Sparse near-field SAR imaging method based on multi-constraint index high-order total variation

Abstract

This invention discloses a sparse near-field SAR imaging method based on high-order total variation with multiple constraint exponents, belonging to the field of radar imaging technology. The method includes the following steps: S1, sampling the scene and preprocessing the scene echo data; S2, calculating the NFCS approximate observation operator; S3, constructing a hybrid sparse near-field SAR imaging model; S4, initializing the parameters of the hybrid sparse near-field SAR imaging model and the NFCS approximate observation operator; S5, updating the hybrid sparse near-field SAR imaging model to complete hybrid constraint sparse near-field SAR imaging; and S6, outputting the imaging results. The method proposed in this invention can accurately reconstruct the target scattering intensity amplitude information, avoiding the underestimation of target scattering point intensity by L1 regularization, while also suppressing speckle noise and avoiding artifacts. It takes into account both the sparsity of the target and the differences in target scattering intensity, improving the imaging performance of complex scenes.

Description

Technical Field

[0001] This invention belongs to the field of radar imaging technology, specifically relating to a sparse near-field SAR imaging method based on high-order total variation with multiple constraint exponents. Background Technology

[0002] Synthetic Aperture Radar (SAR) achieves high-resolution imaging by synthesizing an equivalent large aperture through platform motion, and has important applications in military reconnaissance and geological exploration. Near-field SAR technology can achieve high-precision imaging and has good anti-jamming capabilities, and is widely used in security screening and reversing radar. With the development of SAR imaging technology, SAR imaging faces the dual challenges of massive data volume and high computational complexity.

[0003] Based on L p Regularized SAR imaging methods can effectively promote the sparsity of solutions and suppress noise, making them suitable for imaging sparse scenes. TV regularization can enhance target features in SAR images, preserve edge information, and suppress noise. Combining these two regularization methods linearly to form a new composite regularization function balances target sparsity and continuity, improving imaging performance in complex scenes. However, first-order TV regularization can produce artifacts in imaging. p Regularization underestimates the intensity of the target scattering points, thus affecting image quality. Summary of the Invention

[0004] To address the above problems, this invention proposes a sparse near-field SAR imaging method based on high-order total variation with multiple constraint exponents.

[0005] The technical solution of this invention is: a sparse near-field SAR imaging method based on multi-constraint exponential high-order total variation, comprising the following steps:

[0006] S1. Sample the scene and preprocess the scene echo data;

[0007] S2. Calculate the NFCS approximate observation operator based on the preprocessed scene echo data;

[0008] S3. Construct a hybrid sparse near-field SAR imaging model;

[0009] S4. Initialize the parameters of the hybrid sparse near-field SAR imaging model and the NFCS approximate observation operator;

[0010] S5. Based on the initialization results, update the hybrid sparse near-field SAR imaging model to complete the hybrid constrained sparse near-field SAR imaging.

[0011] S6. Based on near-field SAR hybrid constrained sparse imaging, output the imaging results.

[0012] Furthermore, in S1, the preprocessing includes downsampling, pulse compression, and demodulation.

[0013] Downsampling is performed using radial sampling templates, random sampling templates, or Cartesian sampling templates.

[0014] Furthermore, S2 includes the following sub-steps:

[0015] S21. Calculate the NFCS imaging operator based on the preprocessed scene echo data;

[0016] S22. Calculate the NFCS approximate observation operator based on the NFCS imaging operator.

[0017] Furthermore, in S21, the NFCS imaging operator The expression is:

[0018] ;

[0019] Where Y represents the scene echo data, For Hadamard products, For orientation to Fourier transform, This is the inverse Fourier transform of the orientation. For distance to Fourier transform, For the distance-to-Fourier transform and the distance-to-inverse-Fourier transform, For two-dimensional frequency domain Fourier transform, For the first phase function, For the second phase function, For the third phase function, It is the fourth phase function;

[0020] In S22, the NFCS approximate observation operator The expression is:

[0021] ;

[0022] in, This is a two-dimensional frequency domain inverse Fourier transform. This is the conjugate transpose of the first phase function. This is the conjugate transpose of the second phase function. This is the conjugate transpose of the third phase function. This is the conjugate transpose of the fourth phase function. It is a two-dimensional matrix for scene imaging.

