A PCA-based sparse strip SAR self-focusing imaging method, device and medium

Abstract

The present invention discloses a sparse strip SAR self-focusing imaging method, device and medium based on PCA, the method comprising the following steps: (1) constructing a sparse strip SAR imaging model based on acquired echo data containing phase errors; (2) performing azimuth sub-aperture division processing on the echo data; (3) performing phase-preserving sparse SAR imaging on the sub-aperture echo data; (4) using a PCA algorithm to estimate the phase errors in each sub-aperture echo data; (5) splicing the sub-aperture phase error estimation values ​​to obtain the phase error of the full-aperture echo data; (6) performing phase error correction on the echo data to obtain a focused sparse SAR image. The present invention can effectively estimate and compensate for the phase error contained in the strip SAR data, overcome the disadvantage that the PCA self-focusing algorithm cannot be applied to sparse SAR images under full sampling or downsampling conditions, and realize self-focusing of sparse SAR images.

Description

Technical Field

[0001] The invention belongs to the field of sparse signal processing and microwave imaging, and in particular relates to a PCA-based sparse strip SAR self-focusing imaging method, device and medium. Background Art

[0002] Synthetic Aperture Radar (SAR) is a milestone in the history of radar development. SAR has many advantages, such as being able to work around the clock and in all weather conditions, strong penetration, and a wide detection range. With the continuous development of technology, synthetic aperture radar has been widely used in various fields, such as environmental monitoring, resource exploration, and disaster assessment.

[0003] The motion deviation of the SAR platform and atmospheric disturbances will cause phase errors in the azimuth direction, thereby reducing the quality of SAR images. Even with advanced electronic navigation systems, obtaining focused SAR images is still a major challenge. Therefore, in order to obtain high-quality SAR images, the autofocus algorithm is used to estimate and compensate for unknown phase errors, which plays a vital role in SAR imaging. Phase Curvature Autofocus (PCA) is an efficient and robust autofocus algorithm with good performance in various scenarios and phase error functions. However, since PCA uses adjacent pulses of range-compressed data to estimate the second-order derivative of the phase error, and then obtains the phase error estimate through quadratic integration, it is easy to cause the superposition of phase noise in different pixel units. In the case of low signal-to-noise ratio, the accuracy of the PCA algorithm in estimating the phase error is low.

[0004] Sparse microwave imaging refers to a new microwave imaging system, theory and method formed by introducing sparse signal processing technology into microwave imaging. It can improve the image quality of sparse scenes in full-sampled or downsampled echoes by suppressing noise and sidelobes. However, traditional sparse imaging methods cannot achieve phase preservation, which limits the application of self-focusing algorithms in sparse SAR imaging. The present invention estimates the phase error from the perspective of the image domain and realizes self-focusing of sparse SAR images by utilizing a sparse imaging algorithm that can maintain image phase information. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to provide a PCA-based sparse strip SAR self-focusing imaging method, device and medium to improve the quality of sparse strip SAR imaging.

[0006] Technical solution: The PCA-based sparse strip SAR self-focusing imaging method described in the present invention is specifically implemented as follows:

[0007] (1) Based on the pre-acquired echo data containing phase errors, a sparse strip SAR imaging model is constructed;

[0008] (2) performing azimuth sub-aperture division processing on the echo data to obtain sub-aperture echo data;

[0009] (3) Sparse SAR imaging with phase preservation of sub-aperture echo data based on the BiIST algorithm;

[0010] (4) Estimate the phase error in each sub-aperture echo data based on the PCA algorithm;

[0011] (5) splicing the sub-aperture phase error estimation values ​​obtained in step (4) to obtain the phase error of the full-aperture echo data;

[0012] (6) Perform phase error correction on the full-aperture echo data to obtain a focused sparse SAR image.

[0013] Furthermore, the implementation process of step (1) is as follows:

[0014] In the strip SAR mode, the sparse SAR imaging model of the approximate observation is expressed as:

[0015]

[0016] Among them, S is the echo data; G is the backscattering coefficient of the observation scene; It is an echo simulation operator and an imaging operator for SAR in the range Doppler domain. The reverse process of a and θ r are the two-dimensional downsampling matrices in azimuth and range respectively; v is the Hadamard product; N is Gaussian noise;

[0017] Considering the phase error in azimuth The sparse strip SAR imaging model with phase error is obtained:

[0018]

[0019] in, diag{e} is a diagonal error matrix with main diagonal e. The reconstruction of sparse strip SAR image is achieved by solving the following regularization problem:

[0020]

[0021] in, is the reconstructed sparse strip SAR image, and β is the regularization parameter.

