An ultra-wideband-based synthetic aperture radar imaging method

Abstract

The application relates to a synthetic aperture radar imaging method based on ultra-wideband, and steps are as follows: S1, fixing an ultra-wideband radar on a moving device, collecting ultra-wideband radar baseband data, and acquiring background noise generated by antenna energy leakage; S2, collecting the same moving direction ultra-wideband radar baseband data, and then performing pretreatment; S3, performing two-dimensional fast Fourier transform processing, so that the ultra-wideband radar data is converted from the time domain to the frequency domain; S4, setting a reference distance, calculating a reference function, calculating phase compensation, and focusing a target at the reference distance; S5, realizing stolt interpolation through mapping or bending of a distance frequency axis, and obtaining a basic imaging result after two-dimensional inverse fast Fourier transform; S6, setting a proportional factor function on a depth axis, combining the proportional factor function with the basic imaging result, realizing depth-based energy compensation on a plane image, and finally constructing an environment model contour through extreme value selection. The precision of environment model construction is improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of indoor mapping, and relates to a synthetic aperture radar imaging algorithm, in particular to an ultra-wideband-based synthetic aperture radar imaging method. BACKGROUND

[0002] High-rise building fire has become one of the main disasters threatening the safety of the public and the development of society in the city. High-rise building fire has the characteristics of fast fire spread, difficult evacuation and high difficulty in fire fighting. Due to the complex structure of high-rise buildings and the concentration of personnel, it is difficult to control and escape once on fire. At this time, understanding the plan of the event site can help rescue personnel choose the best path and effectively implement the rescue plan, so the compilation of building plan is of great significance to search and rescue.

[0003] At present, robots equipped with surveying and mapping equipment are usually used to measure indoor plan, and common surveying and mapping equipment includes laser radar, depth camera and RGB camera. Both the camera and the laser are in the visible light band, which is easily affected by light and weather, and is not suitable for all-weather working environment. At the same time, the laser will damage the optical device, and the camera will bring privacy problems. Although millimeter wave radar is currently widely used in automatic driving, it is not suitable for complex indoor environment modeling, and has the problems of poor precision stability, weak penetration mapping ability, large minimum detection distance and high cost, which is not conducive to popularization. The ultra-wideband technology is not affected by light and weather, has high resolution and strong anti-interference ability, and develops rapidly. Ultra-wideband radar has gradually become a research and application hotspot.

[0004] However, unlike laser radar, ultra-wideband radar has a wide beam and weak directivity, and it is difficult to identify small targets at a long distance. Therefore, a reference synthetic aperture radar imaging technology is proposed, which makes the radar move at a constant speed within a certain distance, forms a virtual antenna array, improves the angle resolution, and provides spatial information of different targets.

[0005] Chinese patent document CN 114720985 A discloses a fast imaging method based on NCS, which replaces the phase approximation based on the Taylor formula expansion, uses a phase approximation method based on the Legendre expansion, improves the surveying and mapping bandwidth, and realizes good focusing at the edge of the surveying and mapping band. However, this method is an approximate imaging algorithm, although it has the advantages of not needing interpolation operation, high imaging efficiency and easy combination with motion compensation process, but this algorithm is commonly used in large oblique imaging model and is not suitable for environment model construction.

[0006] Chinese patent document CN 104407349 A discloses a one-station fixed dual-station low-frequency ultra-wideband SAR frequency domain imaging method. First, based on one-station fixed dual-station low-frequency ultra-wideband SAR echo signal, high-order phase error compensation is performed in two-dimensional frequency domain to obtain preprocessed echo signal. Then, the preprocessed echo signal is processed in range direction and azimuth direction to obtain a SAR image. Although this invention compensates high-order phase error in two-dimensional frequency domain, corrects the error caused by two-order Taylor expansion of range-direction coupling term, reduces the influence of high-order error on imaging processing under low-frequency ultra-wideband signal, and further realizes high-precision imaging processing of one-station fixed dual-station low-frequency ultra-wideband SAR, and obtains a good focused image, but the advantage of dual-station SAR in this mode is that it not only has the advantages of traditional spaceborne and airborne dual-station SAR, but also has lower cost, similar imaging processing complexity to single station, and can realize interferometric measurement by configuring multiple receivers; Overall, this method is also for airborne or spaceborne large squint imaging measurement, and cannot be directly used for building a building environment model.

