A method and system for detecting stationary human targets based on FMCW through-wall radar

Abstract

This invention discloses a method and system for detecting stationary human targets based on FMCW through-wall radar, belonging to the field of human target detection technology. The method includes: acquiring the echo signal of FMCW radar and mixing it to obtain an intermediate frequency (IF) signal; sampling the IF signal to obtain a two-dimensional signal matrix; performing a range-dimensional FFT on the two-dimensional signal matrix row by row to obtain a range-slow-time profile; extracting the phase information of all stationary targets based on the range-slow-time profile and calculating the peak interval of the autocorrelation function of the phase information for each target; estimating the signal frequency of each stationary target based on the peak interval of the autocorrelation function; and removing stationary targets whose signal frequencies are not within the range of human breathing frequencies to obtain the stationary human targets. This invention distinguishes stationary human targets from other interference signals by determining whether the signal frequency of the stationary target is within the frequency range of human breathing signals, which can effectively reduce the false alarm probability.

Description

Technical Field

[0001] This invention belongs to the field of human target detection technology, specifically relating to a method and system for detecting stationary human targets based on FMCW through-wall radar. Background Technology

[0002] Through-wall radar transmits electromagnetic waves in a specific frequency band, penetrating non-metallic obstacles such as walls and ruins to receive the echoes from human targets. This allows for the detection and location of people, making it a crucial detection and sensing tool in disaster relief and other applications. A stationary human target refers to a micro-moving target that only exhibits breathing, heartbeat, and slight body movements, without changing its position. A stationary target refers to inanimate objects such as walls and furniture in a scene. Electromagnetic waves with frequencies higher than 3GHz have difficulty penetrating, or cannot penetrate, walls made of materials such as concrete and clay. Therefore, selecting an appropriate frequency band is essential to penetrate these obstacles with minimal attenuation, enabling the detection of concealed human targets indoors.

[0003] Under normal conditions, the human body's heart rate and respiratory rate remain within a stable range. The ranges for human respiratory and heart rates are 0.2–0.5 Hz and 0.8–2.5 Hz, respectively, corresponding to chest cavity vibration amplitudes of 5–15 mm and 2–3 mm. Vital signs detection typically employs millimeter-wave FMCW (Frequency Modulated Continuous Wave) radar, which has a relatively short wavelength. Even minute displacements of the chest cavity during respiration can cause significant phase changes. Therefore, analyzing the phase of the radar's intermediate frequency signal can yield human heart rate and respiratory information, enabling human target detection. However, human heart rate and respiratory signals are easily affected by environmental clutter and noise, leading to a higher false alarm probability in human target detection.

[0004] Therefore, a new method for detecting stationary human targets is needed that can effectively distinguish stationary human targets from other interference signals and reduce the probability of false alarms. Summary of the Invention

[0005] In view of this, the present invention proposes a method and system for detecting stationary human targets based on FMCW through-wall radar, which is used to solve the problem of high false alarm probability in existing stationary human target detection.

[0006] In a first aspect, this invention discloses a method for detecting stationary human targets based on FMCW through-wall radar, the method comprising:

[0007] The echo signal from the FMCW radar is acquired and mixed to obtain an intermediate frequency signal. The intermediate frequency signal is then sampled to obtain a two-dimensional signal matrix.

[0008] Perform a distance-dimensional FFT on the two-dimensional signal matrix row by row to obtain a distance-slow time image;

[0009] Phase information of all stationary targets is extracted from the range-slow time image, and the peak interval of the autocorrelation function of the phase information is calculated for each target.

[0010] The signal frequency of each stationary target is estimated based on the peak interval of the autocorrelation function, and stationary targets whose signal frequencies are not within the range of human breathing frequency are eliminated to obtain stationary human targets.

[0011] Based on the above technical solutions, preferably, the step of sampling the intermediate frequency signal to obtain a two-dimensional signal matrix specifically includes:

[0012] The intermediate frequency signal is sampled by an analog-to-digital converter (ADC) to obtain intermediate frequency sampling data;

[0013] For each P-frame of the intermediate frequency sampling data, one frame is extracted and processed through M×P×T. c After a time period, a two-dimensional signal matrix of radar echoes is obtained; where M represents the total number of frames, T c The scan time length of one chirp.