[0022] Furthermore, the implementation process of step (2) is as follows:

[0023] Based on the sparse strip SAR imaging model, the sparse strip SAR imaging model of the i-th sub-aperture is expressed as:

[0024]

[0025] Then the regularization problem of sparse strip SAR image reconstruction of the i-th sub-aperture is expressed as:

[0026]

[0027] Overlapping pulses are set between adjacent sub-apertures to ensure the continuity of the estimated phase errors between the sub-apertures.

[0028] Furthermore, the implementation process of step (3) is as follows:

[0029] (31) Solution of the (n+1)th iteration of the observation scenario for:

[0030]

[0031] Among them, W (n) Represents the intermediate variable of the nth iteration, whose initialization value is W (0) SAR complex image data of the observed scene in the range Doppler domain; is the sparse solution of the observation scene reconstructed at the nth iteration, and its initialization value is an all-zero matrix;

[0032] (32) The solution of the regularization parameter (n+1)th iteration β (n+1) for:

[0033]

[0034] in, is the K+1th largest element value of the image amplitude; K is the sparsity of the scene; the value range of parameter μ is 0<μ -1 <1;

[0035] (33)W (n) The value W of the (n+1)th iteration (n+1) for:

[0036]

[0037] (34) Sparse solution of the (n+1)th iteration of the observation scene for:

[0038]

[0039] (35) Calculate the residual ε of the (n+)th iteration(n+1) :

[0040]

[0041] If ε (n+1) If the value is greater than the set threshold, the algorithm continues to repeat steps (31) to (35) until it is less than the set threshold after m iterations. The algorithm converges and outputs a sparse SAR image that retains phase information.

[0042] Furthermore, the implementation process of step (4) is as follows:

[0043] (41) The strongest scattering point in each range unit is selected, and a rectangular window is added to each range unit with this point as the center. The SAR image of the i-th subaperture after the n-th iteration of windowing is Where τ is the range sampling time, f η is the azimuth sampling frequency;

[0044] (42) Yes Perform inverse Fourier transform in azimuth to obtain the phase history domain of range compression

[0045]

[0046] Where η is the azimuth sampling time, is the inverse Fourier transform in azimuth;

[0047] (43) The phase curvature of the nth iteration based on the maximum likelihood estimate The calculation formula is:

[0048]

[0049] Among them, Δη is an azimuth sampling point;

[0050] (44) Perform a quadratic integration on the phase curvature and use the summation operation to obtain the phase error estimate; the phase error function of the i-th subaperture estimated at the n-th iteration is It is expressed as:

[0051]

[0052]

[0053] Among them, N a is the number of sampling points in azimuth, It is expressed as:

[0054]

[0055] (45) Calculate the RMS value of the phase error estimated by the nth iteration of the PCA algorithm. If it is greater than the threshold set by the algorithm, the phase error estimated by the nth iteration is compensated to the range compression phase history domain. Then perform azimuth Fourier transform to obtain the sub-aperture SAR image, which is used as the input of the (n+1)th iteration of the PCA algorithm:

[0056]

[0057] Among them, F a {·} is the azimuthal Fourier transform;

[0058] (46) Repeat steps (41)-(45) until the RMS value is less than the set threshold, and obtain the phase error estimate of the i-th sub-aperture

[0059] Furthermore, the implementation process of step (5) is as follows:

[0060] Among the overlapping pulses of adjacent sub-apertures, a pulse position is selected as the splicing point, and the phase error estimation value of each sub-aperture is spliced ​​to obtain the phase error of the full-aperture echo data.

[0061]

[0062] Among them, L sub is the number of pulses in the azimuth direction of the sub-aperture echo data; L m is the number of overlapping pulses of adjacent sub-apertures; k is the position of the splicing point of adjacent sub-aperture pulses in the overlapping pulse, and N is the number of sub-apertures.