[0007] Chinese patent document CN 110850409 A discloses a dual-station synthetic aperture radar imaging method based on time reversal, which constructs a dual-station synthetic aperture radar system by using an ultra-wideband wireless transceiver model and a stepping motor module; constructs a frequency domain transmission matrix SF-MDM and FF-MDM according to target scattering signals; singular value decomposition is performed on the two matrices respectively, and a linear space with a left singular vector as a basis is divided into a signal subspace and a noise subspace; SF and FF imaging functions are constructed by using the two subspaces, and high-resolution imaging in each direction can be realized by using the imaging functions. Although this invention has the advantages of stronger mobility and concealment, flexible measurement; can use ultra-wideband signals for target detection, and has strong robustness; but it is a method proposed for dual-station system platform error, large imaging processing algorithm difficulty and complexity, and is not suitable for single-station ultra-wideband radar environment model application scenarios.

[0008] Therefore, it is necessary to propose a synthetic aperture radar imaging method based on single-station ultra-wideband, to improve the precision and richness of the environment model construction, and to lay a foundation for subsequent path planning and navigation. SUMMARY

[0009] The present application provides a synthetic aperture radar imaging method based on ultra-wideband, which forms a virtual antenna array by moving the radar at a fixed speed, increases the angular resolution of the ultra-wideband radar, and improves the precision of the environment model construction.

[0010] To solve the above technical problems, the technical scheme adopted by the present application is that the synthetic aperture radar imaging method based on ultra-wideband has the following specific steps:

[0011] S1: fixing the ultra-wideband radar on a mobile device and placing it in an open environment, collecting the ultra-wideband radar baseband data and obtaining background noise caused by antenna energy leakage;

[0012] S2: moving the mobile device at a fixed speed in the open environment, collecting the ultra-wideband radar baseband data in the same motion direction, and pre-processing the ultra-wideband radar baseband data to obtain the ultra-wideband radar data for imaging;

[0013] S3: performing two-dimensional fast Fourier transform (2D FFT) processing on the ultra-wideband radar data for imaging obtained in the step S2, converting the ultra-wideband radar data from time domain to frequency domain to obtain frequency domain ultra-wideband radar data;

[0014] S4: setting a reference distance and calculating a reference function, and then calculating phase compensation according to the set reference distance and the reference function, i.e. multiplying the frequency domain ultra-wideband radar data obtained in the step S3 by the reference function to focus the target at the reference distance and partially focus the target at non-reference distances;

[0015] S5: realizing stolt interpolation through mapping or bending of the distance-frequency axis, changing the phase of the data in the two-dimensional frequency domain, adjusting the azimuth and range phase, and eliminating residual phase modulation of the second order and above, and finally obtaining a basic imaging result through two-dimensional inverse fast Fourier transform (2D IFFT);

[0016] S6: considering that the target reflection energy will decay with distance, setting a proportional factor function on the depth axis, combining the proportional factor function with the basic imaging result obtained in the step S5 to realize depth-based energy compensation for the planar image, and finally constructing an environmental model contour through extreme value selection.

[0017] Preferably, the background noise of the ultra-wideband radar is obtained by averaging the ultra-wideband radar baseband data in the step S1, and the formula is:

[0018] a n = mean(a) (1);

[0019] wherein a is the data collected by the ultra-wideband radar in the open environment, mean(·) is an average function, and a n is the background noise of the ultra-wideband radar.

