A radar phase-based stationary human target detection method in complex scenes

Abstract

The present invention relates to a radar phase-based stationary human target detection method in complex scenes, comprising the following steps: shifting radar echo signals in the complex scene from the time domain to the frequency domain to obtain distance information for all target points in the complex scene; extracting a number of stationary target signals from the distance information for all target points; extracting a number of stationary target points in an area of ​​interest from the stationary target signals, and calculating the stationary target point phase for each stationary target point; unwrapping the stationary target point phase of each stationary target point to obtain the true phase of the stationary target point; calculating the absolute average of the true phase of each stationary target point within a target time period; and determining whether each stationary target point is a stationary human target based on the absolute average. This detection method overcomes the problems of existing vital sign detection, such as inaccurate and low sensitivity, resulting in misjudgment of stationary human targets, and slow detection speed, and is still applicable when dynamic and static targets coexist in complex scenes.

Description

Technical Field

[0001] The present invention belongs to the field of radar technology, and in particular relates to a method for detecting stationary human targets in complex scenes based on radar phase. Background Art

[0002] With the rapid development of the Industrial Internet, the demand for sensing is growing. Compared to other detection methods, radar detection offers advantages such as high resolution, wall penetration, all-weather capability, and privacy protection. Traditional radar detection is typically only interested in moving targets, suppressing stationary targets as much as possible. However, in the smart home sector, daily human activities include not only movement but also static states such as sitting and lying down. Furthermore, in national security areas such as earthquake rescue, effective detection of low-speed, slow, and small targets at close range, especially stationary human targets in harsh environments, is a crucial task.

[0003] Investigations have revealed that breathing and heartbeat are important vital signs that distinguish humans from objects such as tables and chairs. Due to the multi-point scattering effect, thoracic spine displacement is the average of breathing and heartbeat. The displacement associated with breathing is on the order of several millimeters, while the displacement associated with the heartbeat is on the order of hundreds of microns. Therefore, simultaneous quantitative extraction of breathing and heartbeat is somewhat difficult. Existing breathing and heartbeat detection methods mainly focus on measuring tiny movements on the surface of the human chest cavity. The breathing and heartbeat signals are then separated and extracted through bandpass filtering or signal decomposition. This method requires several cycles of movement to accumulate sufficient energy to extract breathing and heartbeat, posing a significant challenge to real-time monitoring.

[0004] The prior art proposes a method for accelerating the detection of stationary human targets for impulse-type through-wall radar. First, completely stationary targets in the scene are eliminated through multi-order subtraction. Then, based on the amplitude and periodic characteristics of the breathing and heartbeat signals, points are taken in the fast and slow time dimensions using a sliding window. The time domain is converted to the frequency domain to detect the frequency signal of human respiratory chest movement. As long as a point with a breathing frequency of 0.15Hz to 0.45Hz appears, it is considered a human target point and the detection is terminated. However, this method is only suitable for the case where there are only stationary targets in the scene. The frequency of moving targets is much greater than the frequency of human respiratory chest movement. If there are moving targets in the scene, the frequency signal of human respiratory chest movement will be overwhelmed. In addition, since breathing detection is a weak signal detection, it takes several cycles of breathing movement to accumulate sufficient energy to detect the frequency signal of human respiratory chest movement in the frequency domain, which is very time-consuming.

[0005] In fields such as smart homes and national defense security, scenes may contain moving objects in addition to stationary human subjects. Therefore, in complex scenes, it is necessary to first extract stationary targets. Furthermore, detecting stationary targets based on respiratory chest motion frequency signals requires multiple respiratory cycles to accumulate sufficient energy, resulting in slow stationary target detection. Therefore, existing technologies require detection methods that address the problems of slow stationary human target detection and low sensitivity in complex scenes where both moving and static objects coexist. Summary of the Invention

[0006] In order to solve the above problems existing in the prior art, the present invention provides a method for detecting stationary human targets in complex scenes based on radar phase. The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0007] The embodiment of the present invention provides a method for detecting stationary human targets in complex scenes based on radar phase, comprising the steps of:

[0008] S1. Move the radar echo signal in the complex scene from the time domain to the frequency domain to obtain the distance information of all target points in the complex scene;

[0009] S2. extracting a number of stationary target signals from the distance information of all target points;

[0010] S3, extracting a plurality of stationary target points in an area of ​​interest from the plurality of stationary target signals, and calculating a stationary target point phase for each of the stationary target points;

[0011] S4, unwrapping the stationary target point phase of each stationary target point to obtain the true phase of the stationary target point;

[0012] S5. Calculate the absolute average value of the true phase of each stationary target point within a target time period;

[0013] S6. Determine whether each of the stationary target points is a stationary human target based on the absolute average value.