[0063] Furthermore, the implementation process of step (6) is as follows:

[0064] The phase error of the estimated full-aperture echo data is removed from the echo data, and then a sparse SAR image of the observed scene is obtained by a sparse SAR imaging algorithm based on soft threshold iteration, which is expressed as:

[0065]

[0066] A device according to the present invention includes a memory and a processor, wherein:

[0067] A memory for storing computer programs that can be run on the processor;

[0068] The processor is used to execute the steps of the PCA-based sparse strip SAR self-focusing imaging method as described above when running the computer program.

[0069] A storage medium described in the present invention stores a computer program, and when the computer program is executed by at least one processor, the steps of the PCA-based sparse strip SAR self-focusing imaging method as described above are implemented.

[0070] Beneficial effects: Compared with the prior art, the beneficial effects of the present invention are as follows: based on the constructed sparse strip SAR imaging model, the present invention utilizes the BiIST algorithm to obtain phase-preserved sparse SAR images, and adopts PCA to estimate the phase error. It can realize phase error correction and sparse SAR imaging for full-sampled or down-sampled echo data with phase error, thus overcoming the disadvantage that the PCA autofocusing algorithm cannot be applied to sparse SAR images under full-sampled or down-sampled conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a flow chart of the PCA-based sparse strip SAR self-focusing imaging method;

[0072] Figure 2 The figures are the experimental results of the present invention; wherein (a) is the matched filtering imaging result of the echo data without phase error; (b), (d), and (f) are the matched filtering imaging results with random, quadratic, and sinusoidal phase errors, respectively; (c), (e), and (g) are the result figures obtained by using the present invention after adding the phase errors shown in (b), (d), and (f), respectively. DETAILED DESCRIPTION

[0073] The present invention will be further described below in conjunction with the accompanying drawings.

[0074] The present invention provides a sparse strip SAR self-focusing imaging method based on PCA, such as Figure 1 As shown, the following steps are included:

[0075] Step 1: Based on the acquired echo data containing phase errors, a sparse strip SAR imaging model is constructed.

[0076] In the strip SAR mode, the sparse SAR imaging model of approximate observation can be expressed as:

[0077]

[0078] Among them, S is the echo data; G is the backscattering coefficient of the observation scene; It is an echo simulation operator and an imaging operator for SAR in the range Doppler domain. The reverse process of a and θ r are the two-dimensional downsampling matrices in azimuth and range respectively; ⊙ is the Hadamard product; N is Gaussian noise.

[0079] Inaccurate observation models and transmission interference will cause phase errors in echo data. Since the phase error in the range direction is a linear phase error, it has little effect on the imaging result and can be ignored. Therefore, the present invention only considers the phase error in the azimuth direction, which can be expressed as It can be obtained that the sparse strip SAR imaging model with phase error can be expressed as:

[0080]

[0081] in, diag{e} is a diagonal error matrix with the main diagonal being e. Therefore, the reconstruction of sparse strip SAR images can be achieved by solving the following regularization problem:

[0082]

[0083] in, is the reconstructed sparse strip SAR image, and β is the regularization parameter.

[0084] Step 2: Perform azimuth sub-aperture division on the echo data.

[0085] In strip mode SAR imaging, the antenna illumination direction remains fixed relative to the track direction, and the beam moves along the track at the speed of the radar platform. During the entire echo acquisition time, the illumination area changes continuously. The scattering points distributed in the azimuth direction in the imaging area have different synthetic apertures. Therefore, targets at different azimuth positions in the SAR image have different phase errors. In order to effectively perform self-focusing processing, the echo data needs to be sub-aperture divided in azimuth. Referring to the full-aperture sparse strip SAR imaging model given in step (1), the sparse strip SAR imaging model of the i-th sub-aperture can be expressed as:

[0086]

[0087] It can be obtained that the regularization problem of sparse strip SAR image reconstruction of the i-th sub-aperture can be expressed as:

[0088]

[0089] Since the estimated sub-aperture phase errors need to be spliced ​​later to obtain the phase error of the full-aperture echo data, the sub-aperture division requires setting overlapping pulses between adjacent sub-apertures to ensure the continuity of the estimated phase errors between each sub-aperture.

[0090] Step 3: Perform phase-preserving sparse SAR imaging on the sub-aperture echo data.

