A method and apparatus for single-bit SAR signal recovery based on sparsity judgment

Abstract

This invention provides a method and apparatus for recovering single-bit SAR signals based on sparsity assessment, belonging to the field of signal processing technology. The method includes: performing range migration correction on the single-bit quantized spaceborne SAR echo signal to be recovered, obtaining a two-dimensional linear frequency modulated (LFM) signal; dividing the target imaging region into spatial grids, and establishing a SAR imaging observation model based on the LFM signal within the grid points; establishing a sparsity optimization model based on the SAR imaging observation model, and iteratively updating the sparsity optimization model to obtain an optimal estimate of the scene sparsity; and recovering the corrected echo signal based on the optimal estimate to obtain a reconstructed echo signal that meets quality standards. This solution can solve the problem of difficulty in obtaining target scene sparsity during imaging processing.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a method and apparatus for recovering a single-bit SAR signal based on sparsity judgment. Background Technology

[0002] In recent years, with the continuous development of spaceborne SAR (Synthetic Aperture Radar) technology and the continuous improvement of observation capabilities, the demand for high-resolution and wide-swath observations has led to a continuous increase in the amount of SAR echo data, exacerbating the pressure on satellite-to-ground data transmission, inter-satellite data transmission, and inter-satellite coordination. Under some application conditions, the requirements for radiometric accuracy of imaging results are not very stringent; instead, faster data downlink, processing, and distribution capabilities are required. Therefore, single-bit SAR technology has emerged.

[0003] Related techniques typically utilize sparse priors and use optimization algorithms to recover the relative amplitude of echoes, thereby mitigating interference from strong targets in imaging results. However, the optimization algorithm needs to be solved under the condition that the sparsity is known, which means that the SAR image quality cannot be improved in actual imaging processes where the sparsity is unknown.

[0004] Therefore, there is an urgent need for a single-bit SAR signal recovery method and device based on sparsity judgment to solve the above-mentioned technical problems. Summary of the Invention

[0005] This invention provides a method and apparatus for recovering single-bit SAR signals based on sparsity judgment, which can solve the problem of difficulty in obtaining the sparsity of the target scene during imaging processing.

[0006] In a first aspect, embodiments of the present invention provide a single-bit SAR signal recovery method based on sparsity determination, comprising:

[0007] Range migration correction is performed on the single-bit quantized spaceborne SAR echo signal to be recovered to obtain a two-dimensional linear frequency modulated signal; the two-dimensional linear frequency modulated signal includes two dimensions: azimuth and range.

[0008] The target imaging area is divided into spatial grids, and a SAR imaging observation model is established in the grid points based on the two-dimensional linear frequency modulated signal.

[0009] A sparsity optimization model is established based on the SAR imaging observation model, and the sparsity optimization model is iteratively updated to obtain the optimal estimate of scene sparsity.

[0010] The corrected echo signal is recovered based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard.

[0011] Secondly, embodiments of the present invention also provide a single-bit SAR signal recovery device based on sparsity determination, comprising:

[0012] The correction module is used to perform range migration correction on the single-bit quantized spaceborne SAR echo signal to be recovered, so as to obtain a two-dimensional linear frequency modulated signal; the two-dimensional linear frequency modulated signal includes two dimensions: azimuth and range.

[0013] The modeling module is used to divide the target imaging area into spatial grids and establish a SAR imaging observation model in the grid points obtained based on the two-dimensional linear frequency modulated signal.

[0014] The update module is used to establish a sparsity optimization model based on the SAR imaging observation model, and to iteratively update the sparsity optimization model to obtain the optimal estimate of scene sparsity.

[0015] The recovery module is used to recover the corrected echo signal based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard.

[0016] Thirdly, embodiments of the present invention also provide an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the method described in any embodiment of this specification.

[0017] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform the methods described in any embodiment of this specification.