[0014] Based on the above technical solutions, preferably, the step of extracting the phase information of all stationary targets from the distance-slow time image specifically includes:

[0015] Static clutter suppression processing is applied to each range cell of the range-slow time image to obtain the filtered range-slow time image;

[0016] Energy accumulation in the range dimension is performed on the filtered range-slow time image to obtain the energy accumulation result;

[0017] One-dimensional CFAR detection is performed on the energy accumulation results to obtain the range gate numbers of all stationary targets;

[0018] Based on the index of the range gate where all stationary targets are located, the phase information of the range gate where all stationary targets are located is extracted along the time dimension.

[0019] Based on the above technical solutions, preferably, the formula for static clutter suppression processing of each range cell in the range-slow time image is as follows:

[0020]

[0021] Where m = 0, 1, ..., M-1, m is the frame number in the slow time dimension, and M is the total number of frames; k = 0, 1, ..., K-1, k is the distance cell number in the distance dimension, and K is the total number of distance cells; S(m,k), The m-th frame distance-slow time image before and after static clutter suppression processing;

[0022] The formula for energy accumulation in the range dimension of the filtered range-slow time image is as follows:

[0023]

[0024] Where X(k) is the energy accumulation result.

[0025] Based on the above technical solutions, preferably, the step of calculating the peak interval of the autocorrelation function of the phase information specifically includes:

[0026] Calculate the autocorrelation function of the phase information of each stationary target's range gate:

[0027] The autocorrelation functions are normalized separately to obtain normalized autocorrelation functions;

[0028] Calculate the peak value of the normalized autocorrelation function separately, and calculate the time interval between the peak values ​​of each autocorrelation function as the peak interval of the autocorrelation function.

[0029] Based on the above technical solution, preferably, the formula for calculating the autocorrelation function of the phase information of each stationary target's distance gate is as follows:

[0030]

[0031] Among them, R p (i) represents the autocorrelation function value of the phase information of the range gate of the p-th stationary target, where m and i are both frame numbers. These are the phase information of the m-th and (m+i)-th frames of the distance gate for the p-th stationary target, respectively, where p = 1, 2, ..., P, and P is the total number of detected stationary targets.

[0032] Based on the above technical solutions, preferably, the method further includes:

[0033] The stationary human target is detected multiple times. If the number of times the same distance gate is detected in multiple consecutive detection results is greater than a preset threshold, the stationary human target is determined to be a valid stationary human target, and the valid stationary human target is taken as the final detection result.

[0034] Let the corresponding distance gate index be denoted as a. j , through a j Value calculation of the effective stationary human target distance:

[0035]

[0036] Where D represents the effective distance to a stationary human target, C represents the speed of light, μ is the frequency modulation slope, and T... c The scan time length of one chirp.

[0037] A second aspect of the present invention discloses a system for detecting stationary human targets based on FMCW through-wall radar, the system comprising:

[0038] Preprocessing module: used to acquire the echo signal of FMCW radar, mix it to obtain intermediate frequency signal, sample the intermediate frequency signal to obtain a two-dimensional signal matrix;

[0039] Signal transformation module: used to perform a distance-dimensional FFT on the two-dimensional signal matrix row by row to obtain a distance-slow time image;

[0040] Phase analysis module: used to extract phase information of all stationary targets based on the range-slow time image, and calculate the peak interval of the autocorrelation function of the phase information respectively;

[0041] Target recognition module: used to estimate the signal frequency of each stationary target based on the peak interval of the autocorrelation function, and eliminate stationary targets whose signal frequencies are not within the range of human breathing frequency to obtain stationary human targets.

[0042] A third aspect of the present invention discloses an electronic device comprising: at least one processor, at least one memory, a communication interface, and a bus;

[0043] The processor, memory, and communication interface communicate with each other through the bus.

[0044] The memory stores program instructions that can be executed by the processor, which invokes the program instructions to implement the method as described in the first aspect of the present invention.

[0045] In a fourth aspect, the present invention discloses a computer-readable storage medium storing computer instructions that cause a computer to perform the method described in the first aspect of the present invention.