[0020] Preferably, the pre-processing in the step S2 is background noise removal, normalization, absolute value taking, statistical filtering and Savitzky-Golay (SG) filtering processing in sequence.

[0021] Preferably, the specific steps of the step S2 are:

[0022] S21: first remove the background noise in the ultra-wideband radar baseband data, the formula is:

[0023] c = b - a n (2);

[0024] Wherein, b is the ultra-wideband radar baseband data, a n is the background noise of the ultra-wideband radar, c is the ultra-wideband radar baseband data after removing the background noise;

[0025] S22: the ultra-wideband radar baseband data after removing the background noise in step S21 is normalized, the formula is:

[0026]

[0027] Wherein, c represents the ultra-wideband radar baseband data after removing the background noise, μ c represents the mean of the ultra-wideband radar baseband data after removing the background noise, σ c represents the standard deviation of the ultra-wideband radar baseband data after removing the background noise, d represents the ultra-wideband radar baseband data after removing the background noise and normalization;

[0028] S23: the absolute value and statistical filtering of the ultra-wideband radar baseband data after step S22 processing, the formula is:

[0029]

[0030] Wherein, μ d represents the mean of the ultra-wideband radar baseband data after removing the background noise and normalization, d j | represents the absolute value of the jth value in the ultra-wideband radar baseband data after removing the background noise and normalization, e j is the result of the ultra-wideband radar base station data after background noise removal, normalization, absolute value and statistical filtering;

[0031] S24: using SG filter to twice smooth the ultra-wideband radar data after removing the background noise and normalization, absolute value and statistical filtering processing in step S23, to obtain the ultra-wideband radar data for imaging.

[0032] Preferably, the step S3 is two-dimensional fast Fourier transform (2D FFT) processing, and the specific steps of converting the ultra-wideband radar data from time domain to frequency domain are:

[0033] S31: selecting the echo signal model of the ultra-wideband radar data for imaging;

[0034] S32: performing fast Fourier transform on the ultra-wideband radar data for imaging in the range direction (along the direction of radar wave emission);

[0035] S33: obtaining an approximation of the azimuth direction fast Fourier transform result by using the stationary phase principle in the azimuth direction, thereby obtaining the frequency domain ultra-wideband radar data.

[0036] Preferably, the echo signal model in the step S31 is represented as:

[0037]

[0038] wherein A p represents the signal gain, which is related to the target material, is a Gaussian model, the mean value is the standard deviation is the pulse width σ, r p (t) represents the slant range of the target from the radar, c represents the light speed, f c represents the radar center frequency, τ represents the time corresponding to the range direction, also referred to as the fast time, t represents the azimuth direction time, also referred to as the slow time; the formula of the fast Fourier transform in the step S32 is:

[0039]

[0040] wherein A p represents the signal gain, which is related to the target material, t represents the azimuth direction time, also referred to as the slow time, f τ represents the frequency corresponding to the range direction, σ represents the pulse width, which is the standard deviation of the Gaussian model, f c represents the radar center frequency, r p (t) represents the slant range of the target from the radar, c represents the light speed;

[0041] The formula of the stationary phase principle in the step S33 is:

[0042]

[0043] wherein A p represents the signal gain, which is related to the target material, f τ represents the frequency corresponding to the range direction, σ represents the pulse width, which is the standard deviation of the Gaussian model, f t represents the frequency corresponding to the azimuth direction, v represents the speed of the moving trolley motion, x represents the azimuth direction distance, r represents the range direction depth, f c represents the radar center frequency, c represents the light speed.

[0044] Preferably, the specific step of calculating the phase compensation in the step S4 according to the set reference distance and the reference function is:

[0045] S41: First, set the distance and radar equivalent velocity at the reference distance, calculate the best phase compensation of the ultra-wideband radar data in the two-dimensional frequency domain;

[0046] S42: Then, according to the phase of the ultra-wideband data for imaging in the two-dimensional frequency domain and the phase compensation at the reference distance, calculate the residual phase in the two-dimensional frequency domain, and realize the focusing of the target at the reference distance.