[0014] In one embodiment of the present invention, step S1 includes:

[0015] Mixing and filtering the radar echo signal to obtain a mixed and filtered signal;

[0016] Perform distance-dimensional fast Fourier transform on the mixed and filtered signal to obtain the distance information of all target points in the complex scene:

[0017]

[0018] Among them, Q'(f) is the distance information of the target point, f is the radar carrier frequency, μ is the frequency modulation slope, τ is the instantaneous delay of the echo, c is the speed of light, λ is the signal wavelength, and R is the distance between the stationary target and the radar.

[0019] In one embodiment of the present invention, step S2 includes:

[0020] A zero Doppler method is used to extract several stationary target signals from the distance information of all target points.

[0021] In one embodiment of the present invention, extracting a plurality of stationary target signals from the distance information of all target points using a zero Doppler method includes:

[0022] Performing Doppler fast Fourier transform on the target point distance information, extracting the signal with Doppler frequency of 0, and obtaining the plurality of stationary target signals.

[0023] In one embodiment of the present invention, step S2 includes:

[0024] A signal mean method is used to extract several stationary target signals from the distance information of all target points.

[0025] In one embodiment of the present invention, a signal mean method is used to extract a plurality of stationary target signals from the distance information of all target points, including:

[0026] The N pulses of the distance cells where all target points are located are accumulated and averaged to obtain the average distance information of each target point;

[0027] Combining the average distance information expression of the stationary target point and the average distance information expression of the moving target point, when it is determined that the average distance information of each target point is equal to the target point distance information, the target point is the stationary target signal; wherein,

[0028] The average distance information expression of the stationary target point is:

[0029]

[0030] The average distance information expression of the moving target point is:

[0031]

[0032] Among them, Q SN '(f) is the average distance information of the stationary target, Q MN '(f) is the average distance information of the moving target, f is the radar carrier frequency, μ is the frequency modulation slope, τ is the instantaneous delay of the echo, c is the speed of light, λ is the signal wavelength, R is the distance between the stationary target and the radar, and N is the number of pulses.

[0033] In one embodiment of the present invention, the stationary target point phase includes the phase of the distance unit where the stationary human target is located and the phase of the distance unit where other stationary targets are located, wherein:

[0034] The phase of the distance unit where the stationary human target is located is:

[0035]

[0036] The phase of the distance unit where the other stationary targets are located is:

[0037]

[0038] Among them, R0 is the position of the stationary human body, X(mt s ) is the chest amplitude of the human body changing with time, m is the number of frames, t s is the frame period, and λ is the signal wavelength.

[0039] In one embodiment of the present invention, step S4 includes:

[0040] For each of the stationary target points, when the absolute value of the phase difference of the stationary target point between two adjacent moments is greater than π, the phase of the stationary target point at the next moment is added by 2π or subtracted by 2π to obtain the true phase of the stationary target point.

[0041] In one embodiment of the present invention, step S6 includes:

[0042] When it is determined that the absolute average value of each of the stationary target points is greater than a preset phase amplitude, the stationary target point is a stationary human target;

[0043] When it is determined that the absolute average value of the stationary target point is less than or equal to the preset phase amplitude, the stationary target point is not a stationary human target.

[0044] In one embodiment of the present invention, the preset phase amplitude is 0.1 mm.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. The detection method of the present invention determines whether the target is a stationary human target by calculating the phase of the stationary target point and the true phase of the stationary target point, without further solving the breathing and heartbeat signals. This overcomes the problem of inaccurate and low sensitivity of existing vital sign detection, which leads to misjudgment of stationary human targets. At the same time, the phase change of the human body reflection signal is used to determine the stationary human target. The stationary human target can be detected without waiting too long, the detection time is short, and the possibility of real-time monitoring is improved.

