A method, apparatus, device and medium for detecting respiration and heartbeat

Abstract

The application discloses a kind of respiratory heartbeat detection method, device, equipment and medium.The method comprises: identifying the phase signal of the object to be detected from the target distance image corresponding to the object to be detected;Respectively from the time domain angle and the frequency domain angle, the phase signal is separated, and the corresponding initial time domain signal component and initial frequency domain signal component are obtained;The initial time domain signal component and the initial frequency domain signal component are reconstructed, and the corresponding target frequency domain signal component is obtained;Actual breathing frequency of the object to be detected is obtained based on the target frequency domain signal component, and the actual heartbeat frequency of the object to be detected is obtained by twice harmonic weighting compensation selection to the target frequency domain signal component.The application selects more suitable heartbeat frequency from the target frequency domain signal component by using twice harmonic weighting compensation selection mode, effectively solves the false detection problem of the object to be detected in the interference state, and effectively improves the detection accuracy of breathing and heartbeat frequency.

Description

Technical Field

[0001] This invention relates to the field of physiological signal processing technology, and in particular to a method, apparatus, device, and medium for detecting respiratory heartbeat. Background Technology

[0002] Chest and heart diseases are receiving increasing attention and are considered life-threatening. Respiratory and heartbeat signals are considered crucial physiological signals, effectively assessing a person's health. Experiments have shown that heartbeat signals can be used to effectively prevent, diagnose, and treat heart diseases. Traditional heart rate measurement methods primarily utilize electrocardiograms (ECGs) and pulse oximeters, requiring direct contact with the patient to detect heart activity. However, these methods can cause discomfort, allergies, and even disease transmission. Therefore, non-contact respiratory and heart rate detection technologies are gaining traction and have significant applications in healthcare, personnel monitoring, and disaster relief.

[0003] Traditional non-contact respiratory and heartbeat detection technologies include LiDAR, cameras, Wi-Fi, and radar. LiDAR is a high-precision measurement device, highly sensitive to minute movements, and is widely used in scene reconstruction, autonomous driving, and other fields. However, LiDAR is expensive, and large-scale application would incur huge costs. Cameras can directly reflect the patient's condition, but are easily affected by lighting conditions, patient obstruction, etc., and have poor anti-interference capabilities. Wi-Fi-based respiratory and heartbeat detection is limited by distance and rapid attenuation, and cannot completely acquire respiratory and heartbeat signals. Radar-based respiratory and heartbeat detection, on the other hand, has high accuracy, is sufficiently sensitive to minute human movements, has strong anti-interference capabilities, is not easily affected by lighting conditions or obstructions, and is also inexpensive. Millimeter-wave radar combines the high accuracy of continuous-wave radar with the detection range of ultra-wideband radar. Therefore, using millimeter-wave radar for human respiratory and heartbeat detection has become a common method.

[0004] The technology for detecting breathing and heartbeat based on millimeter-wave radar includes: detecting the phase of the radar echo and interpolating it to compensate for phase information loss due to insufficient sampling rate, then using the ICEEMDAN method to separate the phase information to obtain the breathing and heartbeat signal; or sampling the obtained phase information at different sampling rates to obtain phase information with different amounts of information, then processing and reconstructing the phase information at different sampling rates using Fourier transform and other methods to obtain the breathing and heartbeat signal. However, the above-mentioned breathing and heartbeat signal detection algorithms based on millimeter-wave radar do not consider detection under scene interference or colored noise interference. Furthermore, these methods are prone to false positives and false negatives, and their application scope is limited. Summary of the Invention

[0005] This invention provides a method, apparatus, device, and medium for detecting respiratory heartbeat, in order to solve the technical problems in the prior art that it is impossible to perform detection under scene interference and colored noise interference, and that it is prone to false detection and missed detection and has a limited scope of application.

[0006] According to one aspect of the present invention, a method for detecting respiratory and heartbeat is provided, comprising:

[0007] Identify the phase signal of the target object from the target range profile corresponding to the target object; wherein the target range profile is generated based on the distance information between the millimeter-wave radar and the target object;

[0008] The phase signal is separated from both the time domain and frequency domain perspectives to obtain the corresponding initial time domain signal component and initial frequency domain signal component;

[0009] The initial time-domain signal component and the initial frequency-domain signal component are reconstructed to obtain the corresponding target frequency-domain signal component;

[0010] The actual respiratory rate of the subject under test is obtained based on the target frequency domain signal components, and the actual heart rate of the subject under test is obtained by performing second harmonic weighted compensation selection on the target frequency domain signal components.

[0011] According to another aspect of the present invention, a respiratory and heart rate detection device is provided, comprising:

[0012] The identification module is used to identify the phase signal of the object to be detected from the target range image corresponding to the object to be detected; wherein the target range image is generated based on the distance information between the millimeter-wave radar and the object to be detected;

[0013] The separation module is used to separate the phase signal from the time domain and the frequency domain respectively to obtain the corresponding initial time domain signal component and initial frequency domain signal component;

[0014] The reconstruction module is used to reconstruct the initial time-domain signal component and the initial frequency-domain signal component to obtain the corresponding target frequency-domain signal component;

[0015] The detection module is used to obtain the actual respiratory rate of the object to be detected based on the target frequency domain signal component, and to perform second harmonic weighted compensation selection on the target frequency domain signal component to obtain the actual heart rate of the object to be detected.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the respiratory and heart rate detection method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the respiratory and heartbeat detection method according to any embodiment of the present invention.