[0047] Preferably, the formula for calculating the phase compensation in step S41 is:

[0048]

[0049] Where θ ref represents the phase compensation at the reference distance, f t represents the corresponding frequency in the azimuth direction, f τ represents the corresponding frequency in the range direction, v represents the speed of the moving trolley, x ref is the reference distance set in the azimuth direction, r ref is the reference depth set in the range direction, f c represents the radar center frequency, and c represents the speed of light.

[0050] The formula for calculating the residual phase in the two-dimensional frequency domain in step S42 is:

[0051]

[0052] Where θ RFM represents the residual phase in the two-dimensional frequency domain, f t represents the corresponding frequency in the azimuth direction, f τ represents the corresponding frequency in the range direction, v represents the speed of the moving trolley, x ref is the reference distance set in the azimuth direction, r ref is the reference depth set in the range direction, f c represents the radar center frequency, c represents the speed of light, x represents the azimuth distance, and r represents the range depth.

[0053] Preferably, the specific steps for performing depth-based energy compensation on the basic imaging result and obtaining the environmental model contour through extreme value extraction in step S6 are:

[0054] S61: Depth-based energy compensation is achieved by setting a proportionality factor to weaken the noise influence of targets within 4 meters in depth and enhance the energy of targets between 4-8 meters in depth;

[0055] S62: When constructing the environmental model contour, select a fixed percentage of energy values as an adaptive threshold to reduce the influence of low-power multipath noise and Gaussian white noise and improve the adaptability of the system.

[0056] Preferably, the formula for calculating the scaling factor in step S61 is:

[0057]

[0058] Where r is the target's distance depth, max is the target's distance depth maximum, and thre is the set threshold.

[0059] Compared with existing technologies, the advantages of this invention are: the ultra-wideband synthetic aperture radar (UAV) imaging method is unaffected by lighting or weather, has strong anti-interference capabilities, and compared with millimeter-wave radar, the miniaturization of UAV radar significantly reduces its weight, cost, and energy consumption, offering a cost advantage in a wider range of applications. Furthermore, by combining UAV imaging technology with a virtual antenna array formed by moving the radar at a constant speed over a distance, angular resolution is improved without increasing hardware costs or complex antenna arrays, and spatial information of different targets is provided to the system. This UAV-based UAV imaging method can be widely applied in the field of indoor environment modeling. This method is suitable for mobile vehicles or robots incorporating UAV radar chips, using UAV radar baseband data as the observable, and building intelligent platforms that construct environmental models through movement in unfamiliar environments. Attached Figure Description

[0060] Figure 1 This is a flowchart of the ultra-wideband synthetic aperture radar imaging method of the present invention;

[0061] Figure 2 This is a schematic diagram of a scenario for the ultra-wideband synthetic aperture radar imaging method of the present invention;

[0062] Figure 3 This is the original echo signal of the ultra-wideband synthetic aperture radar imaging method of the present invention;

[0063] Figure 4 This is an image of the ultra-wideband synthetic aperture radar imaging method of the present invention;

[0064] Figure 5 This is an environmental model outline diagram of the ultra-wideband synthetic aperture radar imaging method of the present invention. Detailed Implementation

[0065] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solutions of the present invention more clearly, and should not be used to limit the scope of protection of the present invention.