[0023] Furthermore, S3 includes the following sub-steps:

[0024] S31. Construct a higher-order total variational penalty term;

[0025] S32. Construct an exponential penalty term;

[0026] S33. Construct a hybrid sparse near-field SAR imaging model based on the higher-order total variation penalty term and the exponential penalty term.

[0027] Furthermore, in S31, the higher-order total variational penalty term The expression is:

[0028] ;

[0029] in, For gradient operators, For the rows of the matrix, These are the columns of the matrix;

[0030] In S32, the exponential penalty term The expression is:

[0031] ;

[0032] in, Here are the parameters for exponential regularization. For the current two-dimensional imaging scenario, For index;

[0033] In S33, the expression for constructing the hybrid sparse near-field SAR imaging model is:

[0034] ;

[0035] in, These are the weighting parameters used to control SAR time domain and TFXD. The first regularization parameter is used. This is the second regularization parameter. It is the short-time Fourier transform matrix. For approximate observation operators, This is a representation of two-dimensional echo data vectors.

[0036] Furthermore, S5 includes the following sub-steps:

[0037] S51. Update the current two-dimensional imaging scene based on the hybrid sparse near-field SAR imaging model;

[0038] S52. Update the first auxiliary variable of the hybrid sparse near-field SAR imaging model according to the updated current two-dimensional imaging scene.

[0039] S53. Update the first Lagrange multiplier of the hybrid sparse near-field SAR imaging model according to the updated first auxiliary variable;

[0040] S54. Update the first scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model;

[0041] S55. Update the second auxiliary variable of the hybrid sparse near-field SAR imaging model according to the updated current two-dimensional imaging scene;

[0042] S56. Update the second Lagrange multiplier of the hybrid sparse near-field SAR imaging model according to the updated second auxiliary variable;

[0043] S57. Update the second scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model and complete the hybrid constrained sparse imaging of near-field SAR.

[0044] Furthermore, in S51, the expression for updating the current two-dimensional imaging scene is:

[0045] ;

[0046] in, To perform the first step of the ADMM algorithm on g The result of the iteration. For the number of iterations, These are the weighting parameters used to control SAR time domain and TFXD. For approximate observation operators, For imaging operators, This is the short-time inverse Fourier transform. It is the short-time Fourier transform matrix. As the first Lagrange multiplier, For the second Lagrange multiplier, This is a two-dimensional echo data vector representation. Let be the first auxiliary variable for solving the problem in the t-th iteration. Let be the second auxiliary variable for solving the problem in the t-th iteration. For the first scaling augmented Lagrange multiplier obtained in the t-th iteration, The second scaling augmented Lagrange multiplier is obtained in the t-th iteration;

[0047] In S52, the expression for updating the first auxiliary variable of the hybrid sparse near-field SAR imaging model is:

[0048] ;

[0049] in, To use the ADMM algorithm for Conduct the first The result of the iteration. As the first auxiliary variable, The first regularization parameter is used. for The The result of the iteration. For the current two-dimensional imaging scenario, Exponential regularization with the first auxiliary variable as the parameter. This is a generalized iterative thresholding method. To use the ADMM algorithm for Conduct the first The result of the iteration;

[0050] In S53, the expression for updating the first Lagrange multiplier of the hybrid sparse near-field SAR imaging model is:

[0051] ;

[0052] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for the first The result of the iteration;

[0053] In S54, the expression for updating the first scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model is:

[0054] ;

[0055] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration;

[0056] In S55, the expression for updating the second auxiliary variable of the hybrid sparse near-field SAR imaging model is:

[0057] ;

[0058] in, To use the ADMM algorithm for Conduct the first The result of the iteration. As the second auxiliary variable, This is the second regularization parameter. The higher-order total variation penalty term is the one with the second auxiliary variable as a parameter. for Conduct the first The result of the iteration;

[0059] In S56, the expression for updating the second Lagrange multiplier of the hybrid sparse near-field SAR imaging model is:

[0060] ;

[0061] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for No. The result of the iteration;

[0062] In S57, the second scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model is updated:

[0063] ;

[0064] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration is obtained.