[0018] This invention provides a method and apparatus for single-bit SAR signal recovery based on sparsity assessment. This method leverages the sparse prior characteristics of the scene and employs a sparsity optimization algorithm. Addressing the issue of unknown sparsity in actual imaging scenes, it first performs a sparsity assessment before recovering the echo signal, iteratively estimating the scene's sparsity, and then uses an orthogonal matching pursuit algorithm to recover the single-bit echo signal. This method solves the problem of difficulty in obtaining target scene sparsity during imaging processing by using an adaptive iterative algorithm to estimate the scene's sparsity. Furthermore, compared to traditional compressed sensing-based methods, this method reduces the algorithm's space complexity by using a matrix observation model instead of a vector observation model, making it applicable to wide-swath imaging scenes. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of a single-bit SAR signal recovery method based on sparsity judgment provided in an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of a point target imaging result provided in an embodiment of the present invention;

[0022] Figure 3 This is a schematic diagram of the intensity comparison of false targets provided in an embodiment of the present invention;

[0023] Figure 4 This is a schematic diagram of the simulated imaging result of X-band surface target echo provided in an embodiment of the present invention;

[0024] Figure 5 This is a schematic diagram of the simulated imaging result of C-band surface target echo provided in an embodiment of the present invention;

[0025] Figure 6 This is a schematic diagram of the imaging results of SAR echo data from the Maritime Silk Road-1 satellite, provided in an embodiment of the present invention.

[0026] Figure 7 This is a hardware architecture diagram of an electronic device provided in an embodiment of the present invention;

[0027] Figure 8 This is a structural diagram of a single-bit SAR signal recovery device based on sparsity judgment provided in an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0029] As mentioned earlier, the sparsity of existing technologies is often unknown in actual imaging processes, which makes it difficult to obtain high-quality imaging that meets the requirements after recovering the echo signal.

[0030] Based on this, the concept of the present invention is to first determine the sparsity before recovering the echo signal, obtain the sparsity estimate of the scene through iteration, and then use the orthogonal matching pursuit algorithm to recover the single-bit echo signal.

[0031] The specific implementation of the above concept is described below.

[0032] Please refer to Figure 1 This invention provides a method for recovering a single-bit SAR signal based on sparsity determination, the method comprising:

[0033] Step 100: Perform range migration correction on the single-bit quantized spaceborne SAR echo signal to be recovered to obtain a two-dimensional linear frequency modulated signal; the two-dimensional linear frequency modulated signal includes two dimensions: azimuth and range.

[0034] Step 102: Divide the target imaging area into spatial grids, and establish a SAR imaging observation model in the grid points obtained by dividing the area according to the two-dimensional linear frequency modulated signal.

[0035] Step 104: Establish a sparsity optimization model based on the SAR imaging observation model, and iteratively update the sparsity optimization model to obtain the optimal estimate of scene sparsity.

[0036] Step 106: Recover the corrected echo signal based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard.

[0037] In this embodiment of the invention, based on the sparse prior characteristics of the scene, a sparse optimization algorithm is employed to address the problem of unknown sparsity in actual imaging scenes. Before restoring the echo signal, a sparsity assessment is first performed, and an estimated value of the scene's sparsity is obtained through iteration. Then, an orthogonal matching pursuit algorithm is used to restore the single-bit echo signal. This method solves the problem of difficulty in obtaining the sparsity of the target scene during imaging processing by using an adaptive iterative algorithm to estimate the scene's sparsity. Furthermore, compared to traditional compressed sensing-based methods, this method reduces the algorithm's space complexity by using a matrix observation model instead of a vector observation model, making it applicable to wide-swath imaging scenes.

[0038] The following description Figure 1 The execution method of each step is shown.

[0039] First, for step 100, range migration correction is performed on the single-bit quantized spaceborne SAR echo signal to be recovered to obtain a two-dimensional linear frequency modulated signal.

[0040] Due to range migration, there are coupling terms in the azimuth and range components of the SAR echo signal, which can lead to observation model mismatch. Therefore, range migration correction is required to decouple the azimuth and range signals.

[0041] In this embodiment of the invention, the range migration correction process includes: transforming the echo signal to the range Doppler domain and using linear frequency modulation scaling technology to perform residual range migration correction to obtain first correction data, so that the range migration curves at different range gates are corrected to be the same as the range migration curves at the reference slant range.

[0042] The first correction is transformed to the two-dimensional frequency domain according to the distance-to-Fourier transform, and consistent distance migration correction is performed in the two-dimensional frequency domain to obtain the second correction data;

[0043] The second corrected data is transformed to the range-Doppler domain by the inverse Fourier transform of the range direction, and the transformed second corrected data is multiplied by the azimuth compensation factor to obtain a two-dimensional linear frequency modulated signal; wherein, the compensation factor is used to compensate for the residual phase introduced by the linear scaling; the two-dimensional linear frequency modulated signal includes two dimensions: azimuth and range.