[0047] 2. The detection method of the present invention is still applicable when dynamic and static targets coexist in complex scenes. For complex scenes where dynamic and static targets coexist, the static targets are extracted first and then the moving targets, avoiding phase interference of the moving targets and improving the scope of application of the detection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A schematic flow chart of a method for detecting stationary human targets in complex scenarios based on radar phase according to an embodiment of the present invention;

[0049] Figure 2 A schematic flow chart of another method for detecting stationary human targets in complex scenarios based on radar phase provided by an embodiment of the present invention;

[0050] Figure 3 The original signal spectrum after mixing and filtering of a radar phase-based stationary human target detection method in complex scenes provided by an embodiment of the present invention;

[0051] Figure 4 A spectrum after fast Fourier transform in the range dimension of a stationary human target detection method based on radar phase in complex scenes provided by an embodiment of the present invention;

[0052] Figure 5 A radar phase-based stationary human target detection method in complex scenarios provided by an embodiment of the present invention uses a zero-Doppler method to extract a stationary target spectrum;

[0053] Figure 6 A radar phase-based stationary human target detection method for complex scenes provided by an embodiment of the present invention uses a signal mean method to extract stationary target spectra respectively;

[0054] Figure 7 A radar phase-based stationary human target detection method for complex scenes provided by an embodiment of the present invention assumes that a target spectrum within 0 to 8 meters is selected;

[0055] Figure 8 A radar phase-based stationary human target detection method for complex scenes provided by an embodiment of the present invention includes a phase change spectrum of a stationary human target within 5 consecutive seconds;

[0056] Figure 9 The present invention provides a method for detecting stationary human targets in a complex scenario based on radar phase in a short time, and a phase change spectrum of non-stationary human targets within 5 consecutive seconds. DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below with reference to specific examples, but the embodiments of the present invention are not limited thereto.

[0058] Example 1

[0059] See Figure 1 and Figure 2 , Figure 1 A schematic diagram of a flow chart of a method for detecting stationary human targets in complex scenarios based on radar phase according to an embodiment of the present invention is provided. Figure 2 A flowchart of another radar phase-based stationary human target detection method in complex scenarios provided by an embodiment of the present invention. Assuming there are a moving target and a stationary human target in the scene, where the moving target moves arbitrarily in the scene and the stationary target stands still in front of the radar, the detection method includes the following steps:

[0060] S1. Move the radar echo signal in the complex scene from the time domain to the frequency domain to obtain the distance information of all target points in the complex scene.

[0061] Specifically, the radar transmission signal expression is:

[0062]

[0063] Among them, A e is the amplitude of the transmitted signal, f c is the radar carrier frequency, μ is the frequency modulation slope, and t is the signal transmission time.

[0064] The expression of the radar receiving signal is:

[0065]

[0066] Among them, A r is the received signal amplitude, τ is the instantaneous delay of the echo, for a stationary target R is the distance between the stationary target and the radar, c is the speed of light 3×10 8 m / s.

[0067] The radar echo signal is mixed and filtered to obtain the mixed and filtered signal. The expression of the mixed and filtered signal is:

[0068] Q(t)=s(t)×r * (t) = Aexp[j2π(f R t+Φ)]

[0069] Where A is the amplitude of the mixed and filtered signal, μ is the frequency modulation slope, R is the distance between the stationary target and the radar, c is the speed of light, τ is the instantaneous delay of the echo, f c is the radar carrier frequency, and λ is the signal wavelength.

[0070] like Figure 3 As shown, Figure 3The original signal spectrum after mixing and filtering of a radar phase-based stationary human target detection method in complex scenes provided by an embodiment of the present invention, Figure 3 The signal in is a time domain signal.

[0071] In order to make the distance features in the signal more obvious, the distance dimension fast Fourier transform is performed on the signal after mixing and filtering, and the signal is moved from the time domain to the frequency domain to obtain the distance information of all target points in the complex scene.

[0072] Among them, the expression of distance dimension fast Fourier transform is:

[0073]

[0074] Among them, Q'(f) is the distance information of the target point, f is the radar carrier frequency, μ is the frequency modulation slope, τ is the instantaneous delay of the echo, c is the speed of light, λ is the signal wavelength, and R is the distance between the stationary target and the radar.