[0046] The present invention has the following advantages over the prior art:

[0047] 1) This invention acquires and analyzes the intermediate frequency signal of FMCW radar to obtain a range-slow time image, extracts the phase information of all stationary targets from it, and determines the frequency information of each stationary target by the peak interval of the autocorrelation function. By judging whether the frequency of the stationary target is within the frequency range of the human breathing signal, the stationary human target can be distinguished from other interference signals, which can effectively reduce the probability of false alarms during detection and improve the accuracy of detecting stationary human targets behind the wall.

[0048] 2) This invention effectively reduces or eliminates the influence of clutter by performing static clutter suppression processing on each range cell of the range-slow time image, thereby improving the detection sensitivity of stationary target signals. Subsequently, energy accumulation in the range dimension is performed on the filtered range-slow time image to improve the signal-to-noise ratio, thus better distinguishing between stationary targets and clutter. One-dimensional CFAR detection can effectively detect the range gate number of all stationary targets and accurately extract the phase information of stationary targets in complex backgrounds, thereby helping to achieve reliable detection and localization of stationary targets.

[0049] 3) This invention calculates the autocorrelation function of the phase information of each stationary target's range gate, determines the signal period and frequency based on the peak interval of each normalized autocorrelation function, and obtains the frequency of the stationary target. This allows for more accurate extraction of the frequency domain features of the stationary target, thus improving the radar's ability to detect stationary targets.

[0050] 4) To further reduce false alarms, this invention performs multiple detections and verifications on identified stationary targets to determine valid stationary human targets. The distance to the valid stationary human targets is calculated based on the index of the same distance door detected multiple times, thereby achieving accurate detection and positioning of concealed human targets indoors. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0052] Figure 1 This is a flowchart of the method for detecting stationary human targets based on FMCW through-wall radar according to the present invention;

[0053] Figure 2 This is a schematic diagram of the range gate for two stationary targets obtained by one-dimensional CFAR detection based on the energy accumulation results of a certain range-slow time image.

[0054] Figure 3 for Figure 2 A schematic diagram of the phase information of the distance gate for the two detected targets;

[0055] Figure 4 for Figure 3 A schematic diagram of the normalized autocorrelation function and its peak value for the phase information of two stationary targets. Detailed Implementation

[0056] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0057] Extensive field data on human targets behind walls show that weak heartbeat signals are easily obscured by environmental clutter and noise, making them difficult to observe, while respiratory signals are relatively robust vital signs. Therefore, this invention proposes a method and system for detecting stationary human targets based on FMCW through-wall radar. This method achieves the detection of stationary targets behind walls by first converting the L-band FMCW radar intermediate frequency signal to obtain a range-slow-time profile. Then, the phase information of the stationary target is calculated based on the range-slow-time profile. The phase frequency is estimated using the peak value of the autocorrelation function. Finally, by determining whether the frequency falls within the range of human breathing frequencies, stationary human targets are distinguished from other interference signals, effectively reducing the probability of false alarms.

[0058] Please see Figure 1 This invention discloses a method for detecting stationary human targets based on FMCW through-wall radar, the method comprising:

[0059] S1. Acquire the echo signal of the FMCW radar, mix it to obtain an intermediate frequency signal, sample the intermediate frequency signal to obtain a two-dimensional signal matrix.

[0060] Step S1 specifically includes the following sub-steps:

[0061] S11. Transmit electromagnetic waves outward through the FMCW radar and receive echo signals.

[0062] Specifically, when the FMCW radar is operating, the modulation signal generator produces a modulation signal to control the voltage-controlled oscillator to generate a high-frequency continuous transmission signal, which is then emitted as electromagnetic waves through the antenna.

[0063]

[0064] s t (t) represents the transmitted signal at time t, A t Let f0 be the amplitude of the transmitted signal, f0 be the starting frequency of the linear frequency modulated continuous wave, and μ be the frequency modulation slope. The initial phase is random.

[0065] Simultaneously, one of the transmitted signals is also sent to the mixer as the local oscillator signal.

[0066] The emitted electromagnetic waves propagate through space and scatter upon encountering a target. The reflected electromagnetic wave signal is a delayed version of the emitted signal; therefore, the echo signal can be represented as:

[0067]

[0068] Among them, s r (t) represents the echo signal at time t, A r τ is the amplitude of the echo signal. d Delay for the echo signal.