[0065] Furthermore, S6 includes the following sub-steps:

[0066] S61. Calculate iterative parameters based on near-field SAR hybrid constrained sparse imaging;

[0067] S62. Determine whether the iteration parameters simultaneously satisfy the following conditions. and If so, return to S5 and continue updating the hybrid sparse near-field SAR imaging model; otherwise, output the imaging result. For iteration parameters, For error parameters, t represents the maximum number of iterations, and t is the initial value of the number of iteration steps.

[0068] Furthermore, in S61, the iteration parameters The expression is:

[0069] ;

[0070] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration is obtained.

[0071] The beneficial effects of this invention are: compared with traditional matched filtering algorithms, the method proposed in this invention can effectively suppress additive noise and clutter; compared with the composite regularization method combining L1 regularization and TV regularization, the method proposed in this invention can accurately reconstruct the target scattering intensity amplitude information, avoid the underestimation of the target scattering point intensity by L1 regularization, suppress speckle noise and avoid artifacts, take into account the sparsity of the target and the difference in target scattering intensity, and improve the imaging performance of complex scenes. Attached Figure Description

[0072] Figure 1 This is a flowchart of a sparse near-field SAR imaging method based on high-order total variation with multiple constraint exponents. Detailed Implementation

[0073] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0074] like Figure 1 As shown, this invention provides a sparse near-field SAR imaging method based on multi-constraint exponential high-order total variation, comprising the following steps:

[0075] S1. Sample the scene and preprocess the scene echo data;

[0076] S2. Calculate the NFCS approximate observation operator based on the preprocessed scene echo data;

[0077] S3. Construct a hybrid sparse near-field SAR imaging model;

[0078] S4. Initialize the parameters of the hybrid sparse near-field SAR imaging model and the NFCS approximate observation operator;

[0079] S5. Based on the initialization results, update the hybrid sparse near-field SAR imaging model to complete the hybrid constrained sparse near-field SAR imaging.

[0080] S6. Based on near-field SAR hybrid constrained sparse imaging, output the imaging results.

[0081] In this embodiment of the invention, S1 includes downsampling, pulse compression, and demodulation.

[0082] Downsampling is performed using radial sampling templates, random sampling templates, or Cartesian sampling templates.

[0083] In this embodiment of the invention, S2 includes the following sub-steps:

[0084] S21. Calculate the NFCS imaging operator based on the preprocessed scene echo data;

[0085] S22. Calculate the NFCS approximate observation operator based on the NFCS imaging operator.

[0086] In this embodiment of the invention, in S21, the improved inverse algorithm of Near Field Chirp Scaling (NFCS) is used as the approximate observation matrix. Near-field broadband correction is performed by converting the echo data to the two-dimensional frequency domain, enabling the Chirp Scaling algorithm to be applied in the near field. NFCS imaging operator The expression is:

[0087] ;

[0088] Where Y represents the scene echo data, For Hadamard products, For orientation to Fourier transform, This is the inverse Fourier transform of the orientation. For distance to Fourier transform, For the distance-to-Fourier transform and the distance-to-inverse-Fourier transform, For two-dimensional frequency domain Fourier transform, For the first phase function, For the second phase function, For the third phase function, It is the fourth phase function;

[0089] In S22, the NFCS approximate observation operator is obtained by performing the inverse operation of the imaging operator. The expression is:

[0090] ;

[0091] Among them, This is a two-dimensional frequency domain inverse Fourier transform. This is the conjugate transpose of the first phase function. This is the conjugate transpose of the second phase function. This is the conjugate transpose of the third phase function. This is the conjugate transpose of the fourth phase function. It is a two-dimensional matrix for scene imaging.

[0092] In this embodiment of the invention, S3 includes the following sub-steps:

[0093] S31. Construct a higher-order total variational penalty term;

[0094] S32. Construct an exponential penalty term;

[0095] S33. Construct a hybrid sparse near-field SAR imaging model based on the higher-order total variation penalty term and the exponential penalty term.

