Separation sparse SAR imaging method and device based on echo simulation and medium

Abstract

The invention discloses a separation sparse SAR imaging method and device based on echo simulation and a medium, and the method comprises the steps: employing different norms to respectively restrain a non-sparse target and a sparse target, and constructing a separation imaging mathematical model; designing an optimization solution algorithm based on the separation imaging mathematical model; according to the matched filtering operator, an echo simulation operator in the algorithm is designed; and respectively obtaining imaging results of a non-sparse target and a sparse target in the scene through iteration. According to the separation sparse SAR imaging method based on echo simulation, the difference between a non-sparse target and a sparse target in a scene is fully considered, the sparse characteristic of the target is grasped, sparse constraint is carried out on the non-sparse target and the sparse target in the scene, the image of the target is obtained through algorithm iteration, and the target resolution is improved. Therefore, the separation imaging frame provided by the invention can process echo data containing multiple types of targets and effectively realize separation imaging of the targets, and technical support is provided for processing of subsequent target detection and the like.

Description

Technical Field

[0001] The invention belongs to the field of sparse signal processing and microwave imaging, and in particular relates to a separation sparse SAR imaging method, device and medium based on echo simulation. Background Art

[0002] Synthetic Aperture Radar (SAR) is a radar that achieves synthetic aperture by moving the radar, obtains a narrow beam width, and thus improves the radar's resolution in azimuth. It has been widely used in various fields. Sparse SAR imaging technology, which is developed by combining sparse signal processing technology with SAR imaging, has great research value. Sparse SAR imaging technology can improve the target signal-to-noise ratio and image quality. In addition, sparse SAR imaging technology can effectively reduce the pressure of data transmission and storage, and reduce the cost and resource consumption of the system. At the same time, sparse imaging technology also has higher sensitivity and accuracy in target detection and recognition, which helps to analyze and monitor targets in complex scenes. Therefore, sparse SAR imaging technology has broad application prospects in military reconnaissance, disaster monitoring, resource exploration and other fields.

[0003] The sparse imaging method based on the observation matrix cannot achieve sparse reconstruction of large-scale scenes due to the large amount of observation matrix calculation, while the sparse imaging method based on the echo simulation operator can achieve approximate reconstruction of large-scale scenes. The sparse SAR imaging method based on regularization achieves sparse recovery of the observation scene by solving the LASSO problem. Usually, the modeling process based on the LASSO problem only characterizes the sparsity of a single target and cannot achieve separate imaging of different targets. Compared with the traditional single target modeling method, the sparse separation imaging model also introduces the Frobenius norm constraint for non-sparse targets, which can achieve separate imaging of non-sparse targets and sparse targets. The present invention will carry out separate imaging of different targets based on sparse signal processing technology, and propose an algorithm based on the separation sparse SAR imaging model. The algorithm is based on the echo simulation operator, and can achieve accurate recovery and reconstruction for both sparse and non-sparse targets. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to propose a separation sparse SAR imaging method, device and medium based on echo simulation to achieve imaging of non-sparse targets and sparse targets.

[0005] Technical solution: The present invention discloses a separation sparse SAR imaging method based on echo simulation, comprising the following steps:

[0006] (1) Based on the sparsity of different targets, different norms are used as regularization terms to constrain non-sparse targets and sparse targets, and a two-dimensional separated sparse SAR imaging model is constructed;

[0007] (2) The iterative soft threshold algorithm (IST) is used to find the optimal solution to the minimization problem of the two-dimensional separation sparse SAR imaging model using an alternating iteration strategy;

[0008] (3) Design the corresponding echo simulation operator according to the characteristics of the matched filter operator;

[0009] (4) The imaging results of non-sparse targets and sparse targets in the scene are obtained through iteration.