[0044] The specific calculation process in the above steps is well known to those skilled in the art and will not be elaborated here.

[0045] Then, for step 102, the target imaging area is divided into spatial grids, and a SAR imaging observation model is established in the grid points obtained based on the two-dimensional linear frequency modulated signal.

[0046] In this embodiment of the invention, the SAR imaging observation model is established in the following manner:

[0047] The target imaging region is divided into P×Q spatial grids; wherein, the number of spatial grids is determined by the swath width, range gate and azimuth gate of the target imaging region, and each value in the grid represents the backscattering coefficient with that grid as the scattering center;

[0048] The SAR imaging observation model is established based on the two-dimensional linear frequency modulated signal:

[0049]

[0050] Where S is the echo signal matrix; X is the original sparse matrix directly established based on the two-dimensional linear frequency modulated signal; Φ A Φ is the azimuth observation matrix composed of the time shift of the azimuth linear frequency modulated signal; R The range observation matrix is ​​composed of the time shift of the range-direction linear frequency modulated signal.

[0051] Specifically, the scene is divided into P×Q spatial grids. To recover the echo using single-bit compressed sensing theory, a SAR imaging observation model must first be established. The SAR echo signal is divided into range and azimuth components as follows:

[0052]

[0053] Where t is the azimuth time; τ is the range time; T a is the target illumination time; c is the speed of light; is the distance between the grid point and the radar; T p k is the pulse width. r To adjust the frequency.

[0054] After range migration correction, the SAR echo signal is decoupled into a two-dimensional linear frequency modulated signal, thus obtaining the matrix observation model of the SAR echo signal. Where, Φ A The azimuth observation matrix is ​​composed of the time shift of the azimuth linear frequency modulated signal.

[0055]

[0056] in, Na is the number of azimuth points of the echo signal, f c R is the carrier frequency; a (i) represents the distance between the grid point in the azimuth direction and the radar when the range direction is at the reference range.

[0057] Φ R The range observation matrix is ​​composed of the time shift of the range linear frequency modulated signal:

[0058]

[0059] in, R r (j) represents the distance between the grid point in the range direction and the radar when the azimuth direction is at the reference range.

[0060] For step 104, a sparsity optimization model is established based on the SAR imaging observation model, and the sparsity optimization model is iteratively updated to obtain the optimal estimate of scene sparsity.

[0061] In this embodiment of the invention, the optimal estimate is obtained in the following manner:

[0062] A sparsity optimization model is established based on the SAR imaging observation model:

[0063]

[0064] in, Y is the reconstructed sparse matrix; Y is the echo matrix of the two-dimensional linear frequency modulated signal.

[0065] The sparse optimization model is iteratively updated using the orthogonal matching pursuit algorithm to obtain the optimal estimate:

[0066] S11. Select multiple atoms that satisfy the correlation criteria of the observation matrix from the two-dimensional linear frequency modulated signal to form a support set;

[0067] S12. Calculate the inner product of the support set and the echo matrix of the two-dimensional linear frequency modulated signal;

[0068] S13. Determine whether the calculation result satisfies the RIP constraint, and update the sparsity estimate based on the determination result to complete one iteration of calculation:

[0069] Determine the initial value k for the sparsity estimate:

[0070] k=(k a +k b ) / 2

[0071] Where, k a =0,k b =M, where M is the upper limit of sparsity;

[0072] Determine whether the calculation result satisfies the RIP constraint conditions. If:

[0073]

[0074] Then, if the sparsity estimate is greater than the target estimate, k is set to... b The value of is updated to the sparsity estimate at this time; where, For the atoms of the supporting set; To support the inner product of the set and the distance migration corrected echo matrix; δ 2K The confidence level of the RIP constraint;

[0075] like:

[0076]

[0077] Then, if the sparsity estimate is less than the target estimate, k is set to... a The value is updated to the sparsity estimate at this time.

[0078] S14. Repeat steps S1-S4 until the number of iterations is reached or the calculation results simultaneously satisfy all RIP constraints, and output the optimal estimate of sparsity.