[0021] The technical solution of this invention identifies the phase signal of the target object from the target range image generated based on the distance information between the millimeter-wave radar and the target object, and separates the phase signal from both the time and frequency domains. This ensures decomposition efficiency while improving decomposition speed. At the same time, by using a second harmonic weighted compensation selection method to select a more suitable heartbeat frequency from the target frequency domain signal components, the problem of false detection of the target object under interference is effectively solved, and the detection accuracy of breathing and heartbeat frequencies is effectively improved.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a respiratory and heartbeat detection algorithm provided in an embodiment of the present invention;

[0025] Figure 2 This is a flowchart of a respiratory and heart rate detection method provided in an embodiment of the present invention;

[0026] Figure 3 This is a schematic diagram illustrating the configuration of raw data acquired by a single receiving antenna according to an embodiment of the present invention;

[0027] Figure 4This is a CA_CFAR detection block diagram provided in an embodiment of the present invention;

[0028] Figure 5 This is a flowchart of another respiratory and heart rate detection method provided in an embodiment of the present invention;

[0029] Figure 6 This is a flowchart illustrating the implementation of a symplectic geometric similarity transformation provided in an embodiment of the present invention;

[0030] Figure 7 This is a schematic diagram of the results of an initial distance profile and a target distance profile provided in an embodiment of the present invention;

[0031] Figure 8 This is a schematic diagram of CFAR detection results and phase signals provided in an embodiment of the present invention;

[0032] Figure 9 This is a schematic diagram illustrating the separation of initial time-domain signal components and initial frequency-domain signal components provided in an embodiment of the present invention;

[0033] Figure 10 This is a schematic diagram of the result of a target time-domain signal component and a target frequency-domain signal component provided in an embodiment of the present invention;

[0034] Figure 11 This is a schematic diagram of respiratory rate test results and heart rate test results provided in an embodiment of the present invention;

[0035] Figure 12 This is a schematic diagram of a millimeter-wave radar system and the object to be detected in an actual scene according to an embodiment of the present invention;

[0036] Figure 13 This is a schematic diagram of the structure of a respiratory and heart rate detection device provided in an embodiment of the present invention;

[0037] Figure 14 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0038] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0039] It should be noted that the terms "first," "second," "initial," "target," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0040] This invention separates the phase signal from both the time and frequency domains, improving decomposition speed while maintaining efficiency. It also utilizes a second harmonic weighted compensation method to select a more suitable respiratory and heartbeat signal, effectively solving the problem of false detection under interference. Furthermore, this method can process colored noise interference to obtain respiratory and heartbeat signals with high accuracy.

[0041] Figure 1 This is a flowchart of a respiratory and heartbeat detection algorithm provided in an embodiment of the present invention. Figure 1 As shown, firstly, a one-dimensional FFT is performed on the obtained raw echo data ADC. Then, mean cancellation is used to process the initial range image to suppress static clutter in the scene, obtaining the corresponding target range image. Secondly, the range cell with the highest energy and the range cell with the highest phase variance are statistically identified using one-dimensional CFAR and the phase variance information of each range cell. Based on this, the range cell with the highest variance and energy is selected, and a sampling point is extracted from each frame. Simultaneously, the phase information of the human body is composed using data from multiple frames, which includes the human body's breathing and heartbeat information. For the obtained phase information, the SG-VMD algorithm is used to process the human body phase information and separate the breathing and heartbeat signals. Since it is necessary to select accurate breathing and heartbeat signals from the signal components, the second harmonic weighted compensation selection method is used to select each signal component. This method can select accurate breathing and heartbeat frequencies.

[0042] Specifically, the millimeter-wave radar system transmits a Linear Frequency Modulation Continuous Wave (LFMCW) signal, operating in MIMO mode. The received signals from multiple channels are mixed to obtain intermediate frequency (IF) signals. The acquisition board then acquires these IF signals according to the Nyquist sampling theorem to obtain the raw ADC data. This raw ADC data contains target information, and target-related information can be obtained through signal processing.

[0043] The raw ADC data is subjected to a 1D FFT, followed by clutter suppression using mean cancellation. A suitable range cell is then selected using one-dimensional CFAR and phase variance, and phase information is extracted from this range cell. The phase information is then processed using the SG-VMD method to decompose it into multiple signal components. Finally, a second harmonic weighted compensation selection method is used to select a suitable respiratory and heartbeat signal.

[0044] In one embodiment, Figure 2 This is a flowchart of a breathing and heartbeat detection method provided in an embodiment of the present invention. This embodiment is applicable to situations where breathing and heartbeat are detected in the presence of interference noise. The method can be executed by a breathing and heartbeat detection device, which can be implemented in hardware and / or software and can be configured in an electronic device.

[0045] In this embodiment, before identifying the phase signal of the target object from the target range image, the method further includes: performing FFT processing on the pre-acquired raw ADC data to obtain a corresponding initial range image; wherein the raw ADC data contains distance information based on the millimeter-wave radar and the target object; and filtering the initial range image using mean cancellation to obtain a corresponding target range image.

[0046] The ADC raw data contains distance information between the radar and the stationary human body to be detected. By performing FFT processing on the ADC raw data in the fast time dimension, i.e. the distance dimension, the obtained spectrum is positively correlated with the distance. Based on the relationship between the two, the frequency is converted into the target distance, and the distribution of the target on the range image can be obtained.

[0047] Figure 3 This is a schematic diagram illustrating the configuration of raw data acquired by a single receiving antenna according to an embodiment of the present invention. The matrix S of the raw data received by a single receiving antenna is as follows: Figure 3 As shown, the data matrix is ​​an N s Okay, N c ×N s A matrix of columns.

[0048] Where, N sN represents the number of sampling points in a chirp. c N represents the number of chirs contained in each frame. f This represents the total number of frames acquired. The initial distance profile is obtained by performing a one-dimensional FFT on the raw acquired data. The calculation method is as follows:

[0049]

[0050] In equation (1), TR (m,k) W represents the magnitude of the m-th chirp at position k among all chirps. u S is a preset window function. (n-u,m) This represents the data of the null-th sampling point of the m-th chirp.

[0051] After obtaining the initial range image, mean cancellation needs to be performed on it to obtain the corresponding target range image. In addition to noise interference, the echo signal also contains a large amount of stationary clutter. This clutter is mainly caused by the radar's environment, and indoors it is primarily caused by beds, pillows, and miscellaneous objects (bags, mobile phones), etc. The biggest difference between these clutter signals and human signals lies in the Doppler frequency. Therefore, clutter signals can be suppressed in the Doppler dimension to better detect liveness signals. Clutter suppression is achieved using mean cancellation. The distance from the stationary target to the radar antenna is constant, and the time delay of the stationary target on each received pulse is also constant. Averaging all received pulses (received pulses from the stationary target and the moving target) yields the reference received pulse. Then, subtracting the reference received pulse from each received pulse yields the received pulse for the moving target. The core idea is to calculate the difference between the mean and the reference pulse.