[0091] Referring to the regularization problem of sparse strip SAR image reconstruction of the i-th sub-aperture given in step (2), it can be solved by using the BiIST algorithm. Since the BiIST algorithm can obtain a phase-preserving reconstructed SAR image, it can be used in the subsequent PCA-based phase error estimation. Taking the n+1th iteration as an example, the iterative process of BiIST can be expressed as:

[0092] (3.1) Solution of the observation scenario n+1th iteration for:

[0093]

[0094] Among them, W (n) Represents the intermediate variable of the nth iteration, whose initialization value is W (0) SAR complex image data of the observed scene in the range Doppler domain; is the sparse solution of the observation scene reconstructed at the nth iteration, and its initialization value is an all-zero matrix.

[0095] (3.2) The solution of the regularization parameter (n+1)th iteration β (n+1) for:

[0096]

[0097] in, is the K+1th largest element value of the image amplitude; K is the sparsity of the scene; the value range of parameter μ is 0<μ -1 <1.

[0098] (3.3) Intermediate variable W (n) The value W of the (n+1)th iteration (n+1) for:

[0099]

[0100] (3.4) Sparse solution of the (n+1)th iteration of the observation scene for:

[0101]

[0102] (3.5) Calculate the residual ε of the (n+1)th iteration (n+1) :

[0103]

[0104] If ε (n+1) If the value is greater than the set threshold, the algorithm continues to repeat steps (3.1) to (3.5) until it is less than the set threshold after m iterations. The algorithm converges and outputs a sparse SAR image that retains phase information.

[0105] Step 4: The PCA algorithm estimates the phase error in each sub-aperture echo data.

[0106] The sparse SAR image with phase information obtained in step 3 is As the input of the PCA algorithm, taking the nth iteration as an example, the process of the PCA algorithm can be divided into the following four steps:

[0107] (4.1) Select the strongest scattering point in each range unit, and add a rectangular window to each range unit with this point as the center. Calculate the energy of the strongest scattering point. The width of the window function is the number of pixels corresponding to the energy dropping to 10% of the energy of the strongest scattering point. The windowing process can be used to discard data that does not contribute to the phase error estimation, thereby improving the signal-to-noise ratio of the data in the window in the azimuth data domain. Suppose the SAR image of the i-th subaperture after the n-th iteration of windowing is Where τ is the range sampling time, f η is the azimuth sampling frequency.

[0108] (4.2) Azimuth inverse Fourier transform.

[0109] The sub-aperture SAR image after the nth iteration windowing By performing an inverse Fourier transform in azimuth, we can obtain the phase history domain of range compression. It can be expressed as:

[0110]

[0111] Where η is the azimuth sampling time, is the inverse Fourier transform of the azimuth.

[0112] (4.3) Phase curvature estimation.

[0113] Phase curvature based on maximum likelihood estimation at the nth iteration The calculation formula can be expressed as:

[0114]

[0115] Where Δη is an azimuth sampling point. The phase curvature is the second-order derivative of the phase error function. Estimating the phase error of the subaperture image based on the second-order derivative can effectively eliminate the unknown linear phase part in the subaperture phase error function.

[0116] (4.4) Phase error estimation.

[0117] By performing a quadratic integration of the phase curvature, the phase error estimate can be obtained. Since the calculation formula of the phase curvature is obtained through differential operation, the integral operation here can be implemented by summation operation. Thus, the phase error function of the i-th subaperture estimated by the n-th iteration is It can be expressed as:

[0118]

[0119]

[0120] Among them, N a is the number of sampling points in azimuth, It can be expressed as:

[0121]

[0122] (4.5) Update the estimated phase error and obtain the sub-aperture phase error estimate through continuous iteration.

[0123] Calculate the RMS value of the phase error estimated by the nth iteration of the PCA algorithm. If it is greater than the threshold set by the algorithm, the phase error estimated by the nth iteration is compensated to the range compression phase history domain. Then perform azimuth Fourier transform to obtain the sub-aperture SAR image, which is used as the input of the (n+1)th iteration of the PCA algorithm:

[0124]

[0125] Among them, F a {·} is the azimuth Fourier transform. Repeat steps (4.1)-(4.5) until the RMS value is less than the set threshold. This indicates that there is no need to continue estimating the phase error. The phase error estimate of the i-th sub-aperture is obtained.