[0096] In this embodiment of the invention, in S31, the higher-order total variation penalty term The expression is:

[0097] ;

[0098] in, For gradient operators, For the rows of the matrix, These are the columns of the matrix;

[0099] In S32, the exponential penalty term The expression is:

[0100] ;

[0101] in, Here are the parameters for exponential regularization. For the current two-dimensional imaging scenario, For index;

[0102] In S33, the expression for constructing the hybrid sparse near-field SAR imaging model is:

[0103] ;

[0104] in, These are the weighting parameters used to control SAR time domain and TFXD. The first regularization parameter is used. This is the second regularization parameter. It is the short-time Fourier transform matrix. For approximate observation operators, This is a two-dimensional echo data vector representation;

[0105] The unconstrained problem above can be transformed into a constrained problem using the split variable method:

[0106] ;

[0107] in, As the first auxiliary variable, As the second auxiliary variable, the above equation can be transformed into an augmented Lagrange function using the augmented Lagrange multiplier method:

[0108] ;

[0109] in, The first regularization parameter is used. This is the second regularization parameter. As the first auxiliary variable, As the second auxiliary variable, As the first Lagrange multiplier, For the second Lagrange multiplier, For the first scaling augmented Lagrange multiplier, For the second scaling augmented Lagrange multiplier, These are the weighting parameters used to control SAR time domain and TFXD. It is the short-time Fourier transform matrix. This is an expression of the sensing matrix in vector form. This is a two-dimensional echo data vector representation. Exponential regularization with the first auxiliary variable as the parameter.

[0110] In this embodiment of the invention, in S4, the input echo data y is the initial solution. , , , , Imaging operator and approximate observation operator, error parameter The maximum number of iterations is The first Lagrange multiplier Second Lagrange multiplier Let the initial number of iterations be t=0.

[0111] In this embodiment of the invention, S5 includes the following sub-steps:

[0112] S51. Update the current two-dimensional imaging scene based on the hybrid sparse near-field SAR imaging model;

[0113] S52. Update the first auxiliary variable of the hybrid sparse near-field SAR imaging model according to the updated current two-dimensional imaging scene.

[0114] S53. Update the first Lagrange multiplier of the hybrid sparse near-field SAR imaging model according to the updated first auxiliary variable;

[0115] S54. Update the first scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model;

[0116] S55. Update the second auxiliary variable of the hybrid sparse near-field SAR imaging model according to the updated current two-dimensional imaging scene;

[0117] S56. Update the second Lagrange multiplier of the hybrid sparse near-field SAR imaging model according to the updated second auxiliary variable;

[0118] S57. Update the second scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model and complete the hybrid constrained sparse imaging of near-field SAR.

[0119] In this embodiment of the invention, in S51, the expression for updating the current two-dimensional imaging scene is:

[0120] ;

[0121] in, To perform the first step of the ADMM algorithm on g The result of the iteration. For the number of iterations, These are the weighting parameters used to control SAR time domain and TFXD. For approximate observation operators, For imaging operators, This is the short-time inverse Fourier transform. It is the short-time Fourier transform matrix. As the first Lagrange multiplier, For the second Lagrange multiplier, This is a two-dimensional echo data vector representation. Let be the first auxiliary variable for solving the problem in the t-th iteration. Let be the second auxiliary variable for solving the problem in the t-th iteration. For the first scaling augmented Lagrange multiplier obtained in the t-th iteration, The second scaling augmented Lagrange multiplier is obtained in the t-th iteration;

[0122] In S52, the expression for updating the first auxiliary variable of the hybrid sparse near-field SAR imaging model is:

[0123] ;

[0124] in, To use the ADMM algorithm for Conduct the first The result of the iteration. As the first auxiliary variable, The first regularization parameter is used. for The The result of the iteration. For the current two-dimensional imaging scenario, Exponential regularization with the first auxiliary variable as the parameter. This is a generalized iterative thresholding method. To use the ADMM algorithm for Conduct the first The result of the iteration;

[0125] In S53, the expression for updating the first Lagrange multiplier of the hybrid sparse near-field SAR imaging model is:

[0126] ;

[0127] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for the first The result of the iteration;

[0128] In S54, the expression for updating the first scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model is:

[0129] ;

[0130] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration;

[0131] In S55, the expression for updating the second auxiliary variable of the hybrid sparse near-field SAR imaging model is:

[0132] ;

[0133] in, To use the ADMM algorithm for Conduct the first The result of the iteration. As the second auxiliary variable, This is the second regularization parameter. The higher-order total variation penalty term is the one with the second auxiliary variable as a parameter. for Conduct the first The result of the iteration;

[0134] In S56, the expression for updating the second Lagrange multiplier of the hybrid sparse near-field SAR imaging model is:

[0135] ;

[0136] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for No. The result of the iteration;

[0137] In S57, the second scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model is updated:

[0138] ;

[0139] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration is obtained.

[0140] In this embodiment of the invention, S6 includes the following sub-steps:

[0141] S61. Calculate iterative parameters based on near-field SAR hybrid constrained sparse imaging;

[0142] S62. Determine whether the iteration parameters simultaneously satisfy the following conditions. and If so, return to S5 and continue updating the hybrid sparse near-field SAR imaging model; otherwise, output the imaging result. For iteration parameters, For error parameters, t represents the maximum number of iterations, and t is the initial value of the number of iteration steps.

[0143] In this embodiment of the invention, in S61, the iteration parameters The expression is:

[0144] ;

[0145] in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration is obtained.

[0146] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A sparse near-field SAR imaging method based on multi-constraint exponential high-order total variation, characterized in that, Includes the following steps: S1. Sample the scene and preprocess the scene echo data; S2. Calculate the NFCS approximate observation operator based on the preprocessed scene echo data; S3. Construct a hybrid sparse near-field SAR imaging model; S4. Initialize the parameters of the hybrid sparse near-field SAR imaging model and the NFCS approximate observation operator; S5. Based on the initialization results, update the hybrid sparse near-field SAR imaging model to complete the hybrid constrained sparse near-field SAR imaging. S6. Based on near-field SAR hybrid constrained sparse imaging, output the imaging results; In S1, the preprocessing includes downsampling, pulse compression, and demodulation. Among them, radial sampling template, random sampling template or Cartesian sampling template are used for downsampling; S2 includes the following sub-steps: S21. Calculate the NFCS imaging operator based on the preprocessed scene echo data; S22. Calculate the NFCS approximate observation operator based on the NFCS imaging operator; S3 includes the following sub-steps: S31. Construct a higher-order total variational penalty term; S32. Construct an exponential penalty term; S33. Construct a hybrid sparse near-field SAR imaging model based on the higher-order total variation penalty term and the exponential penalty term.

2. The sparse near-field SAR imaging method based on high-order total variation with multiple constraints as described in claim 1, characterized in that, In S21, the NFCS imaging operator The expression is: ； Where Y represents the scene echo data, For Hadamard products, For orientation to Fourier transform, This is the inverse Fourier transform of the orientation. For distance to Fourier transform, For the distance-to-Fourier transform and the distance-to-inverse-Fourier transform, For two-dimensional frequency domain Fourier transform, For the first phase function, For the second phase function, For the third phase function, It is the fourth phase function; In S22, the NFCS approximate observation operator The expression is: ； in, This is a two-dimensional frequency domain inverse Fourier transform. This is the conjugate transpose of the first phase function. This is the conjugate transpose of the second phase function. This is the conjugate transpose of the third phase function. This is the conjugate transpose of the fourth phase function. It is a two-dimensional matrix for scene imaging.

3. The sparse near-field SAR imaging method based on high-order total variation with multiple constraints as described in claim 1, characterized in that, In S31, the higher-order total variational penalty term The expression is: ； in, For gradient operators, For the rows of the matrix, These are the columns of the matrix; In S32, the exponential penalty term The expression is: ； in, Here are the parameters for exponential regularization. For the current two-dimensional imaging scenario, For index; In S33, the expression for constructing the hybrid sparse near-field SAR imaging model is as follows: ； in, These are the weighting parameters used to control SAR time domain and TFXD. The first regularization parameter is used. This is the second regularization parameter. It is the short-time Fourier transform matrix. For approximate observation operators, This is a representation of two-dimensional echo data vectors.