[0079] Specifically, based on the SAR imaging observation model, compressed sensing algorithms can be used to recover the single-bit quantized echo signal and reconstruct the sparse scene. The problem is first formulated as the following optimization problem:

[0080]

[0081] There are various methods to solve the above optimization problem, such as the orthogonal matching pursuit algorithm. This algorithm selects one or a group of atoms that are most relevant to the observation matrix from the signal as the support set in each iteration, then recovers the signal and calculates the residual with the original signal. New atoms are selected through the residual to update the support set, thereby achieving the purpose of signal reconstruction.

[0082] For example, let Λ = sup(X) be the support set of the original sparse matrix X, and let |·| denote the sparsity of the set. Then |Λ| = K, where K is the target sparsity.

[0083] ||X||0=|sup(X)|

[0084]

[0085] Among them, X Λ Let φ represent the matrix consisting of elements in X indexed by Λ. a For Φ A In a certain column, φ r For Φ R If a certain column, then Called an atom, denoted by Φ Λ This represents a submatrix indexed by Λ.

[0086] Furthermore, the orthogonal matching pursuit algorithm requires a preset scene sparsity before iteration. However, the scene sparsity is unknown during actual imaging. Therefore, it is necessary to estimate the scene sparsity first. This can be achieved by obtaining a set of atoms through matching tests, making its potential slightly smaller than the true sparsity. Let... Then the element r in the i-th row and j-th column of V i,j Representative atom Φ T The inner product with matrix Y, i.e., <Φ T ,Y>。 Take the indices of the first k maximum values ​​in V to obtain the set Λ 0 ,|Λ 0 |=k. Let atom Φ satisfy the parameter (2K, δ) 2K The RIP property of ) if Then the estimated sparsity k > K, which serves as the criterion for judging whether the sparsity estimate is excessive. Similarly, if Then the sparsity is estimated to be k < K.

[0087] Furthermore, based on the above sparsity judgment criteria, a suitable sparsity can be estimated during the iteration process when the sparsity is unknown. The specific process is shown in the following example: First, let k a =0,k b =M, where M is the upper limit of sparsity. In the iterative solution process, the sparsity is first determined by k = (k a +k b ) / 2 as a test, if We know for certain that k > K, therefore let k b =k; if We know for certain that k > K, therefore let k b =k. Update the sparsity estimate k = (k) before the next iteration. a +k b The iteration continues until the constraints of both inequalities are satisfied simultaneously or the maximum number of iterations is reached, at which point the iteration ends and the sparsity estimate at this point is the optimal estimate.

[0088] For step 106, the corrected echo signal is recovered based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard.

[0089] In this embodiment of the invention, the recovery process includes the following steps:

[0090] S21. Perform recovery initialization based on the input two-dimensional linear frequency modulated signal, the azimuth observation matrix and range observation matrix of the two-dimensional linear frequency modulated signal, and the optimal estimate; wherein, the recovery initialization includes residual initialization, atom set initialization and iteration count initialization;

[0091] S22. Select an atom that is most relevant to the current residual and calculate the corresponding sparse matrix using the least squares method; add the selected atom to the atom set.

[0092] S23. Update the next residual R based on the sparse matrix. j+1 And end one iteration of calculation:

[0093] R j+1 =Y-Φ j X j

[0094] Where Y is the echo matrix of the two-dimensional linear frequency modulated signal; Φ j X is the atom most relevant to the residual at the j-th iteration; j Let be the sparse matrix obtained by least squares calculation in the j-th iteration;

[0095] S24. Determine if the number of iterations is equal to the optimal sparsity. If so, end the iteration and output the reconstructed echo signal.

[0096]

[0097] Where X0 is the sparse matrix at the time of the last iteration;

[0098] Otherwise, repeat steps S21-S23 until the number of iterations is the same as the optimal sparsity.

[0099] For example, input: SAR echo after single-bit quantization Observation matrix Φ A ,Φ R Sparsity K. Output: Single-bit reconstructed echo.

[0100] Initialize residual R 0 =Y, select an empty atom set Ω0, iteration count j=0; select an atom most relevant to the current residual, i.e., the matrix inner product |<φ a,i φ r,j ,R k-1 >|The largest atom λ k Add the selected atom to the atom set Ω j In this approach, a set is set up to accommodate the atoms selected each time, thus avoiding the repeated selection of atoms; the sparse matrix X is solved using the least squares method. j Update residual R j+1 =Y-ΦX j If j = j + 1, check if the iteration number j has reached the sparsity K. If not, repeat the above steps; otherwise, end the iteration and output the reconstructed echo.