[0052] like Figure 2 As shown, the method includes:

[0053] S110. Identify the phase signal of the object to be detected from the target range image corresponding to the object to be detected.

[0054] The target range profile is generated based on the distance information between the millimeter-wave radar and the target object. In this embodiment, the target object refers to a user whose respiratory rate and heart rate need to be detected. The millimeter-wave radar system can be placed at a suitable location close to the upper body of the target object, and the distance information between the millimeter-wave radar and the target object can be collected based on the millimeter-wave radar system. Then, the distance information is subjected to range-dimensional FFT and mean cancellation to obtain the corresponding target range profile.

[0055] In this embodiment, taking a target range image as a one-dimensional range image as an example, the process of identifying the phase signal is explained against a white Gaussian noise background. Figure 4 This is a CA_CFAR detection block diagram provided by an embodiment of the present invention. The generalized structure of constant false alarm rate (CFAR) detection under a white Gaussian background based on square-law detection is as follows: Figure 4 As shown.

[0056] A certain number of cells around the cell to be detected are selected as reference cells. Under general assumptions, the cell to be detected and the reference cells can be considered to have the same statistical distribution. To prevent information from the cell to be detected from leaking to nearby reference cells and causing erroneous detection of the cell to be detected, some cells even closer to the cell to be detected are usually designated as guard cells. These guard cells do not participate in the corresponding calculations; they are only used to ensure that the information of the cell to be detected is not leaked to the reference cells, thereby enabling effective estimation of clutter power and avoiding the influence of clutter.

[0057] For each distance cell in the obtained one-dimensional signal, the signal is sent for square-law detection and incoherent accumulation to obtain the corresponding statistic X.

[0058] The detection statistic based on the number of reference cells defined by the above method requires clutter power to estimate Z, and then the threshold factor α is calculated. t To form an adaptive detection threshold α t Z. Depending on the estimation strategy, corresponding CFAR techniques can be developed to estimate Z. For example, CA-CFAR. Here, the estimation strategy is characterized in a generalized form, and can be described in the corresponding mathematical form as follows:

[0059] Z = f(X1,X2,…,X) M (2)

[0060] The detection statistic D of the detection unit is compared with the detection threshold α. t Z-comparison, decision output. The detection of one detection unit is now complete.

[0061] Based on the above detection procedure, under condition H1, the detection probability is:

[0062] P d =Pr(D>α) t Z|H1) (3)

[0063] As can be seen from the above formula, the constant false alarm rate (CFAR) detection probability includes two random variables: the detection statistic D of the target cell and the clutter power Z. That is, to determine the detection probability for a target cell, it is necessary to simultaneously determine the detection cell statistic and the power distribution estimation characteristics. To make the derivation easier to understand, this chapter uses Bayesian theory to perform statistical operations on the two variables in two steps. Specifically, we first assume that the clutter power Z is known, and the threshold is fixed at α. tZ, calculate the conditional detection probability P of undulating targets using the statistical characteristics of undulating targets. d (Z); then, based on the statistical characteristics f of the clutter power estimates corresponding to different constant false alarm strategies... Z (z), for the conditional detection probability P d (Z) Perform integration to obtain the detection probability P. d .

[0064] Step 1: Assume that the clutter power estimate Z is known, and the corresponding threshold is a constant α. t Z. Conditional detection probability P of fluctuating targets. d (Z) As in equation (4):

[0065]

[0066] Step 2: Utilize background power PDFf Z (z) conditional detection probability P d (Z) Perform statistical averaging to derive the unconditional detection probability P. d ,Right now:

[0067]

[0068] It is worth noting that the performance of constant false alarm rate (CFAR) detection will depend on the statistical characteristics f of the power estimate. Z (z) (determined by clutter estimation strategy and background characteristics) and statistical characteristics of detection unit detection statistics f Y (y)(determined by the fluctuating target model).

[0069] On the other hand, under condition H0, the false alarm probability is:

[0070]

[0071] Among them, P fa (Z) is the conditional false alarm probability, which is:

[0072]

[0073] This method allows us to obtain range cells that may represent human targets. Then, based on the phase variance of these range cells, we determine whether the current range cell represents a stationary human. Using this method, we can obtain the range cells of the stationary human to be detected. We select chirps at the same location in each frame, and then select the peak value of the range cell of the stationary human to be detected from these chirps as a data point for the discrete phase signal. We statistically analyze the data from all frames using this method and arrange them in chronological order to form the phase signal.

[0074] S120. Separate the phase signal from the time domain and frequency domain perspectives respectively to obtain the corresponding initial time domain signal component and initial frequency domain signal component.

[0075] In the embodiments, the phase signal can be separated by symplectic geometric similarity transformation to obtain the corresponding initial time-domain signal component, and the initial time-domain signal component can be decomposed by VMD decomposition to obtain the corresponding initial frequency-domain signal component.

[0076] S130. Reconstruct the initial time-domain signal component and the initial frequency-domain signal component to obtain the corresponding target frequency-domain signal component.

[0077] In an embodiment, when the maximum amplitude positions of two initial frequency domain signal components are the same or close, the two initial time domain signal components corresponding to these two initial frequency domain signal components can be merged together to obtain a merged time domain signal component. Then, the two initial frequency domain signal components are replaced by the merged time domain signal component, and all time domain signal components are converted into target frequency domain signal components using fast Fourier transform again.

[0078] S140. Obtain the actual respiratory rate of the subject under test based on the target frequency domain signal components, and perform second harmonic weighted compensation selection on the target frequency domain signal components to obtain the actual heart rate of the subject under test.