[0066] Example: Figure 1 As shown, the specific steps of this ultra-wideband synthetic aperture radar imaging method are as follows:

[0067] S1: fixing the ultra-wideband radar on a mobile device and placing it in an open environment for about 10 minutes to collect the ultra-wideband radar baseband data and obtain background noise caused by antenna energy leakage; in the embodiment, the mobile device is a mobile trolley; as shown in FIG. 1; Figure 2

[0068] In the step S1, the background noise of the ultra-wideband radar is obtained by averaging the ultra-wideband radar baseband data, and the formula is:

[0069] a n =mean(a) (1);

[0070] wherein a is the data collected by the ultra-wideband radar in an open environment, mean(·) is an average function, and a n is the background noise of the ultra-wideband radar;

[0071] S2: moving the mobile device at a fixed speed of 0.1 m / s-0.5 m / s in an open environment to collect the ultra-wideband radar baseband data in the same moving direction, and then pre-processing the ultra-wideband radar baseband data to obtain the ultra-wideband radar data for imaging;

[0072] The pre-processing in the step S2 includes background noise removal, normalization, absolute value taking, statistical filtering and Savitzky-Golay (SG) filtering in sequence;

[0073] The specific steps of the step S2 are as follows:

[0074] S21: first removing the background noise in the ultra-wideband radar baseband data, and the formula is:

[0075] c=b-a n (2);

[0076] wherein b is the ultra-wideband radar baseband data, a n is the background noise of the ultra-wideband radar, and c is the ultra-wideband radar baseband data after removing the background noise;

[0077] S22: normalizing the ultra-wideband radar baseband data after removing the background noise in the step S21, and the formula is:

[0078]

[0079] wherein c represents the ultra-wideband radar baseband data after removing the background noise, μ c represents the mean of the ultra-wideband radar baseband data after removing the background noise, and σ c ​a standard deviation of the ultra-wideband radar baseband data after removing background noise, d represents the ultra-wideband radar baseband data after removing background noise and normalization;

[0080] S23: taking absolute value and statistical filtering on the ultra-wideband radar baseband data processed in step S22, the formula is:

[0081]

[0082] wherein μ d represents the average value of the ultra-wideband radar baseband data after removing background noise and normalization, d j represents the absolute value of the jth value in the ultra-wideband radar baseband data after removing background noise and normalization, e j is the result of the ultra-wideband radar base station data after background noise removal, normalization, absolute value taking and statistical filtering;

[0083] S24: using an SG filter to perform secondary smoothing on the ultra-wideband radar data after removing background noise and normalization, taking absolute value and statistical filtering in step S23, to obtain ultra-wideband radar data for imaging;

[0084] S3: performing two-dimensional fast Fourier transform (2D FFT) processing on the ultra-wideband radar data for imaging obtained in step S2, to convert the ultra-wideband radar data from time domain to frequency domain, to obtain frequency domain ultra-wideband radar data;

[0085] The specific steps of performing two-dimensional fast Fourier transform (2D FFT) processing on the ultra-wideband radar data in step S3 to convert the ultra-wideband radar data from time domain to frequency domain are:

[0086] S31: selecting an echo signal model of the ultra-wideband radar data for imaging;

[0087] The echo signal model in step S31 is represented as:

[0088]

[0089] wherein A p is a signal gain, which is related to target material, is a Gaussian model, the mean value is the standard deviation is pulse width σ, r p (t) is the slant range of the target from the radar, c is the speed of light, f c is the radar center frequency, τ is the time corresponding to the distance, also known as fast time, t is the azimuth time, also known as slow time;

[0090] S32: performing fast Fourier transform on the ultra-wideband radar data for imaging in the distance direction (along the direction of radar wave emission);

[0091] The formula of the fast Fourier transform in the step S32 is:

[0092]

[0093] wherein A p represents the signal gain, which is related to the target material, t represents the azimuth time, also known as the slow time, f τ represents the corresponding frequency in the range direction, σ represents the pulse width, which is the standard deviation of the Gaussian model, f c represents the radar center frequency, r p (t) represents the slant range of the target from the radar, and c represents the light speed;

[0094] S33: An approximate azimuth fast Fourier transform result is obtained by using the stationary phase principle in the azimuth direction, so as to obtain the frequency domain ultra-wideband radar data;

[0095] The formula of the stationary phase principle used in the step S33 is:

[0096]