[0101] The feasibility of the above method is demonstrated below with an example:

[0102] This embodiment is configured with an Intel(R) Core(TM) i7-11800H@2.30GHz, 64GB of memory, and a 64-bit Microsoft Windows 11 system. The simulation software used is MATLAB (R2020a).

[0103] The simulation uses a 2x2 dot matrix scene with a dot matrix spacing of 200m. The imaging area is 1000m x 1000m, divided into 500x500 grids. The simulation parameters for the remaining point targets are shown in Table 1 below.

[0104] Table 1

[0105]

[0106] Figure 2The results of imaging using the original echo, the single-bit quantized echo, and the echo recovered after single-bit quantization are presented. Figure 2 The comparison results show that after single-bit quantization, the echo signal, due to quantization noise and the cross-modulation of strong scattering point amplitudes, leads to sidelobe elevation at the real point target and the appearance of paired false targets. However, after using the proposed algorithm to recover the echo, the original target is accurately recovered while the point target sidelobes and false targets are significantly suppressed. Since the main application scenario of this algorithm is sparsely populated marine ship scenarios, in practical applications, the false targets introduced by single-bit quantization will seriously affect subsequent ship target detection. Therefore, it is necessary to focus on the algorithm's suppression effect on false targets. Figure 3 The intensity comparison at the false target location is given. As can be seen from the figure, the relative intensity of the false target caused by single-bit quantization is above -10dB. However, after single-bit quantization echo recovery, the intensity at this location drops to below -45dB. The suppression effect of the false target is above 30dB, indicating that the method has an excellent suppression effect on false targets.

[0107] Figure 4 The results are experimental findings from X-band surface target echo simulation imaging. The simulation scenario is a naval ship scenario, and the simulation parameters of the X-band surface target are shown in Table 2 below:

[0108] Table 2

[0109]

[0110] Figure 4 The original scene (a) is a distant sea vessel target, (b) is the imaging result after single-bit quantization of the echo signal. It can be seen that there are many false targets in the imaging result. (c) is the imaging result after the single-bit quantized echo signal is restored. False targets are significantly suppressed.

[0111] Figure 5 The results are experimental findings from C-band surface target echo simulation imaging. The simulation scenario is a naval ship scenario. The simulation parameters of the C-band surface target are shown in Table 3 below:

[0112] Table 3

[0113]

[0114]

[0115] Figure 5 The original scene (a) is a distant sea vessel target, (b) is the imaging result after single-bit quantization of the echo signal. It can be seen that there are many false targets in the imaging result. (c) is the imaging result after the single-bit quantized echo signal is restored. False targets are significantly suppressed.

[0116] Finally, the processing results of the actual echo data from the Maritime Silk Road-1 satellite are presented. The imaging scene is also a maritime ship target. Imaging processing is performed on the original echo data, the echo data after single-bit quantization, and the single-bit echo data recovered by the proposed method. Figure 6 The comparison of imaging results is shown. The target in the red box is the real target, while the target in the green box is a false target caused by single-bit quantization. The comparison of the results shows that the method proposed in this invention can preserve the real target while suppressing false targets caused by single-bit quantization.

[0117] like Figure 7 , Figure 8 As shown, this embodiment of the invention provides a single-bit SAR signal recovery device based on sparsity judgment. The device embodiment can be implemented in software, hardware, or a combination of both. From a hardware perspective, as... Figure 7 The diagram shown is a hardware architecture diagram of an electronic device for recovering a single-bit SAR signal based on sparsity judgment, provided in an embodiment of the present invention. (Except for...) Figure 7 In addition to the processor, memory, network interface, and non-volatile memory shown, the electronic device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing packets. Taking software implementation as an example, such as... Figure 8 As shown, a device in a logical sense is formed by the CPU of its host electronic device reading the corresponding computer program from non-volatile memory into memory and running it. This embodiment provides a single-bit SAR signal recovery device based on sparsity judgment, comprising:

[0118] The correction module 800 is used to perform range migration correction on the single-bit quantized spaceborne SAR echo signal to be recovered, so as to obtain a two-dimensional linear frequency modulated signal; the two-dimensional linear frequency modulated signal includes two dimensions: azimuth and range.