[0075] like Figure 4 As shown, Figure 4 The embodiment of the present invention provides a method for detecting stationary human targets in complex scenes based on radar phase, and a spectrum after distance-dimensional fast Fourier transform. Figure 4 In the figure, the upper part is the spectrum diagram after Fourier transform, and the lower part is the range image after Fourier transform. The moving target and the stationary target in the scene are around 8m and 4m respectively.

[0076] S2. Extract several stationary target signals from the distance information of all target points.

[0077] Since the Doppler frequency of a stationary target is 0, the zero Doppler method can be used to extract several stationary target signals from the distance information of all target points. Specifically, after performing a distance dimension fast Fourier transform, the Doppler dimension fast Fourier transform is performed on the target point distance information to extract the signal with a Doppler frequency of 0 to obtain several stationary target signals. Stationary target signals include stationary people, tables, chairs, etc. Figure 5 As shown, Figure 5 The embodiment of the present invention provides a method for detecting stationary human targets in complex scenes based on radar phase, which uses the zero Doppler method to extract the stationary target spectrum. Figure 5 In the equation, the horizontal axis is the distance dimension, and the vertical axis is the velocity dimension. Figure 5 Only stationary target signals with a speed of 0 are retained.

[0078] In another specific embodiment, a signal mean method is used to extract several stationary target signals from all target point distance information. Specifically, the steps include:

[0079] First, the N pulses of the distance unit where all target points (including stationary target points and moving target points) are located are accumulated and averaged to obtain the average distance information of each target point.

[0080] Combining the average distance information expression of the stationary target point and the average distance information expression of the moving target point, when it is judged that the average distance information of each target point is equal to the target point distance information, the target point is a stationary target signal.

[0081] Specifically, the average distance information of the stationary target point is expressed as:

[0082]

[0083] The average distance information expression of the moving target point is:

[0084]

[0085] Among them, Q SN '(f) is the average distance information of the stationary target, Q MN '(f) is the average distance information of the moving target, f is the radar carrier frequency, μ is the frequency modulation slope, τ is the instantaneous delay of the echo, c is the speed of light, λ is the signal wavelength, R is the distance between the stationary target and the radar, and N is the number of pulses.

[0086] From the above expressions of the average distance information of the stationary target point and the average distance information of the moving target point, it can be seen that the average distance information of the stationary target point is equal to the target point distance information obtained by the distance dimension fast Fourier transform in step S1, that is, Q' SN (f)≈Q'(f); and for the moving target point, since f R Very small, Therefore, the moving target points that are averaged after accumulation can be ignored. Therefore, when it is judged that the average distance information of each target point is equal to the target point distance information, the target point is a stationary target signal. Figure 6 As shown, Figure 6 The embodiment of the present invention provides a method for detecting stationary human targets in complex scenes based on radar phase, which uses the signal mean method to extract the stationary target spectra. Figure 6 In the equation, the horizontal axis is the distance dimension, and the vertical axis is the velocity dimension. Figure 6 Only the stationary target signal with a speed of 0 is retained, and Figure 5 and Figure 6 The extracted stationary target signals are consistent.

[0087] S3. Extracting a plurality of stationary target points in the region of interest from the plurality of stationary target signals, and calculating a stationary target point phase for each stationary target point.

[0088] First, a number of stationary target points in the region of interest are extracted from a number of stationary target signals.

[0089] Specifically, the complex scene is divided, and the stationary target points obtained by step S2 in the area of ​​interest are selected. In a specific embodiment, the area where the stationary target may be located is first predicted based on the distance between the ground area and the radar, such as the area of ​​interest in a home scene, the area where people may be active (living room, bedroom, etc.); in earthquake rescue, the depth at which the human body may be buried is predicted, etc. Then, the stationary target points in the area of ​​interest are screened out. In another embodiment, a stationary target signal is first selected, and then the area of ​​interest is selected based on the distance between the ground area and the radar, and then it is determined whether the selected stationary target signal belongs to the area of ​​interest, thereby screening out the stationary target points in the area. As Figure 7 As shown, Figure 7 A radar phase-based stationary human target detection method for complex scenes provided by an embodiment of the present invention assumes that a target spectrum within 0 to 8 meters is selected. Since the stationary target is before 6 meters, the target point before 6 meters is selected. Figure 7 The position of the middle circle is the selected target point.