[0097] wherein A p represents the signal gain, which is related to the target material, f τ represents the corresponding frequency in the range direction, σ represents the pulse width, which is the standard deviation of the Gaussian model, f t represents the corresponding frequency in the azimuth direction, v represents the speed of the moving trolley, x represents the azimuth distance, r represents the range depth, f c represents the radar center frequency, and c represents the light speed;

[0098] S4: A reference distance is set, a reference function is calculated, and then a phase compensation is calculated according to the set reference distance and the reference function; that is, the frequency domain ultra-wideband radar data obtained in the step S3 is multiplied by the reference function, so that the target at the reference distance is focused; and the targets at non-reference distances are partially focused;

[0099] The specific steps for calculating the phase compensation according to the set reference distance and the reference function in the step S4 are as follows:

[0100] S41: First, the distance and the radar equivalent speed are set at the reference distance, and the best phase compensation of the ultra-wideband radar data in the two-dimensional frequency domain is calculated;

[0101] S42: Then, the residual phase in the two-dimensional frequency domain is calculated according to the phase of the ultra-wideband data used for imaging in the two-dimensional frequency domain and the phase compensation at the reference distance, so as to realize the focusing of the target at the reference distance;

[0102] The formula for calculating the phase compensation in step S41 is:

[0103]

[0104] wherein θ ref represents the phase compensation at the reference distance, f t represents the corresponding frequency in the azimuth direction, f τ represents the corresponding frequency in the range direction, v represents the speed of the moving trolley, x ref is the reference distance set in the azimuth direction, r ref is the reference depth set in the range direction, f c represents the radar center frequency, and c represents the speed of light.

[0105] The formula for calculating the residual phase in the two-dimensional frequency domain in step S42 is:

[0106]

[0107] wherein θ RFM represents the residual phase in the two-dimensional frequency domain, f t represents the corresponding frequency in the azimuth direction, f τ represents the corresponding frequency in the range direction, v represents the speed of the moving trolley, x ref is the reference distance set in the azimuth direction, r ref is the reference depth set in the range direction, f c represents the radar center frequency, c represents the speed of light, x represents the azimuth distance, and r represents the range depth.

[0108] S5: Stolt interpolation is achieved through mapping or bending of the range frequency axis, the data phase in the two-dimensional frequency domain is changed, the azimuth and range phase is adjusted, and the residual phase modulation of the second order and above is eliminated, and finally the basic imaging result is obtained through two-dimensional inverse fast Fourier transform (2D IFFT);

[0109] S6: Considering that the target reflection energy will decay with the increase of the distance, a proportional factor function is set on the depth axis, the basic imaging result obtained in step S5 is combined with the proportional factor function to achieve depth-based energy compensation for the planar image, and finally the environmental model contour is constructed through extreme value selection;

[0110] The specific steps for performing depth-based energy compensation on the basic imaging result and obtaining the environmental model contour through extreme value extraction in step S6 are as follows:

[0111] S61: Depth-based energy compensation is achieved by setting a proportional factor to weaken the influence of noise on targets within 4 meters in depth and enhance the energy of targets at a depth of 4-8 meters.

[0112] The proportional factor calculation formula in the step S61 is:

[0113]

[0114] Wherein, r is the depth of the target distance direction, max is the maximum value of the target distance direction depth, and thre is the set threshold value.

[0115] S62: When constructing the environment model profile, a fixed percentage of 70%-90% of the energy value is selected as the adaptive threshold value, the influence of low power multipath noise and Gaussian white noise is reduced, and the adaptability of the system is improved.

[0116] The synthetic aperture radar imaging method based on ultra-wideband of the application has simple structure, improves the angle resolution without increasing the hardware cost and complex antenna array, and provides spatial information of different targets for the system. The synthetic aperture radar imaging method based on ultra-wideband can be widely applied to the intelligent platform for constructing the environment model in an unfamiliar environment. Figure 3 The original echo signal is shown in FIG. 1, and the specific effect result is shown in FIG. 2. Figure 4 Figure 5 Figure 4 The imaging graph is shown in FIG. 3. Figure 5 The environment model profile graph is shown in FIG. 4.