[0069] In an embodiment of the present invention, assuming a stationary human target is at a distance of R0 from the FMCW radar, the relationship between the rise and fall of the chest cavity caused by the human body's breathing and heartbeat and time is r(t), and the speed of light is C, then the echo signal delay...

[0070] S12. The echo signal is mixed with the local oscillator signal through a mixer to obtain the intermediate frequency signal.

[0071] Will be with echo signal s r (t) and the local oscillator signal s t (t) is mixed to obtain the intermediate frequency signal s. b (t), the formula is:

[0072]

[0073] Among them, s b (t) is the intermediate frequency signal, s t (t) is the local oscillator signal, i.e., the transmitted signal.

[0074] Based on the above intermediate frequency signal s b From the formula for (t), we can obtain the frequency of the intermediate frequency signal as: The phase of the intermediate frequency signal is: λ is the wavelength of the electromagnetic wave.

[0075] S13. The intermediate frequency signal is sampled by an analog-to-digital converter (ADC) to obtain intermediate frequency sampling data.

[0076] The intermediate frequency signal s is processed by an analog-to-digital converter (ADC). b (t) is sampled, and the formula is:

[0077]

[0078] Among them, s b (n) represents the intermediate frequency sampling data, n = 0, 1, ..., N-1, where n represents the sampling point number and N is the total number of sampling points.

[0079] S14. Extract one frame every P frames from the intermediate frequency sampling data, and process it through M×P×T. c After a certain period of time, a two-dimensional signal matrix s of radar echo is obtained.b (m,n).

[0080] Where M represents the total number of frames, T c The scan time length of one chirp.

[0081] S2. Perform a distance-dimensional FFT on the two-dimensional signal matrix row by row to obtain a distance-slow time image.

[0082] For a two-dimensional signal matrix s b Calculate the distance dimension FFT row by row (m,n) to obtain the distance-slow time image S(m,k), where m is the frame number of the slow time dimension and k is the distance cell number of the distance dimension.

[0083] In the obtained distance-slow time image S(m,k), the amplitude spectrum of the distance dimension can reflect the magnitude of different frequency components in the intermediate frequency signal. Different frequency components correspond to the distances of different objects that cause the reflected echo. The larger the peak value of the amplitude spectrum, the stronger the reflected echo signal caused by the object at the corresponding distance.

[0084] S3. Extract the phase information of all stationary targets based on the distance-slow time image, and calculate the peak interval of the autocorrelation function of the phase information respectively.

[0085] Step S3 specifically includes the following sub-steps:

[0086] S31. Perform static clutter suppression processing on each range cell of the range-slow time image to obtain the filtered range-slow time image.

[0087] Since the echoes from stationary targets on walls and indoors have inherent zero-frequency components and high signal strength, their sidelobes can interfere with or even drown out human vital signals, so stationary clutter suppression is necessary.

[0088] Specifically, static clutter suppression is performed on each range cell along the time dimension of the range-slow time image S(m,k), using the following formula:

[0089]

[0090] Where m = 0, 1, ..., M-1, m is the frame number in the slow time dimension, and M is the total number of frames; k = 0, 1, ..., K-1, k is the distance cell number in the distance dimension, and K is the total number of distance cells; S(m,k), This is the m-th frame distance-slow time image before and after static clutter suppression processing.

[0091] S32. Perform energy accumulation in the distance dimension on the filtered distance-slow time image to obtain the energy accumulation result.

[0092] Specifically, the magnitude matrix is ​​obtained by taking the modulus of the filtered range-slow time image. Then, the energy of each distance unit is accumulated along the time dimension, using the following formula:

[0093]

[0094] Where X(k) is the energy accumulation result.

[0095] S33. Perform one-dimensional CFAR detection on the energy accumulation results to obtain the range gate numbers of all stationary targets.

[0096] In an embodiment of the present invention, one-dimensional CFAR detection is performed on the energy accumulation result X(k) to obtain the index of the distance gate for all stationary targets: a1, a2, ..., a P , where P is the total number of detected stationary targets. For example Figure 2 The diagram shows a distance gate for two stationary targets obtained by one-dimensional CFAR detection based on the energy accumulation results of a certain distance-slow time image.