[0090] Then, for each stationary target point, its stationary target point phase is calculated.

[0091] Specifically, the stationary target point phase includes the phase of the distance unit where the stationary human target is located and the phase of the distance unit where other stationary targets are located.

[0092] Assuming that a stationary human body is located R0 meters in front of the radar, the distance expression is:

[0093] R=R0+X(mt s )

[0094] Where R0 is the position of the stationary human body, t s is the frame period, m represents the frame number, X(mt s ) represents the chest amplitude of the human body that changes with time.

[0095] From step S2, we can know that the expression of the human body stationary target signal is:

[0096]

[0097] The phase expression of the distance unit where the stationary human target is located is:

[0098]

[0099] Other stationary target signals are expressed as:

[0100]

[0101] The phase expression of the distance unit where other stationary targets are located is:

[0102]

[0103] Among them, R0 is the position of the stationary human body, X(mt s ) is the chest amplitude of the human body changing with time, m is the number of frames, t s is the frame period, and λ is the signal wavelength.

[0104] See Figure 8 and Figure 9 , Figure 8 The embodiment of the present invention provides a method for detecting stationary human targets in complex scenes based on radar phase, and a phase change spectrum of a stationary human target within 5 seconds. Figure 9 The embodiment of the present invention provides a method for detecting stationary human targets in a complex scene based on radar phase in a short time. The phase change spectrum of non-stationary human targets within 5 seconds is obtained by Figure 8 and Figure 9 It can be seen that there is a big difference between the phase change of stationary human targets and the phase change of non-stationary human targets.

[0105] S4. Unwrap the stationary target point phase of each stationary target point to obtain the true phase of the stationary target point.

[0106] Specifically, the true phase variation caused by chest motion can be greater than π, so phase wrapping should be considered when extracting the phase. For each stationary target point, if the absolute value of the phase difference between two adjacent moments is greater than π, the resulting 2π jitter must be compensated for. This is done by adding or subtracting 2π (±2π) from the phase of the next stationary target point to obtain the true phase of the stationary target point. Without signal dewarping, the signal will always be between [-π, π] and will not reflect the true phase variation caused by chest motion.

[0107] Furthermore, when the phase difference of the stationary target point between two adjacent moments is greater than π, the phase of the stationary target point at the next moment is subtracted by 2π; when the phase difference of the stationary target point between two adjacent moments is less than π, the phase of the stationary target point at the next moment is added by 2π.

[0108] S5. Calculate the absolute average of the true phase of each stationary target point within the target time period.

[0109] Specifically, for each stationary target point, the absolute average value of the true phases of all stationary target points within a period of time is calculated. For example, the absolute average value of the true phases of all stationary target points within 5 seconds is calculated.

[0110] S6. Determine whether each stationary target point is a stationary human target based on the absolute average value.

[0111] Specifically, for each stationary target point, the relationship between its absolute average value and the preset phase amplitude is determined: when the absolute average value of each stationary target point is determined to be greater than the preset phase amplitude, the stationary target point is a stationary human target; when the absolute average value of the stationary target point is determined to be less than or equal to the preset phase amplitude, the stationary target point is not a stationary human target.

[0112] In a specific embodiment, the preset phase amplitude is 0.1 mm.

[0113] In a specific embodiment, after determining whether a stationary target point is a stationary human target or not, determine whether the number of iterations i is greater than or equal to the number n of stationary target points. If so, end the iteration; if not, set i=i+1 and return to step S3 to continue extracting several stationary target points in the area of ​​interest from several stationary target signals.

[0114] The detection method of this embodiment determines whether the target is a stationary human target by calculating the phase of the stationary target point and the true phase of the stationary target point, without further solving the breathing and heartbeat signals. This overcomes the problem of inaccurate and low sensitivity of existing vital sign detection, which leads to misjudgment of stationary human targets. At the same time, the phase change of the human body reflection signal is used to determine the stationary human target. The stationary human target can be detected without waiting too long, the detection time is short, and the possibility of real-time monitoring is improved.