[0079] In this embodiment, the signal frequency of each target frequency domain signal component is acquired, and the target frequency domain signal components whose signal frequencies fall within a preset respiratory frequency range are assigned to a respiratory frequency candidate set. The frequency of the target frequency domain signal component with the largest amplitude is selected from the respiratory frequency candidate set as the actual respiratory frequency. Simultaneously, the target frequency domain signal components whose signal frequencies fall within a preset heart rate frequency range are assigned to an initial heart rate candidate set. When the target frequency domain signal component whose signal frequency falls within a preset second harmonic range has a 2:1 relationship with one of the target frequency domain signal components in the initial heart rate candidate set, the amplitude of the target frequency domain signal component within the preset second harmonic range is directly multiplied by a preset weighting factor and applied to one of the target frequency domain signal components in the initial heart rate candidate set. If the target frequency domain signal component whose signal frequency falls within the preset second harmonic range does not have a 2:1 relationship with one of the target frequency domain signal components in the initial heart rate candidate set, the amplitude of the target frequency domain signal component within the preset second harmonic range is directly multiplied by a preset weighting factor and applied as the amplitude of the target frequency domain signal component in the initial heart rate candidate set.

[0080] The technical solution of this embodiment identifies the phase signal of the target object from the target range image generated based on the distance information between the millimeter-wave radar and the target object, and separates the phase signal from both the time domain and frequency domain perspectives. This ensures decomposition efficiency while improving decomposition speed. At the same time, by using the second harmonic weighted compensation selection method to select a more suitable heartbeat frequency from the target frequency domain signal components, the false detection problem of the target object under interference is effectively solved, and the detection accuracy of breathing and heartbeat frequencies is effectively improved.

[0081] In one embodiment, Figure 5 This is a flowchart of another respiratory and heart rate detection method provided by an embodiment of the present invention. This embodiment further explains the phase signal separation process, signal component reconstruction process, and respiratory and heart rate detection process based on the above embodiments. Figure 5 As shown, the method includes:

[0082] S210. Identify the phase signal of the object to be detected from the target range image corresponding to the object to be detected.

[0083] The target range profile is generated based on the distance information between the millimeter-wave radar and the object to be detected.

[0084] S220. The phase signal is separated by symplectic geometric similarity transformation to obtain the corresponding initial time-domain signal components.

[0085] In this embodiment, the phase information is separated from the time domain using the symplectic geometric similarity transformation method, which can separate the phase information without destroying the essential characteristics of the signal.

[0086] In one embodiment, S220 includes: determining the trajectory matrix of the phase signal according to the time-series delay equivalence topology; obtaining the corresponding coefficient transition matrix based on the trajectory matrix, the covariance matrix corresponding to the trajectory matrix, the pre-configured symplectic geometric orthogonal matrix and the upper triangular matrix; obtaining the corresponding reconstruction initial component based on the coefficient transition matrix; and transforming the reconstruction initial component using a diagonal averaging method to obtain the corresponding time-domain signal component.

[0087] Figure 6 This is a flowchart illustrating the implementation of a symplectic geometric similarity transformation provided in an embodiment of the present invention. Figure 6 As shown, the implementation of the symplectic geometric similarity transformation method includes the following process:

[0088] Assume the phase signal obtained from step three is x(t) = x1, x2, ..., xt. n , where n is the length of the x-phase signal, and the trajectory matrix X can be obtained by using this signal according to the time series delay equivalent topology method.

[0089]

[0090] Where r is the time delay, q is the embedding dimension, and m = n - (q - 1)r. r is typically set to 1. The embedding dimension is chosen appropriately based on the needs. In this embodiment, the embedding dimension is 20. After obtaining the trajectory matrix using the above method, the covariance matrix Y of the trajectory matrix is ​​obtained.

[0091]

[0092] The Hamiltonian matrix can be obtained using this covariance matrix. Mathematically, this can be expressed as:

[0093]

[0094] However, this Hamiltonian matrix is ​​insufficient for symplectic geometric transformations, therefore it needs to be processed again, i.e., Q = H. 2 This matrix is ​​still a Hamiltonian matrix. Using this Hamiltonian matrix, we can obtain the symplectic geometric matrix:

[0095]

[0096] Here, S is a symplectic geometric orthogonal matrix, and R is an upper triangular matrix. However, S and M are unknown at this point. The Householder matrix P can be used to replace the symplectic geometric orthogonal matrix S. And matrix P can be obtained from the Householder matrix G.

[0097]

[0098] The required matrix G can be obtained by performing QR decomposition on the Q matrix in the above equation. Where G... i For matrix Y 2 The eigenvectors of the system can be used to decompose the signal components without changing the system structure. The coefficient transition matrix can be obtained using matrix G and the trajectory matrix X.

[0099]

[0100] The initial components can then be reconstructed using this coefficient transition matrix:

[0101] A i =G i *W, (14)

[0102] The obtained initial reconstructed component Ai is an m*q dimensional matrix, therefore, it needs to be converted into a one-dimensional component using diagonal averaging. The matrix A can be represented as a... ij(1≤i≤m, 1≤j≤q), and we can assume q * =min(m,q) and m * =max(m, q). When m < q, otherwise We can obtain the following by using diagonal averaging:

[0103]

[0104] The diagonal averaging method described above yields q initial time-domain signal components. However, these initial time-domain signal components may still share the same frequency components, necessitating further decomposition in the frequency domain. This invention utilizes VMD decomposition to further decompose the initial time-domain signal components in the frequency domain.

[0105] S230. The initial time-domain signal components are decomposed using the VMD decomposition method to obtain the corresponding initial frequency-domain signal components.

[0106] In one embodiment, S230 includes: acquiring the number of modes, penalty factor, and first preset value of a pre-configured time-domain signal component; determining corresponding target constraints based on target optimization conditions, alternating direction multiplier algorithm, and augmented Lagrange equation; wherein the target optimization conditions are used to minimize the sum of bandwidths of the initial time-domain signal components; and converting the initial time-domain signal components into corresponding initial frequency-domain signal components based on the number of modes, penalty factor, target constraints, and first preset value, using fast Fourier transform.

[0107] Assume that the initial time-domain signal components consist of multiple bandwidth signals with a center frequency, and that these components exhibit sparsity in the frequency domain. Therefore, the sum of the bandwidths of these multiple signals should be minimized. Thus, by using the minimization of this bandwidth sum as the optimization objective, a corresponding optimization equation is constructed.