[0097] S34. Based on the index of the range gate where all stationary targets are located, extract the phase information of the range gate where all stationary targets are located along the time dimension.

[0098] In embodiments of the present invention, for P distance gates a1, a2, ..., a P Extract phase information of P stationary targets along the time dimension. p=1,2,…,P, m=0,1,…,M-1, Figure 3 As shown Figure 2 A schematic diagram of the phase information of the distance gates of the two detected targets, where a) is target one and b) is target two.

[0099] This invention effectively reduces or eliminates the influence of clutter by performing static clutter suppression processing on each range cell of the range-slow time image, thereby improving the detection sensitivity of stationary target signals. Subsequently, energy accumulation in the range dimension is performed on the filtered range-slow time image to improve the signal-to-noise ratio, thus better distinguishing between stationary targets and clutter. One-dimensional CFAR detection can effectively detect the range gate number of all stationary targets and accurately extract the phase information of stationary targets in complex backgrounds, thereby helping to achieve reliable detection and localization of stationary targets.

[0100] S35. Calculate the peak interval of the autocorrelation function of the phase information respectively.

[0101] Step S35 specifically includes the following sub-steps:

[0102] S351. Calculate the autocorrelation function of the phase information of the range gates of P stationary targets respectively. The formula is:

[0103]

[0104] Among them, R p (i) represents the autocorrelation function value of the phase information of the range gate of the p-th stationary target, where m and i are both frame numbers. These are the phase information of the m-th and (m+i)-th frames of the distance gate for the p-th stationary target, respectively, where p = 1, 2, ..., P, and P is the total number of detected stationary targets.

[0105] S352. Normalize the autocorrelation functions respectively to obtain the normalized autocorrelation functions:

[0106]

[0107] Among them, R p,norm (i) is the autocorrelation function of the phase information of the range gate of the p-th stationary target.

[0108] S353. Calculate the peak value of the normalized autocorrelation function, and calculate the time interval between the peak values ​​of each autocorrelation function as the peak interval of the autocorrelation function.

[0109] Specifically, the normalized autocorrelation function R is detected. p,norm The signal peak value of (n) is used to derive R by establishing a peak index. p,norm The time interval between the peaks of (n) is the peak interval of the autocorrelation function. For example... Figure 4 As shown Figure 3 The diagram shows the normalized autocorrelation function and peak values ​​of the phase information of two stationary targets, where a) represents target one and b) represents target two. According to... Figure 4 It can be seen that the signal peak value and peak interval of each stationary target can be identified more quickly by using the normalized autocorrelation function.

[0110] This invention calculates the autocorrelation function of the phase information of each stationary target's range gate, determines the signal period and frequency based on the peak interval of each normalized autocorrelation function, and obtains the frequency of the stationary target. This allows for more accurate extraction of the frequency domain characteristics of stationary targets, improving the radar's ability to detect stationary targets.

[0111] S4. Estimate the signal frequency of each stationary target based on the peak interval of the autocorrelation function, and eliminate stationary targets whose signal frequencies are not within the range of human breathing frequency to obtain stationary human targets.

[0112] Calculate the normalized autocorrelation function R p,normThe average value of the peak intervals of (n) is the period of the signal. The reciprocal of the period of the signal is converted into the frequency, which is the signal frequency of the corresponding stationary target.

[0113] As an example, calculate Figure 4 The peak intervals of the autocorrelation functions of the two stationary targets were calculated, and their average values ​​were obtained. The periods of the two stationary targets, one and two, were 3.2s and 0.97s, respectively, which were converted into frequencies of 0.3125Hz and 1.0294Hz, respectively. Given that the range of human breathing frequency is 0.2 to 0.5Hz, the frequency of stationary target two is not within this range. Therefore, stationary target two is judged to be a non-human target and is removed. The remaining targets can be considered as stationary human targets.

[0114] S5. Perform multiple detections on the stationary human target to determine the effective stationary human target and calculate the distance of the effective stationary human target.

[0115] To further reduce false alarms, the identified target can be detected and judged multiple times. If the number of times the same distance gate is detected in n consecutive detection results is greater than or equal to a preset threshold Y, then the stationary human target is determined to be a valid stationary human target, and the valid stationary human target is taken as the final detection result.