[0115] The detection method of this embodiment is still applicable when moving and static targets coexist in complex scenes. For complex scenes where moving and static targets coexist, the static targets are extracted first and then the moving targets, avoiding phase interference of the moving targets and improving the scope of application of the detection method.

[0116] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. For those skilled in the art of the present invention, without departing from the concept of the present invention, several simple deductions or substitutions can be made, which should be considered to fall within the scope of protection of the present invention.

Claims

1. A radar phase-based stationary human target detection method in complex scenes, characterized by: Including steps: S1. Move the radar echo signal in the complex scene from the time domain to the frequency domain to obtain the distance information of all target points in the complex scene; S2. extracting a number of stationary target signals from the distance information of all target points; S3, extracting a plurality of stationary target points in an area of ​​interest from the plurality of stationary target signals, and calculating a stationary target point phase for each of the stationary target points; S4, unwrapping the stationary target point phase of each stationary target point to obtain the true phase of the stationary target point; S5. Calculate the absolute average value of the true phase of each stationary target point within a target time period; S6. Determine whether each of the stationary target points is a stationary human target based on the absolute average value.

2. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 1 is characterized in that: Step S1 includes: Mixing and filtering the radar echo signal to obtain a mixed and filtered signal; Perform distance-dimensional fast Fourier transform on the mixed and filtered signal to obtain the distance information of all target points in the complex scene: Among them, Q'(f) is the distance information of the target point, f is the radar carrier frequency, μ is the frequency modulation slope, τ is the instantaneous delay of the echo, c is the speed of light, λ is the signal wavelength, and R is the distance between the stationary target and the radar.

3. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 1 is characterized in that: Step S2 includes: A zero Doppler method is used to extract several stationary target signals from the distance information of all target points.

4. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 3 is characterized in that: Extracting several stationary target signals from the distance information of all target points using the zero Doppler method includes: Performing Doppler fast Fourier transform on the target point distance information, extracting the signal with Doppler frequency of 0, and obtaining the plurality of stationary target signals.

5. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 1 is characterized in that: Step S2 includes: A signal mean method is used to extract several stationary target signals from the distance information of all target points.

6. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 5 is characterized in that: The signal mean method is used to extract several stationary target signals from the distance information of all target points, including: The N pulses of the distance cells where all target points are located are accumulated and averaged to obtain the average distance information of each target point; Combining the average distance information expression of the stationary target point and the average distance information expression of the moving target point, when it is determined that the average distance information of each target point is equal to the target point distance information, the target point is the stationary target signal; wherein, The average distance information expression of the stationary target point is: The average distance information expression of the moving target point is: Among them, Q SN '(f) is the average distance information of the stationary target, Q MN '(f) is the average distance information of the moving target, f is the radar carrier frequency, μ is the frequency modulation slope, τ is the instantaneous delay of the echo, c is the speed of light, λ is the signal wavelength, R is the distance between the stationary target and the radar, N is the number of pulses, and t is the time it takes to transmit the signal.

7. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 1 is characterized in that: The stationary target point phase includes the phase of the distance unit where the stationary human target is located and the phase of the distance unit where other stationary targets are located, wherein: The phase of the distance unit where the stationary human target is located is: The phase of the distance unit where the other stationary targets are located is: Among them, R0 is the position of the stationary human body, X(mt s ) is the chest amplitude of the human body changing with time, m is the number of frames, t s is the frame period, and λ is the signal wavelength.

8. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 1 is characterized in that: Step S4 includes: For each of the stationary target points, when the absolute value of the phase difference of the stationary target point between two adjacent moments is greater than π, the phase of the stationary target point at the next moment is added by 2π or subtracted by 2π to obtain the true phase of the stationary target point.

9. The method for detecting stationary human targets in complex scenes based on radar phase according to claim 1 is characterized in that: Step S6 includes: When it is determined that the absolute average value of each of the stationary target points is greater than a preset phase amplitude, the stationary target point is a stationary human target; When it is determined that the absolute average value of the stationary target point is less than or equal to the preset phase amplitude, the stationary target point is not a stationary human target.

10. The radar phase-based stationary human target detection method in a cluttered scene according to claim 9, characterized in that: The preset phase amplitude is 0.1 mm.

Images