[0119] The modeling module 802 is used to divide the target imaging area into spatial grids and establish a SAR imaging observation model in the grid points obtained by dividing the area based on the two-dimensional linear frequency modulated signal.

[0120] The update module 804 is used to establish a sparsity optimization model based on the SAR imaging observation model, and to iteratively update the sparsity optimization model to obtain the optimal estimate of scene sparsity.

[0121] The recovery module 806 is used to recover the corrected echo signal based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard.

[0122] In this embodiment of the invention, when the correction module 800 performs range migration correction on the single-bit quantized spaceborne SAR echo signal to be recovered to obtain a two-dimensional linear frequency modulated (LFM) signal, it specifically performs the following operations: transforming the echo signal to the range-Doppler domain and using LFM scaling technology to perform residual range migration correction to obtain first correction data, so that the range migration curves at different range gates are corrected to be the same as the range migration curve at the reference slant range; transforming the first correction to the two-dimensional frequency domain according to the range-direction Fourier transform, and performing consistent range migration correction in the two-dimensional frequency domain to obtain second correction data; transforming the second correction data to the range-Doppler domain according to the range-direction Fourier transform, and multiplying the transformed second correction data by an azimuth compensation factor to obtain a two-dimensional LFM signal; wherein, the compensation factor is used to compensate for the residual phase introduced by the linear scaling.

[0123] In this embodiment of the invention, when the modeling module 802 performs spatial grid division of the target imaging region and establishes a SAR imaging observation model in the divided grid points based on the two-dimensional linear frequency modulated signal, it specifically performs the following operations:

[0124] The target imaging region is divided into P×Q spatial grids; wherein, the number of spatial grids is determined by the swath width, range gate and azimuth gate of the target imaging region, and each value in the grid represents the backscattering coefficient with that grid as the scattering center;

[0125] The SAR imaging observation model is established based on the two-dimensional linear frequency modulated signal:

[0126]

[0127] Where S is the echo signal matrix; X is the original sparse matrix directly established based on the two-dimensional linear frequency modulated signal; Φ A Φ is the azimuth observation matrix composed of the time shift of the azimuth linear frequency modulated signal; R The range observation matrix is ​​composed of the time shift of the range-direction linear frequency modulated signal.

[0128] In this embodiment of the invention, when the update module 804 performs the following operations to establish a sparsity optimization model based on the SAR imaging observation model and iteratively update the sparsity optimization model to obtain the optimal estimate of scene sparsity: Establishing a sparsity optimization model based on the SAR imaging observation model:

[0129]

[0130] in, Y is the reconstructed sparse matrix; Y is the echo matrix of the two-dimensional linear frequency modulated signal.

[0131] The sparse optimization model is iteratively updated using the orthogonal matching pursuit algorithm to obtain the optimal estimate.

[0132] In this embodiment of the invention, the sparsity optimization model is solved according to the orthogonal matching pursuit algorithm to obtain the optimal estimate of sparsity when the signs are consistent. The steps include: S11, selecting multiple atoms from the two-dimensional linear frequency modulated signal that satisfy the correlation standard of the observation matrix to form a support set; S12, calculating the inner product of the support set and the echo matrix of the two-dimensional linear frequency modulated signal; S13, determining whether the calculation result satisfies the RIP constraint condition, and updating the sparsity estimate according to the determination result to complete one iteration calculation; S14, repeating steps S1-S4 until the number of iterations is reached or the calculation result simultaneously satisfies all RIP constraints, and outputting the optimal estimate of sparsity.

[0133] In this embodiment of the invention, determining the relationship between the calculation result and the RIP discriminant, and updating the sparsity estimate based on the determination result, includes: determining the initial value k of the sparsity estimate:

[0134] k=(k a +k b ) / 2

[0135] Where, k a =0,k b =M, where M is the upper limit of sparsity;

[0136] Determine whether the calculation result satisfies the RIP constraint conditions. If:

[0137]

[0138] Then, if the sparsity estimate is greater than the target estimate, k is set to... b The value of is updated to the sparsity estimate at this time; where, For the atoms of the supporting set; To support the inner product of the set and the distance migration corrected echo matrix; δ 2K The confidence level of the RIP constraint;

[0139] like:

[0140]

[0141] Then, if the sparsity estimate is less than the target estimate, k is set to... a The value is updated to the sparsity estimate at this time.