[0116] Meanwhile, the distance gate index corresponding to this effective stationary human target is denoted as a. j , through a j Value calculation of the distance between the effective stationary human target and the radar:

[0117]

[0118] Where D represents the distance between the effective stationary human target and the radar, C represents the speed of light, μ is the frequency modulation slope, and T... c The scan time length of one chirp.

[0119] This invention verifies the effectiveness of the detection results through multiple tests and calculates the distance of a valid stationary human target based on the index of the same distance door detected multiple times, thereby achieving accurate detection and positioning of concealed human targets indoors.

[0120] In summary, this invention acquires and analyzes the intermediate frequency signals of an FMCW radar to obtain a range-slow time image, extracts the phase information of all stationary targets from it, and determines the frequency information of each stationary target by the peak interval of the autocorrelation function. By judging whether the frequency of the stationary target is within the frequency range of the human breathing signal, it can distinguish stationary human targets from other interference signals, which can effectively reduce the probability of false alarms during detection and improve the accuracy of stationary human target detection.

[0121] Corresponding to the above method embodiments, the present invention also discloses a system for detecting stationary human targets based on FMCW through-wall radar, the system comprising:

[0122] Preprocessing module: used to acquire the echo signal of FMCW radar, mix it to obtain intermediate frequency signal, sample the intermediate frequency signal to obtain a two-dimensional signal matrix;

[0123] Signal transformation module: used to perform a distance-dimensional FFT on the two-dimensional signal matrix row by row to obtain a distance-slow time image;

[0124] Phase analysis module: used to extract phase information of all stationary targets based on the range-slow time image, and calculate the peak interval of the autocorrelation function of the phase information respectively;

[0125] Target recognition module: used to estimate the signal frequency of each stationary target based on the peak interval of the autocorrelation function, and eliminate stationary targets whose signal frequencies are not within the range of human breathing frequency to obtain stationary human targets.

[0126] The above system embodiments and method embodiments are one-to-one correspondences. For a brief description of the system embodiments, please refer to the method embodiments.

[0127] The present invention also discloses an electronic device, comprising: at least one processor, at least one memory, a communication interface, and a bus; wherein the processor, memory, and communication interface communicate with each other through the bus; the memory stores program instructions executable by the processor, and the processor calls the program instructions to implement the aforementioned method of the present invention.

[0128] The present invention also discloses a computer-readable storage medium that stores computer instructions, which cause the computer to implement all or part of the steps of the method described in the embodiments of the present invention. The storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0129] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, meaning they can be distributed across multiple network units. Those skilled in the art can select some or all of the modules to achieve the purpose of this embodiment without any inventive effort, based on actual needs.

[0130] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting stationary human targets based on FMCW through-wall radar, characterized in that, The method includes: The echo signal from the FMCW radar is acquired and mixed to obtain an intermediate frequency signal. The intermediate frequency signal is then sampled to obtain a two-dimensional signal matrix. Perform a distance-dimensional FFT on the two-dimensional signal matrix row by row to obtain a distance-slow time image; Phase information of all stationary targets is extracted from the range-slow time image, and the peak interval of the autocorrelation function of the phase information is calculated for each target. The signal frequency of each stationary target is estimated based on the peak interval of the autocorrelation function, and stationary targets whose signal frequencies are not within the range of human breathing frequency are eliminated to obtain stationary human targets; The calculation of the peak interval of the autocorrelation function of the phase information specifically includes: Calculate the autocorrelation function of the phase information of each stationary target's range gate: The autocorrelation functions are normalized separately to obtain normalized autocorrelation functions; Calculate the peak value of the normalized autocorrelation function separately, and calculate the time interval between the peak values ​​of each autocorrelation function as the peak interval of the autocorrelation function. The formula for calculating the autocorrelation function of the phase information of each stationary target's range gate is as follows: ； in, Let m and i be the autocorrelation function values ​​of the phase information of the range gate of the p-th stationary target, where m and i are both frame numbers. These are the phase information of the m-th and (m+i)-th frames of the distance gate for the p-th stationary target, respectively. , This represents the total number of stationary targets detected. The method further includes: The stationary human target is detected multiple times. If the number of times the same distance gate is detected in multiple consecutive detection results is greater than a preset threshold, the stationary human target is determined to be a valid stationary human target, and the valid stationary human target is used as the final detection result. Let the corresponding distance gate index be denoted as a j ,pass a j Value calculation of the effective stationary human target distance: ； Where D represents the effective distance to a stationary human target, and C represents the speed of light. For frequency modulation slope, The scan time length of one chirp.

