wave sequence currBreathwave at the current moment, namely the respiratory wave monitored in real time, realizing the visualization of the respiratory wave, continuously skipping to the step 2 for circulation, and realizing the respiratory monitoring of the indoor object in a resting state.

Preferably, the excitation signal x_t(n) is formed by splicing a section of windowed linear or sinusoidal frequency modulation signal and a full 0 sequence, wherein the frequency modulation signal period is T, the full 0 sequence signal duration is T ', the total signal duration is T ═ T + T', namely, the duty ratio of the signal isWherein the frequency of the chirp signal varies with time f₁(n) and phase variation with time u₁(n) are as follows:

frequency variation over time f of sinusoidal FM signals₂(n) and phase variation with time u₂(n) are as follows:

the expression for the frequency modulated signal is as follows:

x_t1(n)＝cos(u(n))

u (n) is u₁(n) or u₂(n)；

The expression for the all 0 sequence signal is as follows:

x_t2(n)＝[0,2*0,…n*0]

the transmitted excitation signal is expressed as follows:

x_t(n)＝x_t1(Tf_s)+x_t2(T′f_s)

where n is the number of sampling points, i.e. n is 1,2, …, tf_s，f_sFor the system sampling frequency, T_sIn order to be the sampling period of the system,

preferably, the channel frequency response sequence h (k) at the current time is calculated as follows:

wherein H (k) represents the current timeAt a frequency ofThe sequence of channel frequency responses of the first and second channels,n is the power of 2 closest to the size of N, X_t(k) Representing the frequency spectrum of the excitation signal, X_r(k) Representing the frequency spectrum of the received signal, due to the frequency range of the excitation signal being f_c,f_c+B]So f_c<f<f_c+ B, the final value range of k is as follows:

thereby extracting H_r(k)。

Preferably, the matrix Buffer size is a × b, which is expressed as follows:

wherein the content of the first and second substances,H_{r_a}(k) h indicating the moment of the stored row a_r(k) And (4) sequencing. a is generally equal to or more than 150.

Preferably, the Buffer is a full state indicator matrix Buffer full H_r(k) The number of rows of the sequence is equal to a; the de-trending of each column of data in Buffer is the subtraction of an optimal (least squares or polynomial) fit curve from each column of data.

Preferably, the calculation formula of the maximum value r (k) is as follows:

R(k)＝max(IFFT(X(w)X^*(w)))

wherein X (w) represents the frequency spectrum of each column in Buffer, X^*(w) represents a conjugate spectrum.

Preferably, the written-in configured FinalFRs matrix cache is obtained by initializing the FinalFRs matrix to a 0 matrix with the size of a × j, directly assigning values, and normalizing by using the normalization value.

Preferably, in step 8, the following traversal operation is performed on j columns of data in FinalFRs:

adding corresponding elements of the sequence, and recording as Wave;

step two, after the absolute value of the Wave sequence is taken, all elements are added and summed to obtain a value which is recorded as Sum;

step three, if Sum is less than LastSum, subtracting the current l-th row sequence amplified by 2 times from Wave, and executing step two again to avoid the waveform with the phase difference being nearly a half cycle, wherein LastSum is the value obtained by the last traversal, and is initially 0, and l is 1,2, … and j;

fourthly, assigning the current Sum to LastSum;

and after traversing, averaging the values of the Wave sequence except j to obtain the current synthesized respiratory Wave sequence which is marked as currBreath Wave.

In order to enable the continuous refreshing respiratory waveform to be continuous, waveforms which are at the previous moment and have a phase difference of nearly half a cycle need to be avoided, the following steps are adopted:

adding two sequences of currBreath wave and LastBreath wave separately and corresponding elements, operating according to the step II, and respectively recording the obtained values as Sum2, Sum1 and tempSum, wherein LastBreath wave is a respiratory wave sequence synthesized at the last moment;

and (2) if the tempSum is less than or equal to Sum2 or the tempSum is less than or equal to Sum1, negating each numerical value of the currBreath wave sequence.

Preferably, the sound wave transmitter and the sound wave receiver are arranged at the same place or different places indoors, the indoor environment is relatively closed, and the single-transmitting single-receiving mode or the multiple-transmitting single-receiving mode is adopted.

Compared with the prior art, the method can solve the problem that the traditional respiratory monitoring based on the acoustic ranging and the respiratory airflow Doppler effect is sensitive to the position of a target, and can realize respiratory monitoring independent of the position.

Drawings

FIG. 1 is a flow chart of the method of the present invention.