4. The sparse near-field SAR imaging method based on multi-constraint exponential high-order total variation as described in claim 1, characterized in that, S5 includes the following sub-steps: S51. Update the current two-dimensional imaging scene based on the hybrid sparse near-field SAR imaging model; S52. Update the first auxiliary variable of the hybrid sparse near-field SAR imaging model according to the updated current two-dimensional imaging scene. S53. Update the first Lagrange multiplier of the hybrid sparse near-field SAR imaging model according to the updated first auxiliary variable; S54. Update the first scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model; S55. Update the second auxiliary variable of the hybrid sparse near-field SAR imaging model according to the updated current two-dimensional imaging scene; S56. Update the second Lagrange multiplier of the hybrid sparse near-field SAR imaging model according to the updated second auxiliary variable; S57. Update the second scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model and complete the hybrid constrained sparse imaging of near-field SAR.

5. The sparse near-field SAR imaging method based on high-order total variation with multiple constraints as described in claim 4, characterized in that, In step S51, the expression for updating the current two-dimensional imaging scene is: ； in, To perform the first step of the ADMM algorithm on g The result of the iteration. For the number of iterations, These are the weighting parameters used to control SAR time domain and TFXD. For approximate observation operators, For imaging operators, This is the short-time inverse Fourier transform. It is the short-time Fourier transform matrix. As the first Lagrange multiplier, For the second Lagrange multiplier, This is a two-dimensional echo data vector representation. Let be the first auxiliary variable for solving the problem in the t-th iteration. Let be the second auxiliary variable for solving the problem in the t-th iteration. For the first scaling augmented Lagrange multiplier obtained in the t-th iteration, The second scaling augmented Lagrange multiplier is obtained in the t-th iteration; In step S52, the expression for updating the first auxiliary variable of the hybrid sparse near-field SAR imaging model is: ； in, To use the ADMM algorithm for Conduct the first The result of the iteration. As the first auxiliary variable, The first regularization parameter is used. for The The result of the iteration. For the current two-dimensional imaging scenario, Exponential regularization with the first auxiliary variable as the parameter. This is a generalized iterative thresholding method. To use the ADMM algorithm for Conduct the first The result of the iteration; In S53, the expression for updating the first Lagrange multiplier of the hybrid sparse near-field SAR imaging model is: ； in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for the first The result of the iteration; In S54, the expression for updating the first scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model is: ； in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration; In S55, the expression for updating the second auxiliary variable of the hybrid sparse near-field SAR imaging model is: ； in, To use the ADMM algorithm for Conduct the first The result of the iteration. As the second auxiliary variable, This is the second regularization parameter. The higher-order total variation penalty term is the one with the second auxiliary variable as a parameter. for Conduct the first The result of the iteration; In S56, the expression for updating the second Lagrange multiplier of the hybrid sparse near-field SAR imaging model is: ； in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for No. The result of the iteration; In S57, the second scaling augmented Lagrange multiplier of the hybrid sparse near-field SAR imaging model is updated: ； in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration is obtained.

6. The sparse near-field SAR imaging method based on multi-constraint exponential high-order total variation as described in claim 1, characterized in that, S6 includes the following sub-steps: S61. Calculate iterative parameters based on near-field SAR hybrid constrained sparse imaging; S62. Determine whether the iteration parameters simultaneously satisfy the following conditions. and If so, return to S5 and continue updating the hybrid sparse near-field SAR imaging model; otherwise, output the imaging result. For iteration parameters, For error parameters, To determine the maximum number of iterations, This is the initial value for the number of iteration steps.

7. The sparse near-field SAR imaging method based on multi-constraint exponential high-order total variation as described in claim 6, characterized in that, In S61, the iteration parameters The expression is: ； in, To use the ADMM algorithm for Conduct the first The result of the iteration. To use the ADMM algorithm for Conduct the first The result of the iteration is obtained.

Images