[0126] Step 5: The sub-aperture phase error estimates are concatenated to obtain the phase error of the full-aperture echo data.

[0127] Perform steps 3 to 4 on the echo data of each sub-aperture to obtain the phase error estimate of each sub-aperture. In the overlapping pulses of adjacent sub-apertures, select a pulse position as the splicing point, splice the phase error estimate of each sub-aperture, and obtain the phase error of the full-aperture echo data. It can be expressed as:

[0128]

[0129] Among them, L sub is the number of pulses in the azimuth direction of the sub-aperture echo data; L mis the number of overlapping pulses of adjacent sub-apertures; k is the position of the splicing point of adjacent sub-aperture pulses in the overlapping pulse, and N is the number of sub-apertures.

[0130] Step 6: Perform phase error correction on the echo data to obtain a focused sparse SAR image.

[0131] The phase error of the estimated full-aperture echo data is removed from the echo data, and then a sparse SAR image of the observed scene is obtained by a sparse SAR imaging algorithm based on iterative soft thresholding (IST), which can be expressed as:

[0132]

[0133] The present invention also provides a device, including a memory and a processor, wherein: the memory is used to store a computer program that can be run on the processor; the processor is used to execute the steps of the PCA-based sparse strip SAR self-focusing imaging method as described above when running the computer program.

[0134] The present invention also provides a storage medium having a computer program stored thereon, and when the computer program is executed by at least one processor, the steps of the PCA-based sparse strip SAR self-focusing imaging method as described above are implemented.

[0135] Taking the measured data of the C-band Sentinel-1 satellite as an example, the present invention is verified according to the parameters in Table 1:

[0136] Table 1 SAR system parameters

[0137]

[0138] The experimental results are as follows Figure 2 As shown in the figure, the echo data is randomly downsampled by 50% in the azimuth and range directions respectively. Figure 2 (a) is the matched filtering imaging result of the downsampled echo data without adding phase error; Figure 2 (b), (d), and (f) are the matched filter imaging results with random, quadratic, and sinusoidal phase errors added, respectively; Figure 2 In (c), (e), and (g), respectively, Figure 2The results obtained by using the sparse strip SAR self-focusing imaging method based on PCA of the present invention after the phase errors in (b), (d) and (f). From the experimental results, it can be found that there is an obvious energy dispersion phenomenon in the imaging result of the downsampled data using the classical matched filtering algorithm. After adding different types of phase errors, the defocusing phenomenon of the image obtained by the matched filtering method is more serious. The sparse strip SAR self-focusing imaging method based on PCA of the present invention can effectively solve such problems. For the downsampled echo data with different types of phase errors added, not only can sparse SAR imaging be realized, but also the phase error can be effectively compensated to obtain high-quality focused sparse images.

Claims

1. A sparse strip SAR self-focusing imaging method based on PCA, characterized in that: The following steps are involved: (1) Based on the pre-acquired echo data containing phase errors, a sparse strip SAR imaging model is constructed; (2) performing azimuth sub-aperture division processing on the echo data to obtain sub-aperture echo data; (3) Sparse SAR imaging with phase preservation of sub-aperture echo data based on the BiIST algorithm; (4) Estimate the phase error in each sub-aperture echo data based on the PCA algorithm; (5) splicing the sub-aperture phase error estimation values ​​obtained in step (4) to obtain the phase error of the full-aperture echo data; (6) Perform phase error correction on the full-aperture echo data to obtain a focused sparse SAR image; The implementation process of step (3) is as follows: (31) Solution of the (n+1)th iteration of the observation scenario for: Among them, W (n) Represents the intermediate variable of the nth iteration, whose initialization value is W (0) SAR complex image data of the observed scene in the range Doppler domain; is the sparse solution of the observation scene reconstructed at the nth iteration, and its initialization value is an all-zero matrix; (32) The solution of the regularization parameter (n+1)th iteration β (n+1) for: in, is the K+1th largest element value of the image amplitude; K is the sparsity of the scene; the value range of parameter μ is 0<μ -1 <1; (33)W (n) The value W of the (n+1)th iteration (n+1) for: (34) Sparse solution of the (n+1)th iteration of the observation scene for: (35) Calculate the residual ε of the (n+1)th iteration (n+1) : If ε (n+1) If the value is greater than the set threshold, the algorithm continues to repeat steps (31) to (35) until it is less than the set threshold after m iterations. The algorithm converges and outputs a sparse SAR image that retains phase information.