FIG. 2 is a model of the acoustic channel transmission for monitoring respiration according to the method of the present invention.

Fig. 3 is a time domain diagram of the transmitted excitation signal.

Fig. 4 is a plot of Channel Frequency Response (CFR).

Fig. 5 is a graph of the CFR sequences in the first few columns with the strongest autocorrelation.

FIG. 6 is a graph of a respiratory waveform monitored by the method of the present invention.

Detailed Description

The embodiments of the present invention will be described in detail below with reference to the drawings and examples.

As shown in fig. 1, the present invention is a method for location-independent respiration monitoring based on acoustic environment response, comprising the following steps:

firstly, an acoustic transmitter Tx and an acoustic receiver Rx are arranged at the same place or different places in an indoor environment, wherein the acoustic transmitter Tx circularly transmits a set excitation signal x_t(n) the sonic receiver Rx receives in real time without blocking the signal data x of equal length to said excitation signal_r(n) and further, the CFR sequence is parsed in conjunction with the transmitted and received signal data.

Writing the CFR sequence into a matrix with a fifo function, calculating autocorrelation values of each section of data (each column of data) of each frequency point in the time dimension in the matrix, further selecting the first sections of sequences with the strongest autocorrelation for continuous waveform synthesis, and finally smoothing the synthesized waveform sequence to obtain the respiratory wave monitored in real time.

The acoustic transmitter Tx of the present invention is an electronic device capable of transmitting ultrasonic signals in air, and has a certain available bandwidth and a large transmission power, for example: a sound box with ultrasonic frequency band transmitting capability, a reverse piezoelectric transducer and the like; the sound wave receiver Rx is a microphone or a piezoelectric transducer with an ultrasonic receiving frequency band, and has better performance effects such as non-directivity, high sensitivity and the like. In this embodiment, as a core of signal processing, the operation unit may directly adopt a PC, and then connect with a speaker and a microphone to form an audio transceiving system. If the independent product is to be realized, a system-level chip or a DSP chip can be adopted and combined with a receiving and sending unit for design. The arrangement of the transceiving units can adopt the form of transceiving integration or transceiving variant and the like, in addition, the indoor environment is relatively closed, and the number of the transceiving units can select the form of single-transmitting single-receiving, multi-directional transmitting single-receiving, multi-transmitting multi-receiving and the like according to the size of the specific indoor environment.

In fact, as shown in the indoor acoustic channel transmission model of fig. 2, the enclosed indoor environment has rich multi-path reflection, and the received signal is divided into two parts, one part is the signal received by the microphone directly or indirectly via the chest and abdomen reflection of the measured person, and the other part is the signal received by the acoustic receiver Rx completely via the static environment reflection, so the echo signal can be modeled as follows:

wherein the content of the first and second substances,the attenuation coefficient of the ith reflected signal which is directly or indirectly reflected from the measured person and received by the sound wave receiver is obtained;the attenuation coefficient of the jth reflected signal reflected from the static environment and received by the acoustic receiver; Δ nⁱAnd Δ n^jIs the corresponding time delay in terms of sample points. In practice, only a portion of the multipath signal may be received by the acoustic receiver Rx, which is referred to as a valid multipath reflected signal, as shown by the solid line in fig. 2, and a multipath reflected signal that is not received by the acoustic receiver is referred to as an invalid multipath reflected signal, as shown by the dashed line in fig. 2.

If the transceiver, the measured object and the static environment are regarded as an integral system, the breathing process accompanied by the fluctuation of the breast can cause the physical channel of sound wave transmission to change, and mainly comes from the following aspects:

1. the fluctuation of the chest leads to the dynamic change of the number of effective reflected signals, namely N in the formula is changed;

2. the fluctuation of the breast causes a dynamic change in the attenuation coefficient of the effective reflected signal, i.e. in the above formulaA dynamic change occurs.

In summary, the breathing process can change the indoor sound wave transmission channel, and in turn, the real-time change of the indoor sound wave transmission channel parameter can reflect breathing. Therefore, the breathing monitoring independent of the position can be realized by monitoring the change of the indoor sound wave propagation channel parameters in real time.