[0142] It is understood that the structures illustrated in the embodiments of the present invention do not constitute a specific limitation on a single-bit SAR signal recovery device based on sparsity determination. In other embodiments of the present invention, a single-bit SAR signal recovery device based on sparsity determination may include more or fewer components than illustrated, or combine some components, split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0143] The information interaction and execution process between the modules in the above-mentioned device are based on the same concept as the method embodiment of the present invention, and the specific details can be found in the description of the method embodiment of the present invention, and will not be repeated here.

[0144] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements a single-bit SAR signal recovery method based on sparsity judgment in any embodiment of this invention.

[0145] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program causes the processor to perform a single-bit SAR signal recovery method based on sparsity judgment according to any embodiment of this invention.

[0146] Specifically, a system or apparatus equipped with a storage medium may be provided, on which software program code implementing the functions of any of the embodiments described above is stored, and the computer (or CPU or MPU) of the system or apparatus may read and execute the program code stored in the storage medium.

[0147] In this case, the program code read from the storage medium can itself implement the function of any of the above embodiments, and therefore the program code and the storage medium storing the program code constitute part of the present invention.

[0148] Examples of storage media used to provide program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, program code can be downloaded from a server computer via a communication network.

[0149] Furthermore, it should be clear that not only can the program code read by the computer be executed, but also the operating system or other components operating on the computer can be instructed based on the program code to perform some or all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0150] Furthermore, it is understood that the program code read from the storage medium is written to the memory set in the expansion board inserted into the computer or to the memory set in the expansion module connected to the computer. Then, based on the instructions of the program code, the CPU or other components installed on the expansion board or expansion module execute some and all of the actual operations, thereby realizing the function of any of the above embodiments.

[0151] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0152] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as ROM, RAM, magnetic disk, or optical disk.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recovering a single-bit SAR signal based on sparsity judgment, characterized in that, include: Range migration correction is performed on the single-bit quantized spaceborne SAR echo signal to be recovered to obtain a two-dimensional linear frequency modulated signal; the two-dimensional linear frequency modulated signal includes two dimensions: azimuth and range. The target imaging area is divided into spatial grids, and a SAR imaging observation model is established in the grid points based on the two-dimensional linear frequency modulated signal. A sparsity optimization model is established based on the SAR imaging observation model, and the sparsity optimization model is iteratively updated to obtain the optimal estimate of scene sparsity, including: A sparsity optimization model is established based on the SAR imaging observation model: in, This is the original sparse matrix directly established based on the two-dimensional linear frequency modulated signal; The azimuth observation matrix is ​​composed of the time shift of the azimuth linear frequency modulated signal; The range observation matrix is ​​composed of time shifts of the range-direction linear frequency modulated signal; Y is the reconstructed sparse matrix; Y is the echo matrix of the two-dimensional linear frequency modulated signal. The sparse optimization model is iteratively updated using the orthogonal matching pursuit algorithm to obtain the optimal estimate. The corrected echo signal is recovered based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard.

2. The method according to claim 1, characterized in that, The process of performing range migration correction on the single-bit quantized spaceborne SAR echo signal to be recovered to obtain a two-dimensional linear frequency modulated signal includes: The echo signal is transformed into the range Doppler domain, and the residual range migration is corrected using linear frequency modulation scaling technology to obtain the first correction data, so that the range migration curves at different range gates are corrected to be the same as the range migration curves at the reference slant range. The first correction is transformed to the two-dimensional frequency domain according to the distance-to-Fourier transform, and consistent distance migration correction is performed in the two-dimensional frequency domain to obtain the second correction data; The second corrected data is transformed to the range-Doppler domain by the inverse Fourier transform of the range direction, and the transformed second corrected data is multiplied by the azimuth compensation factor to obtain a two-dimensional linear frequency modulated signal; wherein, the compensation factor is used to compensate for the residual phase introduced by the linear scaling.