[0010] Furthermore, the implementation process of step (1) is as follows:

[0011] The two-dimensional separated sparse SAR imaging model contains different targets and is expressed as:

[0012]

[0013] in, Represents the received echo data containing different targets with a size of M×N, represents the non-sparse targets in the observation scene, represents the sparse targets in the observation scene, and Respectively represent X C and X T The echo simulation operator, and its corresponding matched filter imaging operator is and N represents the noise matrix;

[0014] Based on the reconstruction error, the Frobenius norm and l 1，1 The norm is used as a regularization term to constrain the background image and the target signal image. The target is obtained by solving the following regularization problem:

[0015]

[0016] Among them, ||·|| F is the Frobenius norm, ||·|| 1，1 is the joint matrix norm, which is defined as:

[0017]

[0018] Among them, X T (i, j) represents the matrix X T The element in the i-th row and j-th column of .

[0019] Furthermore, the implementation process of step (2) is as follows:

[0020] Use the optimization algorithm to iteratively update the non-sparse target X in the observation scene alternately C and the sparse target X in the observation scene T Solve the problem; where update X C When using the gradient descent method, update X T Iterative soft thresholding algorithm (IST) is used to process l 1,1 Norm; the algorithm inputs echo data Y and operator and Suppose the non-sparse target of sparse reconstruction and sparse targets is initialized to 1, the iteration step is u, the error parameter is ε, and the number of iterations k is set to 1;

[0021] When the residual ρ k >ε, calculate the non-sparse target estimate:

[0022]

[0023] Compute a sparse target estimate:

[0024]

[0025] Among them, prox μ (·) is the threshold operator of IST, and the residual is obtained by the following expression:

[0026]

[0027] Among them, J(X C , X T ) is the minimization problem to be solved, and finally the reconstructed non-sparse and sparse targets

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

[0029] According to the characteristics of the matched filter imaging operator, the corresponding echo simulation operator is designed, and the back-projection algorithm is selected as the matched filter imaging operator, which is expressed as:

[0030]

[0031] Among them, C r (·) represents the range pulse compression operation, Ψ represents the phase to be compensated, represents the phase compensation operation, S a (·) represents the azimuthal coherent accumulation operation, represents the Hadamard product;

[0032] The inverse imaging operator, that is, the echo simulation operator, is derived from the matched filter imaging operator based on the back-projection algorithm and is expressed as:

[0033]

[0034] in,(·) -1 represents the corresponding inverse operation, (·) * Represents the conjugate transpose operation of a matrix.

[0035] Furthermore, the implementation process of step (4) is as follows:

[0036] The imaging results of non-sparse targets and sparse targets in the scene are obtained separately through iteration; the echo data containing different targets are taken as input, and the separated imaging of non-sparse targets and sparse targets is realized by constructing a separated sparse SAR imaging model based on echo simulation.

[0037] The storage medium of the present invention stores a computer program, and when the computer program is executed by at least one processor, the steps of the above-mentioned separation sparse SAR imaging method based on echo simulation are implemented.

[0038] A device according to the present invention includes a memory and a processor, wherein:

[0039] A memory for storing computer programs that can be run on the processor;

[0040] The processor is used to execute the steps of the above-mentioned separation sparse SAR imaging method based on echo simulation when running the computer program.

[0041] Beneficial effects: Compared with the prior art, the beneficial effects of the present invention are: 1. Compared with the traditional sparse imaging method based on the observation matrix, the present invention is based on the echo simulation operator operation, which reduces the amount of calculation and realizes sparse imaging of a large range of observation scenes; 2. Compared with the existing sparse SAR imaging algorithm, the present invention realizes separate imaging of different targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flow chart of the separation sparse SAR imaging method based on echo simulation;

[0043] Figure 2 The target detection results of the present invention applied to simulated data; wherein (a) is the imaging result of a non-sparse target; and (b) is the imaging result of a sparse target. DETAILED DESCRIPTION

[0044] The present invention will be further described below in conjunction with the accompanying drawings.

[0045] like Figure 1 As shown, the present invention proposes a separation sparse SAR imaging model and method based on echo simulation, comprising the following steps:

[0046] Step 1: Establish a mathematical model for separation imaging based on target sparsity: Based on the sparsity of different targets, different norms are used as regularization terms to constrain non-sparse targets and sparse targets, and a mathematical model for separation imaging is constructed.