[0108]

[0109] Where v k (t) represents one of the signal components in an initial time-domain signal component (which can be understood as dividing an initial time-domain signal component into k signal components), w k Let be the center frequency of the signal component, and s be the initial time-domain signal component input. This constrained problem can be effectively solved using equation (16), the alternating multiplier method (ADMM), and the augmented Lagrange equation. The resulting update formula is:

[0110]

[0111]

[0112]

[0113] By iterating continuously using equation (17), the desired signal components can be separated. Here, α is the penalty factor, and λ is the Lagrange term, which can be randomly initialized. This is the frequency domain representation of the original signal. Using VMD requires manually determining the number of modes and the penalty factor for decomposition. The parameters used in this invention are: 6 modes and a penalty factor of 200. The initial time-domain signal components obtained from the symplectic geometric transformation are fed into VMD for decomposition. The specific decomposition process is as follows: the Lagrange terms are initialized to 0, the center frequency of each component is initialized to 0, the number of modes is 6, and the penalty factor is 200, i.e. w i =0, k=6, α=200, τ=0.0204. Then, the time-domain signal component v is transformed using Fast Fourier Transform. k (t) frequency domain conversion signal Then, using (17) for iteration, when condition (18) is met, the signal decomposition is completed, and the corresponding initial frequency domain signal components are obtained. Typically, the first preset value ξ = 1 × 10⁻⁶ is set. -5 .

[0114]

[0115] S240. Identify and extract two initial frequency domain signal components with the same or similar maximum amplitude positions.

[0116] S250. Combine the two initial time-domain signal components corresponding to the two initial frequency domain signal components to obtain the corresponding combined time-domain signal components.

[0117] S260. Replace two initial frequency domain signal components with the same or similar maximum amplitude positions by merging time domain signal components.

[0118] S270. Reuse the Fast Fourier Transform to convert the combined time-domain signal components into corresponding candidate frequency-domain signal components.

[0119] S280. Combine the unmerged initial frequency domain signal components and candidate frequency domain signal components to form the corresponding target frequency domain signal components.

[0120] In this embodiment, after obtaining multiple initial time-domain signal components and initial frequency-domain signal components, it is necessary to reconstruct the multiple components obtained by decomposing them using the relationship between the frequencies of the signal components. That is, q×k signal components can be obtained through symplectic geometric similarity transformation and VMD decomposition. The initial time-domain signal components are transformed into initial frequency-domain signal components using Fast Fourier Transform (FFT). When the maximum amplitude positions of two initial frequency-domain signal components are the same or close, their initial time-domain signal components are added together to obtain a merged time-domain signal component. This merged time-domain signal component replaces the two initial frequency-domain signal components, and FFT is used again to transform all time-domain signal components into target frequency-domain signal components. This process is repeated for all decomposed signal components until all signal components have been traversed, thus completing the reconstruction of the signal components. Simultaneously, the merged time-domain signal component and other unmerged time-domain signal components are reconstructed into the corresponding target time-domain signal component.

[0121] S290. Assign the target frequency domain signal components whose signal frequency is within the preset heartbeat frequency range to the initial heartbeat frequency candidate set.

[0122] The preset heart rate range can be a pre-configured range that conforms to the heart rate of a normal person. In this embodiment, the signal frequency of each target frequency domain signal component is acquired, and the target frequency domain signal components whose signal frequencies fall within the preset heart rate range are assigned to the initial heart rate candidate set.

[0123] S2100. Based on the relationship between the signal frequencies of the first type of target frequency domain signal components and the signal frequencies of the second type of target frequency domain signal components, the amplitude of the second type of target frequency domain signal components in the initial heartbeat frequency candidate set is updated to obtain the corresponding target heartbeat frequency candidate set.

[0124] Among them, the first type of target frequency domain signal component is the target frequency domain signal component whose signal frequency is within the preset second harmonic range; the second type of target frequency domain signal component is the target frequency domain signal component in the initial heartbeat frequency candidate set.

[0125] In one embodiment, if the signal frequency of one of the first type of target frequency domain signal components is twice that of one of the second type of target frequency domain signal components, then the sum of the product of the amplitude of one of the first type of target frequency domain signal components and the preset weighting factor and the amplitude of one of the second type of target frequency domain signal components is taken as the amplitude of one of the second type of target frequency domain signal components.

[0126] If the signal frequency of one of the first-type target frequency domain signal components does not meet the double relationship with the signal frequency of one of the second-type target frequency domain signal components, then the product of the amplitude of one of the first-type target frequency domain signal components and the preset weighting factor is used as the amplitude of one of the second-type target frequency domain signal components, thus obtaining the corresponding target heartbeat frequency candidate set.

[0127] S2110. Select the frequency of the target frequency domain signal component with the largest amplitude from the candidate set of target heartbeat frequencies as the actual heartbeat frequency of the object to be detected.

[0128] S2120. Assign the target frequency domain signal component whose signal frequency is within the preset respiratory frequency range to the respiratory frequency candidate set.

[0129] S2130. Select the frequency of the target frequency domain signal component with the largest amplitude from the candidate set of respiratory frequencies as the actual respiratory frequency of the object to be detected.

[0130] In this embodiment, target frequency domain signal components whose signal frequencies are within a preset respiratory frequency range are assigned to a respiratory frequency candidate set; the frequency of the target frequency domain signal component with the largest amplitude is selected from the respiratory frequency candidate set as the actual respiratory frequency of the object to be detected.

[0131] To address potential issues such as respiratory harmonics and noise interference, the second harmonic of the heartbeat can be used for assessment. Assume the signal frequency of each target frequency domain signal component is f. i The amplitude of the target frequency domain signal component is A(f i Simultaneously, an initial candidate set of heartbeat frequencies Ω is established. Target frequency domain signal components already within the preset heartbeat frequency range are directly added to the initial candidate set of heartbeat frequencies. Assume the preset second harmonic range of the heartbeat signal is l. second h second Update each possible signal component using the following formula:

[0132]

[0133] When a target frequency domain signal component within a preset second harmonic range has a 2:1 relationship with a target frequency domain signal component in the initial heartbeat frequency candidate set, the amplitude of the target frequency domain signal component is multiplied by a preset weighting factor and applied to the amplitude of the corresponding target frequency domain signal component in the initial heartbeat frequency candidate set.