3. The method according to claim 1, characterized in that, The step of dividing the target imaging region into spatial grids and establishing a SAR imaging observation model in the divided grid points based on the two-dimensional linear frequency modulated signal includes: The target imaging region is divided into A spatial grid; wherein the number of grids in the spatial grid is determined by the swath width, range gate, and azimuth gate of the target imaging region, and each value in the grid represents the backscattering coefficient with that grid as the scattering center; The SAR imaging observation model is established based on the two-dimensional linear frequency modulated signal: Where S is the echo signal matrix; This is the original sparse matrix directly established based on the two-dimensional linear frequency modulated signal; The azimuth observation matrix is ​​composed of the time shift of the azimuth linear frequency modulated signal; The range observation matrix is ​​composed of the time shift of the range-direction linear frequency modulated signal.

4. The method according to claim 3, characterized in that, The step of solving the sparsity optimization model using the orthogonal matching pursuit algorithm to obtain the optimal estimate of sparsity when the signs are consistent includes: S11. Select multiple atoms that satisfy the correlation criteria of the observation matrix from the two-dimensional linear frequency modulated signal to form a support set; S12. Calculate the inner product of the support set and the echo matrix of the two-dimensional linear frequency modulated signal; S13. Determine whether the calculation result satisfies the RIP constraint condition, and update the sparsity estimate based on the determination result to complete one iteration calculation; S14. Repeat steps S1-S4 until the number of iterations is reached or the calculation results simultaneously satisfy all RIP constraints, and output the optimal estimate of sparsity.

5. The method according to claim 4, characterized in that, The determination of the relationship between the calculation result and the RIP discriminant, and the updating of the sparsity estimate based on the determination result, includes: Determine the initial values ​​of the sparsity estimates : in, , M is the upper limit of sparsity; Determine whether the calculation result satisfies the RIP constraint conditions. If: Then, if the sparsity estimate is greater than the target estimate, then... The value of is updated to the sparsity estimate at this time; where, For the atoms of the supporting set; The inner product of the support set and the distance migration corrected echo matrix; The confidence level of the RIP constraint; like: Then, if the sparsity estimate is determined to be less than the target estimate, then... The value is updated to the sparsity estimate at this time.

6. The method according to claim 1, characterized in that, The step of recovering the corrected echo signal based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard includes: S21. Perform recovery initialization based on the input two-dimensional linear frequency modulated signal, the azimuth observation matrix and range observation matrix of the two-dimensional linear frequency modulated signal, and the optimal estimate; wherein, the recovery initialization includes residual initialization, atom set initialization and iteration count initialization; S22. Select an atom that is most relevant to the current residual and calculate the corresponding sparse matrix using the least squares method; add the selected atom to the atom set. S23. Update the residual for the next iteration based on the sparse matrix. And end one iteration of calculation: Wherein, Y is the echo matrix of the two-dimensional linear frequency modulated signal; The atom most relevant to the residual at the j-th iteration; Let be the sparse matrix obtained by least squares calculation in the j-th iteration; S24. Determine if the number of iterations is equal to the optimal sparsity. If so, end the iteration and output the reconstructed echo signal. : in, This is the sparse matrix at the time of the last iteration; Otherwise, repeat steps S21-S23 until the number of iterations is the same as the optimal sparsity.

7. A single-bit SAR signal recovery device based on sparsity judgment, characterized in that, include: The correction module is used to perform range migration correction on the single-bit quantized spaceborne SAR echo signal to be recovered, so as to obtain a two-dimensional linear frequency modulated signal; the two-dimensional linear frequency modulated signal includes two dimensions: azimuth and range. The modeling module is used to divide the target imaging area into spatial grids and establish a SAR imaging observation model in the grid points obtained based on the two-dimensional linear frequency modulated signal. The update module is used to establish a sparsity optimization model based on the SAR imaging observation model, and to iteratively update the sparsity optimization model to obtain the optimal estimate of scene sparsity, including: A sparsity optimization model is established based on the SAR imaging observation model: in, This is the original sparse matrix directly established based on the two-dimensional linear frequency modulated signal; The azimuth observation matrix is ​​composed of the time shift of the azimuth linear frequency modulated signal; The range observation matrix is ​​composed of time shifts of the range-direction linear frequency modulated signal; Y is the reconstructed sparse matrix; Y is the echo matrix of the two-dimensional linear frequency modulated signal. The sparse optimization model is iteratively updated using the orthogonal matching pursuit algorithm to obtain the optimal estimate. The recovery module is used to recover the corrected echo signal based on the optimal estimate to obtain a reconstructed echo signal that meets the quality standard.

8. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed in a computer, causes the computer to perform the method described in any one of claims 1-6.

Images