[0047] The two-dimensional separated sparse SAR imaging model containing different targets can be expressed as:

[0048]

[0049] in, Represents the received echo data containing different targets with a size of M×N, represents the non-sparse targets in the observation scene, represents the sparse targets in the observation scene, and Respectively represent X C and X T The echo simulation operator, and its corresponding matched filter imaging operator can be expressed as and N represents the noise matrix.

[0050] Based on the reconstruction error, the Frobenius norm and l 1，1 The norm is used as a regularization term to constrain the background image and the target signal image. The target can be obtained by solving the following regularization problem:

[0051]

[0052] Among them, ||·|| F is the Frobenius norm, ||·|| 1，1 is the joint matrix norm, which is defined as:

[0053]

[0054] Among them, X T (i, j) represents the matrix X T The element in the i-th row and j-th column of .

[0055] Step 2: Design an optimized algorithm for solving the separation imaging model: Adopt the iterative shrinkage-threshold (IST) algorithm framework and use an alternating iteration strategy to find the optimal solution to the separation model minimization problem.

[0056] According to the separation optimization model obtained in step 1, use the optimization algorithm to iteratively update X alternately. C With X T Solve the problem. In which, update X C When using the gradient descent method, update X T Iterative soft thresholding algorithm is used to process l 1，1 The algorithm inputs the echo data Y and the operator and Suppose the non-sparse target of sparse reconstruction and sparse targets is initialized to 1, the iteration step is u, the error parameter is ε, and the number of iterations k is set to 1.

[0057] When the residual ρ k >ε, perform the following steps:

[0058] 1) Calculate the non-sparse target estimate:

[0059]

[0060] 2) Calculate the sparse target estimate:

[0061]

[0062] Among them, prox μ (·) is the threshold operator of IST, and the residual can be obtained by the following expression:

[0063]

[0064] Among them, J(X C , X T ) is the minimization problem we are trying to solve, and finally outputs the reconstructed non-sparse and sparse targets

[0065] Step 3: Design of echo simulation operator in the algorithm.

[0066] According to the characteristics of the matched filter imaging operator, the corresponding echo simulation operator is designed, and the back-projection algorithm is selected as the matched filter imaging operator, which can be expressed as:

[0067]

[0068] Among them, C r (·) represents the range pulse compression operation, Ψ represents the phase to be compensated, represents the phase compensation operation, S a (·) represents the azimuthal coherent accumulation operation, Denotes the Hadamard product.

[0069] The inverse imaging operator, that is, the echo simulation operator, can be derived from the matched filter imaging operator based on the back-projection algorithm and can be expressed as:

[0070]

[0071] in,(·) -1 represents the corresponding inverse operation, (·) * Represents the conjugate transpose operation of a matrix.

[0072] Step 4: Target separation imaging.

[0073] The imaging results of non-sparse targets and sparse targets in the scene are obtained through iteration. The echo data containing different targets are taken as input, and the separation imaging of non-sparse targets and sparse targets is realized through the constructed separation sparse SAR imaging model based on echo simulation.

[0074] The present invention also provides a storage medium having a computer program stored thereon. When the computer program is executed by at least one processor, the steps of the above-mentioned separation sparse SAR imaging method based on echo simulation are implemented.

[0075] 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 above-mentioned separation sparse SAR imaging method based on echo simulation when running the computer program.

[0076] In order to verify the effectiveness of the method of the present invention, a separation sparse SAR imaging model and method based on echo simulation are simulated and verified according to the parameters shown in Table 1.

[0077] Table 1 Simulated spaceborne SAR parameters

[0078]

[0079] The scene is set to contain a non-sparse target of 1000×1000 size with random amplitude (in the range of [0.01, 0.1]) and random phase (in the range of [-π, π]) and a sparse single point target with amplitude of 1 and phase of 0, and the echo data of the scene are separated and imaged. Figure 2 is the target detection result of the present invention applied to the simulated data, Figure 2 (a) Imaging results of non-sparse targets; Figure 2 (b) The imaging result of sparse targets. Figure 2 The results show that both the non-sparse targets and the sparse targets in the scene are well restored. Therefore, the method proposed in the present invention can achieve the separation imaging of the non-sparse targets and the sparse targets in the scene.