Referring to fig. 1, the method comprises the following steps:

a. referring to fig. 3, the transmitted excitation signal x_t(n) designed as a very short wideband signal consisting of a windowed linear or sinusoidal FM signal x_t1(n) and all 0 sequences x_t2(n) splicing, wherein the period of the frequency modulation signal is T, the time length of the full 0 sequence signal is T ', the total time length of the signal is T ═ T + T', namely, the duty ratio of the signal isThe frequency and the phase of the linear frequency modulation signal are respectively shown as the following formulas 1 and 3 along with time change, the sinusoidal frequency modulation is shown as the following formulas 2 and 4, the expressions of the frequency modulation signal are respectively shown as the following formula 5, the expression of the full 0 sequence signal is shown as the formula 6, and finally, the expression of the transmitted excitation signal is shown as the formula 7:

x_t1(n) ═ cos (u (n)) -formula 5

x_t2(n)＝[0,2*0,…n*0]Formula 6

x_t(tf_s)＝x_t1(Tf_s)+x_t2(T′f_s) Formula 7

In the above equation, the system sampling rate is f_s，n is the number of sampling points, i.e. n is 1,2, …, tf_s，f_cIndicating the starting frequency of the frequency modulation, in order to avoid the influence of ambient noise and to disturb the perceived object, f_cGenerally, the specific value is more than or equal to 18KHz, and is determined according to the available bandwidth of the transceiver and the audible sound wave frequency band of a perception object, wherein B represents the frequency modulation bandwidth;

b. using fast Fourier transform algorithm to receive signal x_r(n) with a known transmitted excitation signal x_t(n) solving the frequency spectrum to calculate a Channel Frequency Response (CFR) sequence at the current time in the indoor environment, represented by h (k), by the following equation:

in the above formula 8, H (k) represents that the system is at the current time and the frequency isThe sequence of CFR of (a),n is the power of 2 closest to the size of N, X_t(k) Representing the frequency spectrum of the excitation signal, X_r(k) Representing the frequency spectrum of the received signal, due to the frequency range of the excitation signal being f_c,f_c+B]So f_c<f<f_c+ B, the value range of k is as follows, thus extracting H_r(k) The sequence is plotted as4 is shown in the specification;

c. setting an axb with fifo functionThe matrix Buffer of the size is used for converting the H of the current moment_r(k) Written from the end of the Buffer, the column direction of the matrix represents the time dimension, and the row direction represents [ f [ ]_c,f_c+B]A frequency within the range;

d. when the Buffer is in a full state, performing subsequent operation, otherwise, resuming the operations of b, c and d;

e. performing detrending on each column of data in the Buffer, namely a CFR sequence in a time dimension;

f. and (3) quickly calculating autocorrelation of CFR sequences of each column of the Buffer by using a time domain convolution theorem, and acquiring a maximum value R (k) as follows:

R(k)＝max(IFFT(X(w)X^*(w))) formula 11

In the above formula, X (w) represents the frequency spectrum of each column in Buffer, X^*(w) represents a conjugate spectrum;

g. according to a set autocorrelation threshold value R and a parameter j, selecting a front j-column CFR sequence with the strongest autocorrelation from the Buffer, writing the CFR sequence into a set FinalFRs matrix cache, and normalizing the CFR sequence, wherein the sequence is plotted as shown in FIG. 5;

h. and (3) synthesizing continuous waveforms of data in FinalFRs, and traversing j columns of data as follows:

adding corresponding elements of the sequence, and recording as Wave;

secondly, after the absolute value of the Wave sequence is obtained, all elements are added and summed, and finally the obtained value is recorded as Sum;

if Sum < LastSum (the last traversal obtained value, initially 0), Wave subtracts the current l-th column sequence (l is 1,2 … j) amplified by 2 times, and step (ii) is executed again to avoid the waveform with the phase difference close to half a cycle, see a reverse waveform shown in fig. 5;

and fourthly, assigning the current Sum to LastSum.

After traversing, averaging the values of the Wave sequence except j to obtain a current synthesized respiratory Wave sequence which is marked as currBreath Wave; finally, in order to make the continuously refreshed respiration waveform continuous, it is necessary to avoid the waveform with the phase difference close to a half cycle at the previous moment, and the following steps are adopted:

adding two sequences of currBreath wave and LastBreath wave (respiratory wave sequence synthesized at last moment) separately and corresponding elements, operating according to the similar step II, and finally obtaining values respectively marked as Sum2, Sum1 and tempSum;

sixthly, if the tempSum is less than or equal to Sum2 or the tempSum is less than or equal to Sum1, negating each numerical value of the currBreathwave sequence.

i. Smoothing the synthesized respiratory wave sequence currBreathwave at the current moment and realizing the visualization of the respiratory wave, as shown in FIG. 6. And continuing jumping to the operation b for circulation, so that the breathing monitoring of the object in the resting state is realized by polling, and the breathing of the indoor object in the resting state is monitored.