[0134] Otherwise, update according to the following formula:

[0135]

[0136] That is, the signal component within the preset second harmonic range is multiplied by a preset weighting coefficient and added to the initial heartbeat frequency candidate set. After updating the amplitudes of all target frequency domain signal components, the final target heartbeat frequency candidate set is obtained. Finally, the frequency with the largest amplitude is selected from the target heartbeat frequency candidate set as the actual heartbeat frequency of the object to be detected. The preset weighting coefficients α and β can be the same value or different values; this is not limited and can be adjusted according to the actual situation.

[0137] The following are the results obtained from the analysis of the measured data:

[0138] Figure 7 This is a schematic diagram illustrating the results of an initial distance image and a target distance image provided in an embodiment of the present invention. Figure 7 As shown, performing a one-dimensional distance FFT on the original data yields the following results: Figure 7 The initial range image is shown on the left. Due to interference from stationary objects, mean cancellation is used to process the initial range image, resulting in the image shown below. Figure 7 The target range image shown on the right can effectively eliminate static clutter in the initial range image and highlight the object to be detected in the scene.

[0139] Figure 8 This is a schematic diagram of CFAR detection results and phase signals provided in an embodiment of the present invention. Figure 8 As shown, after mean cancellation on the initial range image, the corresponding target range image is obtained; then, each chirp in the target range image is detected. The CFAR algorithm is used to detect targets in the scene and simultaneously obtain the range cell where the target object is located, such as... Figure 8 As shown on the left; then, human phase information is extracted according to this distance unit, such as... Figure 8 As shown on the right.

[0140] Figure 9 This is a schematic diagram illustrating the separation of initial time-domain signal components and initial frequency-domain signal components according to an embodiment of the present invention. The separation of the phase signal into respiratory and heartbeat signals, obtained using a symplectic geometric transformation, is shown below. Figure 9 As shown on the left;

[0141] The initial time-domain signal components obtained by the above symplectic geometric similarity transformation are then decomposed using the VMD decomposition method to obtain the following... Figure 9 The initial frequency domain signal components are shown on the right.

[0142] Figure 10 This is a schematic diagram illustrating the results of a target time-domain signal component and a target frequency-domain signal component according to an embodiment of the present invention. The initial time-domain signal component and the initial frequency-domain signal component are reconstructed based on the signal frequency relationship to obtain the following... Figure 10 The reconstructed time-domain signal component (i.e., the target time-domain signal component) and the reconstructed frequency-domain signal component (i.e., the target frequency-domain signal component) are shown.

[0143] Then, the second harmonic weighted compensation selection method is used to select the signal separation to obtain the final result.

[0144] Assuming a data length of 60 seconds is selected and the time sliding window length is set to 3 seconds, the above-mentioned SG-VMD method is used to detect human breathing and heart rate. Figure 11 This is a schematic diagram illustrating the respiratory rate test results and heart rate test results provided in an embodiment of the present invention. Figure 11 As shown, by observing the data processing results, the millimeter-wave radar system can accurately detect the respiratory and heartbeat signals of human targets and obtain key information about these signals, such as heart rate, respiratory rate, and heartbeat interval. Experimental results demonstrate that this algorithm can effectively overcome interference scenarios and noise, significantly reducing the occurrence of false alarms.

[0145] Figure 12 This is a schematic diagram of a millimeter-wave radar system and the object to be detected in an actual scenario, as provided in an embodiment of the present invention. Figure 12 As shown, for scenarios involving blankets and noise interference during human breathing and heartbeat detection, the aforementioned breathing and heartbeat detection algorithm effectively overcomes the effects of interference and noise. It achieves 100% accuracy when the test subject is not covered by a blanket. Even under conditions such as blankets and coats, it still achieves 97.015% accuracy. Furthermore, it exhibits significantly greater robustness compared to other algorithms and can be practically applied to various scenarios.

[0146] Therefore, in practical scenarios, using the SG-VMD algorithm based on millimeter-wave radar for human respiratory and heartbeat detection has the advantages of strong robustness, wide application, high accuracy, and low false detection rate. It can effectively detect the human respiratory and heartbeat status and provide reasonable suggestions for diagnosis and treatment.

[0147] In one embodiment, Figure 13 This is a schematic diagram of the structure of a respiratory and heart rate detection device provided in an embodiment of the present invention. Figure 13 As shown, the device includes: an identification module 310, a separation module 320, a reconstruction module 330, and a detection module 340.

[0148] The identification module 310 is used to identify the phase signal of the object to be detected from the target range image corresponding to the object to be detected; wherein, the target range image is generated based on the distance information between the millimeter-wave radar and the object to be detected.

[0149] The separation module 320 is used to separate the phase signal from the time domain and the frequency domain respectively to obtain the corresponding initial time domain signal component and initial frequency domain signal component;

[0150] The reconstruction module 330 is used to reconstruct the initial time-domain signal component and the initial frequency-domain signal component to obtain the corresponding target frequency-domain signal component;

[0151] The detection module 340 is used to obtain the actual respiratory rate of the subject under test based on the target frequency domain signal components, and to perform second harmonic weighted compensation selection on the target frequency domain signal components to obtain the actual heart rate of the subject under test.

[0152] In one embodiment, the separation module 320 includes:

[0153] The separation unit is used to separate the phase signal using a symplectic geometric similarity transformation to obtain the corresponding initial time-domain signal components;

[0154] The decomposition unit is used to decompose the initial time-domain signal components using the VMD decomposition method to obtain the corresponding initial frequency-domain signal components.

[0155] In one embodiment, the separation unit includes:

[0156] The first determining subunit is used to determine the trajectory matrix of the phase signal according to the time-series delay equivalent topology;

[0157] The second determining subunit is used to obtain the corresponding coefficient transition matrix based on the trajectory matrix, the covariance matrix corresponding to the trajectory matrix, the pre-configured symplectic geometric orthogonal matrix and the upper triangular matrix;

[0158] The third determining sub-unit is used to obtain the corresponding reconstructed initial components based on the coefficient transition matrix;

[0159] The first conversion subunit is used to convert the initial reconstructed components using a diagonal averaging method to obtain the corresponding time-domain signal components.