2. The PCA-based sparse strip SAR self-focusing imaging method according to claim 1, characterized in that: The implementation process of step (1) is as follows: In the strip SAR mode, the sparse SAR imaging model of the approximate observation is expressed as: Among them, S is the echo data; G is the backscattering coefficient of the observation scene; It is an echo simulation operator and an imaging operator for SAR in the range Doppler domain. The reverse process of a and θ r are the two-dimensional downsampling matrices in azimuth and range respectively; ⊙ is the Hadamard product; N is Gaussian noise; Considering the phase error in azimuth The sparse strip SAR imaging model with phase error is obtained: in, diag{e} is a diagonal error matrix with main diagonal e. The reconstruction of sparse strip SAR image is achieved by solving the following regularization problem: in, is the reconstructed sparse strip SAR image, and β is the regularization parameter.

3. The PCA-based sparse strip SAR self-focusing imaging method according to claim 1, characterized in that: The implementation process of step (2) is as follows: Based on the sparse strip SAR imaging model, the sparse strip SAR imaging model of the i-th sub-aperture is expressed as: Then the regularization problem of sparse strip SAR image reconstruction of the i-th sub-aperture is expressed as: Overlapping pulses are set between adjacent sub-apertures to ensure the continuity of the estimated phase errors between the sub-apertures.

4. The PCA-based sparse strip SAR self-focusing imaging method according to claim 1, characterized in that: The implementation process of step (4) is as follows: (41) The strongest scattering point in each range unit is selected, and a rectangular window is added to each range unit with this point as the center. The SAR image of the i-th subaperture after the n-th iteration of windowing is Where τ is the range sampling time, f η is the azimuth sampling frequency; (42) Yes Perform inverse Fourier transform in azimuth to obtain the phase history domain of range compression Where η is the azimuth sampling time, is the azimuthal inverse Fourier transform; (43) The phase curvature of the nth iteration based on the maximum likelihood estimate The calculation formula is: Among them, Δη is an azimuth sampling point; (44) Perform a quadratic integration on the phase curvature and use the summation operation to obtain the phase error estimate; the phase error function of the i-th subaperture estimated at the n-th iteration is It is expressed as: Among them, N a is the number of sampling points in azimuth, It is expressed as: (45) Calculate the RMS value of the phase error estimated by the nth iteration of the PCA algorithm. If it is greater than the threshold set by the algorithm, the phase error estimated by the nth iteration is compensated to the range compression phase history domain. Then perform azimuth Fourier transform to obtain the sub-aperture SAR image, which is used as the input of the (n+1)th iteration of the PCA algorithm: Among them, F a {·} is the azimuthal Fourier transform; (46) Repeat steps (41)-(45) until the RMS value is less than the set threshold, and obtain the phase error estimate of the i-th sub-aperture 5. The PCA-based sparse strip SAR self-focusing imaging method according to claim 1, characterized in that: The implementation process of step (5) is as follows: Among the overlapping pulses of adjacent sub-apertures, a pulse position is selected as the splicing point, and the phase error estimation value of each sub-aperture is spliced ​​to obtain the phase error of the full-aperture echo data. Among them, L sub is the number of pulses in the azimuth direction of the sub-aperture echo data; L m is the number of overlapping pulses of adjacent sub-apertures; k is the position of the splicing point of adjacent sub-aperture pulses in the overlapping pulse, and N is the number of sub-apertures.

6. The PCA-based sparse strip SAR self-focusing imaging method according to claim 1, characterized in that: The implementation process of step (6) is as follows: The phase error of the estimated full-aperture echo data is removed from the echo data, and then a sparse SAR image of the observed scene is obtained by a sparse SAR imaging algorithm based on soft threshold iteration, which is expressed as:

7. A device, characterized in that: comprising a memory and a processor, wherein: A memory for storing computer programs that can be run on the processor; A processor, configured to execute the steps of the PCA-based sparse strip SAR self-focusing imaging method according to any one of claims 1 to 6 when running the computer program.

8. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by at least one processor, the steps of the PCA-based sparse strip SAR self-focusing imaging method are implemented as described in any one of claims 1 to 6.

Images