[0117] The above specific embodiments further specifically describe the purpose, technical scheme and beneficial effects of the application, and it should be understood that the above description is only for specific embodiments of the application, and is not used to limit the application. Any non-essential improvement or direct application of the method concept and technical scheme of the application to other occasions within the protection scope of the application is within the protection scope of the application.​​

Claims

1. A method of ultra-wideband based synthetic aperture radar imaging, characterized in that, The specific steps are: S1: fixing the ultra-wideband radar on a mobile device and placing it in an open environment, collecting ultra-wideband radar baseband data, and obtaining background noise generated by antenna energy leakage; S2: moving the mobile device at a fixed speed in the open environment, collecting the ultra-wideband radar baseband data in the same motion direction, and pre-processing the ultra-wideband radar baseband data to obtain the ultra-wideband radar data for imaging; S3: performing two-dimensional fast Fourier transform processing on the ultra-wideband radar data for imaging obtained in step S2, converting the ultra-wideband radar data from time domain to frequency domain to obtain frequency domain ultra-wideband radar data; S4: setting a reference distance, calculating a reference function, and then calculating phase compensation according to the set reference distance and the reference function to focus the target at the reference distance; S5: realizing stolt interpolation through mapping or bending of the distance-frequency axis, and finally obtaining the basic imaging result through two-dimensional inverse fast Fourier transform; S6: setting a scaling factor function on the depth axis, combining the scaling factor function with the basic imaging result obtained in step S5 to realize depth-based energy compensation for the planar image, and finally constructing an environmental model contour through extreme value selection.

2. The ultra-wideband-based synthetic aperture radar imaging method of claim 1, wherein, In step S1, the background noise of the ultra-wideband radar is obtained by averaging the ultra-wideband radar baseband data, and the formula is: a n = mean(a)(1); where a is the data collected by the ultra-wideband radar in an open environment, mean(·) is an average function, and a n is the background noise of the ultra-wideband radar.

3. The ultra-wideband-based synthetic aperture radar imaging method of claim 1, wherein, The preprocessing in step S2 is background noise removal, normalization, absolute value taking, statistical filtering, and Savitzky-Golay filtering processing in sequence.

4. The ultra-wideband-based synthetic aperture radar imaging method of claim 3, wherein, The specific steps of step S2 are: S21: first remove the background noise in the ultra-wideband radar baseband data, and the formula is: c = b - a n (2); Wherein, b is the baseband data of the ultra-wideband radar, a n is the background noise of the ultra-wideband radar, and c is the baseband data of the ultra-wideband radar after removing the background noise. S22: normalize the ultra-wideband radar baseband data after removing the background noise in step S21, and the formula is: where c denotes the background noise removed ultra-wideband radar baseband data, μ c denotes the mean of the background noise removed ultra-wideband radar baseband data, σ c denotes the standard deviation of the background noise removed ultra-wideband radar baseband data, d denotes the background noise removed and normalized ultra-wideband radar baseband data; S23: take the absolute value and statistical filtering of the ultra-wideband radar baseband data after step S22 processing, and the formula is: wherein μ |d| represents the average value of the background noise removed and normalized ultra-wideband radar baseband data, |d j represents the absolute value of the jth value in the background noise removed and normalized ultra-wideband radar baseband data, e j is the result of the background noise removed, normalized, absolute value taken, and statistical filtering of the ultra-wideband radar base station data; S24: use the SG filter to perform secondary smoothing on the ultra-wideband radar data after removing the background noise and performing normalization, absolute value taking, and statistical filtering in step S23 to obtain the ultra-wideband radar data for imaging.