[0080] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A separation sparse SAR imaging method based on echo simulation, characterized in that: The following steps are involved: (1) Based on the sparsity of different targets, different norms are used as regularization terms to constrain non-sparse targets and sparse targets, and a two-dimensional separated sparse SAR imaging model is constructed; (2) The iterative soft threshold algorithm (IST) is used to find the optimal solution to the minimization problem of the two-dimensional separation sparse SAR imaging model using an alternating iteration strategy; (3) Design the corresponding echo simulation operator according to the characteristics of the matched filter operator; (4) The imaging results of non-sparse targets and sparse targets in the scene are obtained through iteration.

2. The method for separated sparse SAR imaging based on echo simulation according to claim 1, characterized in that: The implementation process of step (1) is as follows: The two-dimensional separated sparse SAR imaging model contains different targets and is expressed as: in, Represents the received echo data containing different targets with a size of M×N, represents the non-sparse targets in the observation scene, represents the sparse targets in the observation scene, and Respectively represent X C and X T The echo simulation operator, and its corresponding matched filter imaging operator is and N represents the noise matrix; Based on the reconstruction error, the Frobenius norm and l 1，1 The norm is used as a regularization term to constrain the background image and the target signal image. The target is obtained by solving the following regularization problem: Among them, ||·|| F is the Frobenius norm, ||·|| 1，1 is the joint matrix norm, which is defined as: Among them, X T (i, j) represents the matrix X T The element in the i-th row and j-th column of .

3. The method for separated sparse SAR imaging based on echo simulation according to claim 1, characterized in that: The implementation process of step (2) is as follows: Use the optimization algorithm to iteratively update the non-sparse target X in the observation scene alternately C and the sparse target X in the observation scene T Solve the problem; where update X C When using the gradient descent method, update X T Iterative soft thresholding algorithm (IST) is used to process l 1，1 Norm; the algorithm inputs echo data Y and operator and Suppose the non-sparse target of sparse reconstruction and sparse targets is initialized to 1, the iteration step is u, the error parameter is ε, and the number of iterations k is set to 1; When the residual ρ k >ε, calculate the non-sparse target estimate: Compute a sparse target estimate: Among them, prox μ (·) is the threshold operator of IST, and the residual is obtained by the following expression: Among them, J(X C , X T ) is the minimization problem to be solved, and finally the reconstructed non-sparse and sparse targets 4. The method for separated sparse SAR imaging based on echo simulation according to claim 1, characterized in that: The implementation process of step (3) is as follows: According to the characteristics of the matched filter imaging operator, the corresponding echo simulation operator is designed, and the back-projection algorithm is selected as the matched filter imaging operator, which is expressed as: Among them, C r (·) represents the range pulse compression operation, Ψ represents the phase to be compensated, represents the phase compensation operation, S a (·) represents the azimuthal coherent accumulation operation, and ο represents the Hadamard product; The inverse imaging operator, that is, the echo simulation operator, is derived from the matched filter imaging operator based on the back-projection algorithm and is expressed as: in,(·) -1 represents the corresponding inverse operation, (·) * Represents the conjugate transpose operation of a matrix.

5. The method for separated sparse SAR imaging based on echo simulation according to claim 1, characterized in that: The implementation process of step (4) is as follows: The imaging results of non-sparse targets and sparse targets in the scene are obtained separately through iteration; the echo data containing different targets are taken as input, and the separated imaging of non-sparse targets and sparse targets is realized by constructing a separated sparse SAR imaging model based on echo simulation.

6. 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 separated sparse SAR imaging method based on echo simulation as claimed in any one of claims 1 to 5 are implemented.

7. An electronic 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 is used to execute the steps of a separation sparse SAR imaging method based on echo simulation as described in any one of claims 1 to 5 when running the computer program.

Images