2. The method for detecting stationary human targets based on FMCW through-wall radar according to claim 1, characterized in that, The sampling of the intermediate frequency signal to obtain a two-dimensional signal matrix specifically includes: The intermediate frequency signal is sampled by an analog-to-digital converter (ADC) to obtain intermediate frequency sampling data; One frame is extracted every P frames from the intermediate frequency sampling data, and then... After a period of time, a two-dimensional signal matrix of radar echoes is obtained; where, Indicates the total number of frames. The scan time length of one chirp.

3. The method for detecting stationary human targets based on FMCW through-wall radar according to claim 1, characterized in that, The extraction of phase information of all stationary targets based on the distance-slow time image specifically includes: Static clutter suppression processing is applied to each range cell of the range-slow time image to obtain the filtered range-slow time image; Energy accumulation in the range dimension is performed on the filtered range-slow time image to obtain the energy accumulation result; One-dimensional CFAR detection is performed on the energy accumulation results to obtain the range gate numbers of all stationary targets; Based on the index of the range gate where all stationary targets are located, the phase information of the range gate where all stationary targets are located is extracted along the time dimension.

4. The method for detecting stationary human targets based on FMCW through-wall radar according to claim 3, characterized in that, The formula for static clutter suppression processing of each range cell in the range-slow time image is as follows: ； in, m is the frame number in the slow time dimension, and M is the total number of frames; k is the index of the distance cell in the distance dimension, and K is the total number of distance cells; , The m-th frame distance-slow time image before and after static clutter suppression processing; The formula for energy accumulation in the range dimension of the filtered range-slow time image is as follows: ； in, This is the result of energy accumulation.

5. A system for detecting stationary human targets based on FMCW through-wall radar, characterized in that, The system includes: Preprocessing module: used to acquire the echo signal of FMCW radar, mix it to obtain intermediate frequency signal, sample the intermediate frequency signal to obtain a two-dimensional signal matrix; Signal transformation module: used to perform a distance-dimensional FFT on the two-dimensional signal matrix row by row to obtain a distance-slow time image; Phase analysis module: used to extract phase information of all stationary targets based on the range-slow time image, and calculate the peak interval of the autocorrelation function of the phase information respectively; Target recognition module: used to estimate the signal frequency of each stationary target based on the peak interval of the autocorrelation function, and eliminate stationary targets whose signal frequencies are not within the range of human breathing frequency to obtain stationary human targets; The calculation of the peak interval of the autocorrelation function of the phase information specifically includes: Calculate the autocorrelation function of the phase information of each stationary target's range gate: The autocorrelation functions are normalized separately to obtain normalized autocorrelation functions; Calculate the peak value of the normalized autocorrelation function separately, and calculate the time interval between the peak values ​​of each autocorrelation function as the peak interval of the autocorrelation function. The formula for calculating the autocorrelation function of the phase information of each stationary target's range gate is as follows: ； in, Let m and i be the autocorrelation function values ​​of the phase information of the range gate of the p-th stationary target, where m and i are both frame numbers. These are the phase information of the m-th and (m+i)-th frames of the distance gate for the p-th stationary target, respectively. , This represents the total number of stationary targets detected. Also includes: The stationary human target is detected multiple times. If the number of times the same distance gate is detected in multiple consecutive detection results is greater than a preset threshold, the stationary human target is determined to be a valid stationary human target, and the valid stationary human target is used as the final detection result. Let the corresponding distance gate index be denoted as a j ,pass a j Value calculation of the effective stationary human target distance: ； Where D represents the effective distance to a stationary human target, and C represents the speed of light. For frequency modulation slope, The scan time length of one chirp.

6. An electronic device, characterized in that, include: At least one processor, at least one memory, a communication interface, and a bus; The processor, memory, and communication interface communicate with each other through the bus. The memory stores program instructions that can be executed by the processor, which invokes the program instructions to implement the method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause the computer to perform the method as described in any one of claims 1 to 4.

Images