[0160] In one embodiment, the decomposition unit includes:

[0161] The acquisition subunit is used to acquire the pre-configured modal number, penalty factor, and first preset value of the time-domain signal component;

[0162] The third determining subunit is used to determine the corresponding target constraints based on the target optimization conditions, the alternating direction multiplier algorithm, and the augmented Lagrange equation; wherein, the target optimization conditions are used to minimize the sum of the bandwidths of the initial time-domain signal components;

[0163] The second conversion subunit is used to convert the initial time-domain signal components into corresponding initial frequency-domain signal components based on the modal number, penalty factor, target constraint conditions and first preset value, and using fast Fourier transform.

[0164] In one embodiment, the reconstruction module 330 includes:

[0165] The identification and extraction unit is used to identify and extract two initial frequency domain signal components that have the same or similar maximum amplitude positions;

[0166] The merging unit is used to merge the two initial time-domain signal components corresponding to the two initial frequency-domain signal components respectively to obtain the corresponding merged time-domain signal component.

[0167] The replacement unit is used to replace two initial frequency domain signal components with the same or similar maximum amplitude positions by merging time domain signal components;

[0168] The conversion unit is used to reuse the fast Fourier transform to convert the combined time-domain signal components into corresponding candidate frequency-domain signal components;

[0169] The constituent unit is used to combine the unmerged initial frequency domain signal components and candidate frequency domain signal components into the corresponding target frequency domain signal components.

[0170] In one embodiment, second harmonic weighted compensation selection is performed based on the target frequency domain signal components to obtain the actual heartbeat frequency of the object to be detected, specifically used for:

[0171] Assign the target frequency domain signal components whose signal frequency is within the preset heartbeat frequency range to the initial heartbeat frequency candidate set;

[0172] Based on the relationship between the signal frequencies of the first type of target frequency domain signal components and the second type of target frequency domain signal components, the amplitude of the second type of target frequency domain signal components in the initial heartbeat frequency candidate set is updated to obtain the corresponding target heartbeat frequency candidate set; wherein, the first type of target frequency domain signal components are target frequency domain signal components whose signal frequencies are within the preset second harmonic range; the second type of target frequency domain signal components are target frequency domain signal components in the heartbeat frequency candidate set;

[0173] The frequency of the target frequency domain signal component with the largest amplitude is selected from the candidate set of target heartbeat frequencies as the actual heartbeat frequency of the object to be detected.

[0174] In one embodiment, based on the relationship between the signal frequencies of the first type of target frequency domain signal components and the signal frequencies of the second type of target frequency domain signal components, the amplitude of the second type of target frequency domain signal components in the initial heartbeat frequency candidate set is updated to obtain the corresponding target heartbeat frequency candidate set, specifically used for:

[0175] If the signal frequency of one of the first type of target frequency domain signal components is twice that of one of the second type of target frequency domain signal components, then the sum of the product of the amplitude of one of the first type of target frequency domain signal components and the preset weighting factor and the amplitude of one of the second type of target frequency domain signal components is taken as the amplitude of one of the second type of target frequency domain signal components.

[0176] If the signal frequency of one of the first-type target frequency domain signal components does not meet the double relationship with the signal frequency of one of the second-type target frequency domain signal components, then the product of the amplitude of one of the first-type target frequency domain signal components and the preset weighting factor is used as the amplitude of one of the second-type target frequency domain signal components, thus obtaining the corresponding target heartbeat frequency candidate set.

[0177] In one embodiment, the actual respiratory rate of the object to be detected is obtained based on the target frequency domain signal components, specifically used for:

[0178] Assign target frequency domain signal components whose signal frequency is within the preset respiratory frequency range to the respiratory frequency candidate set;

[0179] The frequency of the target frequency domain signal component with the largest amplitude is selected from the candidate set of respiratory frequencies as the actual respiratory frequency of the object to be detected.

[0180] In one embodiment, the respiratory detection device further includes:

[0181] The conversion module is used to perform FFT processing on the pre-acquired raw ADC data to obtain the corresponding initial range profile; wherein, the raw ADC data contains distance information based on the millimeter-wave radar and the object to be detected;

[0182] The filtering unit is used to filter the initial range image using mean cancellation to obtain the corresponding target range image.

[0183] The respiratory detection device provided in the embodiments of the present invention can execute the respiratory detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0184] In one embodiment, Figure 14 This is a structural block diagram of an electronic device provided in an embodiment of the present invention, such as... Figure 14The diagram illustrates a schematic representation of an electronic device 10 that can be used to implement embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0185] like Figure 14 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0186] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0187] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as respiratory and heartbeat detection methods.

[0188] In some embodiments, the respiratory and heart rate detection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the respiratory and heart rate detection method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the respiratory and heart rate detection method by any other suitable means (e.g., by means of firmware).

[0189] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0190] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0191] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0192] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0193] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0194] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0195] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0196] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method of breath heartbeat detection, characterized by, The method comprises: identifying a phase signal of a to-be-detected object from a target range image corresponding to the to-be-detected object, wherein the target range image is generated based on distance information between a millimeter wave radar and the to-be-detected object; separating the phase signal from a time domain and a frequency domain respectively to obtain corresponding initial time domain signal components and initial frequency domain signal components; reconstructing the initial time domain signal components and the initial frequency domain signal components to obtain corresponding target frequency domain signal components; obtaining an actual breathing frequency of the to-be-detected object based on the target frequency domain signal components, and obtaining an actual heartbeat frequency of the to-be-detected object by performing second harmonic weighting compensation selection on the target frequency domain signal components; wherein the reconstructing the initial time domain signal components and the initial frequency domain signal components to obtain the corresponding target frequency domain signal components comprises: identifying and extracting two initial frequency domain signal components with the same or similar maximum amplitude position; merging two initial time domain signal components corresponding to the two initial frequency domain signal components to obtain a corresponding merged time domain signal component; replacing the two initial frequency domain signal components with the same or similar maximum amplitude position with the merged time domain signal component; reusing fast Fourier transform to convert the merged time domain signal component into a corresponding candidate frequency domain signal component; combining the initial frequency domain signal components that have not been merged and the candidate frequency domain signal component to obtain a corresponding target frequency domain signal component.