5. The ultra-wideband-based synthetic aperture radar imaging method of claim 3, wherein, The specific steps of performing two-dimensional fast Fourier transform processing in step S3 to convert the ultra-wideband radar data from time domain to frequency domain are: S31: select an echo signal model of the ultra-wideband radar data for imaging; S32: perform fast Fourier transform on the ultra-wideband radar data for imaging in the range direction; S33: obtain the approximate fast Fourier transform result in the azimuth direction by using the stationary phase principle to obtain the frequency domain ultra-wideband radar data.

6. The ultra-wideband-based synthetic aperture radar imaging method of claim 5, wherein, The echo signal model in step S31 is represented as: Wherein, A p is a signal gain, related to a target material, is a Gaussian model, the mean is the standard deviation is the pulse width σ, r p (t) is the slant range of the target and the radar, c is the speed of light, f c is the radar center frequency, τ is the time corresponding to the range direction, also known as fast time, t is the azimuth direction time, also known as slow time; the formula of the fast Fourier transform in the step S32 is: where A p represents the signal gain, and is related to the target material, t represents the azimuth time, also known as the slow time, f τ represents the corresponding frequency in the range direction, σ represents the pulse width, and is the standard deviation of the Gaussian model, f c represents the radar center frequency, r p (t) represents the slant range of the target from the radar, and c represents the speed of light The formula for using the stationary phase principle in step S33 is: where A p represents signal gain, and is related to target material, f τ represents frequency corresponding to range direction, and σ represents pulse width, is standard deviation of Gaussian model, f t represents frequency corresponding to azimuth direction, v represents velocity of moving trolley motion, x represents azimuth direction distance, r represents range direction depth, f c represents radar center frequency, and c represents light speed.

7. The ultra-wideband-based synthetic aperture radar imaging method of claim 5, wherein, The specific steps of calculating the phase compensation according to the set reference distance and the reference function in step S4 are: S41: first set the distance and radar equivalent speed at the reference distance to calculate the phase compensation of the ultra-wideband radar data in the two-dimensional frequency domain; S42: According to the phase compensation of the ultra-wideband data for imaging at the phase and reference distance in the two-dimensional frequency domain, the residual phase in the two-dimensional frequency domain is calculated to realize the focusing of the target at the reference distance.

8. The ultra-wideband-based synthetic aperture radar imaging method of claim 7, wherein, The formula for calculating the phase compensation in step S41 is: where θ ref represents phase compensation at the reference distance, f t represents the corresponding frequency in the azimuth direction, f τ represents the corresponding frequency in the range direction, v represents the speed of the moving trolley, x ref is the reference distance set for the azimuth direction, r ref is the reference depth set for the range direction, f c represents the radar center frequency, and c represents the speed of light. The formula for calculating the residual phase in the two-dimensional frequency domain in step S42 is: where θ RFM represents the residual phase in two-dimensional frequency domain, f t represents the corresponding frequency in azimuth direction, f τ represents the corresponding frequency in range direction, v represents the speed of the moving trolley, x ref is the reference distance set in azimuth direction, r ref is the reference depth set in range direction, f c represents the radar center frequency, c represents the speed of light, x represents the azimuth distance, and r represents the range depth.

9. The ultra-wideband-based synthetic aperture radar imaging method of claim 6, wherein, The specific steps for performing depth-based energy compensation on the basic imaging result and obtaining the environmental model contour through extreme value extraction in step S6 are: S61: Depth-based energy compensation is achieved by setting a proportion factor to weaken the noise influence of depth targets within 4 meters and enhance the energy of depth targets between 4-8 meters; S62: When constructing the environmental model contour, a fixed percentage of energy values is selected as an adaptive threshold to reduce the influence of low-power multipath noise and Gaussian white noise and improve the adaptability of the system.

10. The ultra-wideband-based synthetic aperture radar imaging method of claim 9, wherein, The formula for calculating the proportion factor in step S61 is: Wherein, r is the depth of the target distance direction, max is the maximum value of the depth of the target distance direction, and thre is the set threshold value.

Images