2. The method of claim 1, wherein, The separating the phase signal from a time domain and a frequency domain respectively to obtain corresponding initial time domain signal components and initial frequency domain signal components comprises: separating the phase signal by using a symplectic geometric similarity transformation method to obtain the corresponding initial time domain signal components; decomposing the initial time domain signal components by using a VMD decomposition method to obtain the corresponding initial frequency domain signal components.

3. The method of claim 2, wherein, The separating the phase signal by using a symplectic geometric similarity transformation method to obtain the corresponding initial time domain signal components comprises: determining a trajectory matrix of the phase signal according to a time sequence delay equivalence topology; obtaining a corresponding coefficient transfer matrix according to the trajectory matrix, a covariance matrix corresponding to the trajectory matrix, a pre-configured symplectic geometric orthogonal matrix, and an upper triangular matrix; obtaining a corresponding reconstructed initial component based on the coefficient transfer matrix; converting the reconstructed initial component by using a diagonal average method to obtain the corresponding initial time domain signal components.

4. The method of claim 2, wherein, The decomposing the initial time domain signal components by using a VMD decomposition method to obtain the corresponding initial frequency domain signal components comprises: obtaining a modal number, a penalty factor, and a first preset value of a pre-configured time domain signal component; determining a corresponding target constraint condition based on a target optimization condition, an alternating direction multiplier algorithm, and an augmented Lagrange equation; wherein the target optimization condition is used to minimize the sum of bandwidths of the initial time domain signal components; based on the modal number, the penalty factor, the target constraint condition, and the first preset value, and by using fast Fourier transform, converting the initial time domain signal components into corresponding initial frequency domain signal components.

5. The method as claimed in claim 1, wherein, The compensating for the second harmonic of the target frequency domain signal component to obtain the actual heartbeat frequency of the to-be-detected object comprises: allocating a target frequency domain signal component with a signal frequency in a preset heartbeat frequency range to an initial heartbeat frequency candidate set; updating the amplitude of a second type target frequency domain signal component in the initial heartbeat frequency candidate set based on the relationship between the signal frequency of the first type target frequency domain signal component and the signal frequency of the second type target frequency domain signal component to obtain a corresponding target heartbeat frequency candidate set; wherein the first type target frequency domain signal component is a target frequency domain signal component with a signal frequency in a preset second harmonic range; the second type target frequency domain signal component is a target frequency domain signal component in the initial heartbeat frequency candidate set; selecting the frequency of the target frequency domain signal component with the largest amplitude from the target heartbeat frequency candidate set as the actual heartbeat frequency of the to-be-detected object.

6. The method of claim 5, wherein, The updating the amplitude of a second type target frequency domain signal component in the initial heartbeat frequency candidate set based on the relationship between the signal frequency of the first type target frequency domain signal component and the signal frequency of the second type target frequency domain signal component to obtain a corresponding target heartbeat frequency candidate set comprises: if the signal frequency of one of the first type target frequency domain signal components and the signal frequency of one of the second type target frequency domain signal components satisfy a double relationship, then taking the product of the amplitude of the one of the first type target frequency domain signal components and a preset weighting factor as the amplitude of the one of the second type target frequency domain signal components; if the signal frequency of one of the first type target frequency domain signal components and the signal frequency of one of the second type target frequency domain signal components do not satisfy a double relationship, then taking the product of the amplitude of the one of the first type target frequency domain signal components and a preset weighting factor as the amplitude of the one of the second type target frequency domain signal components to obtain a corresponding target heartbeat frequency candidate set.

7. The method as claimed in claim 1, wherein, The obtaining the actual breathing frequency of the to-be-detected object based on the target frequency domain signal component comprises: allocating a target frequency domain signal component with a signal frequency in a preset breathing frequency range to a breathing frequency candidate set; selecting the frequency of the target frequency domain signal component with the largest amplitude from the breathing frequency candidate set as the actual breathing frequency of the to-be-detected object.

8. The method according to any one of claims 1 to 7, characterized in that, The method further comprises: performing FFT processing on pre-acquired ADC original data to obtain a corresponding initial range image; wherein the ADC original data contains distance information between the millimeter wave radar and the to-be-detected object; performing filtering on the initial range image using mean cancellation to obtain a corresponding target range image.

9. A respiratory heartbeat detection apparatus, characterized by comprise: an identification module configured to identify a phase signal of the to-be-detected object from a target range image corresponding to the to-be-detected object; wherein the target range image is generated based on distance information between the millimeter wave radar and the to-be-detected object; a separation module configured to separate the phase signal from the time domain and the frequency domain respectively to obtain a corresponding initial time domain signal component and an initial frequency domain signal component; The reconstruction module is configured to reconstruct the initial time domain signal components and the initial frequency domain signal components to obtain corresponding target frequency domain signal components; The detection module is configured to obtain an actual breathing frequency of the to-be-detected object based on the target frequency domain signal components, and to obtain an actual heartbeat frequency of the to-be-detected object by performing second harmonic weighting compensation selection on the target frequency domain signal components. The reconstruction module comprises: An identification and extraction unit configured to identify and extract two initial frequency domain signal components with the same or similar maximum amplitude positions; A merging unit configured to merge two initial time domain signal components corresponding to the two initial frequency domain signal components to obtain corresponding merged time domain signal components; A replacement unit configured to replace the two initial frequency domain signal components with the same or similar maximum amplitude positions with the merged time domain signal components; A conversion unit configured to convert the merged time domain signal components into corresponding candidate frequency domain signal components by using fast Fourier transform; A composition unit configured to compose the initial frequency domain signal components that have not been merged and the candidate frequency domain signal components to obtain corresponding target frequency domain signal components.

10. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the breathing and heartbeat detection method in any one of claims 1-8.

11. A computer readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for enabling the processor to execute the breathing and heartbeat detection method in any one of claims 1-8 when executed.

Images