Sparse SAR Joint Imaging Method, Device and Medium Based on Multi-Band Data

Abstract

The present invention discloses a sparse SAR joint imaging method, device and medium based on multi-band data. The method uses a sparse model combining the norm and the norm to constrain the multi-band data, and constructs a multi-band sparse SAR joint imaging model; solves the multi-band sparse SAR joint imaging model to obtain the optimal solution of the minimization problem of the model; designs a corresponding echo simulation operator according to the characteristics of the matched filtering operator to realize the iteration of the algorithm; and obtains the joint imaging result of the multi-band data of the scene through iteration. The present invention takes into account the characteristic that the scene support sets are consistent under different bands, grasps the sparse characteristics of the target, performs sparse constraint on the scene, reduces the sidelobes, effectively obtains the image of the target through algorithm iteration, and provides technical support for subsequent image processing.

Description

Technical Field

[0001] The present invention belongs to the fields of sparse signal processing and microwave imaging, and particularly relates to a sparse SAR joint imaging method, device, and medium based on multi-band data. Background Art

[0002] Synthetic Aperture Radar (SAR) improves the azimuth image resolution through the synthetic aperture in the azimuth direction, and has higher imaging quality than traditional radar imaging. It has now been widely used in many fields. By combining sparse signal processing technology and applying the principle of compressive sensing to the SAR processing field, it is possible to further improve the imaging quality on the basis of traditional SAR imaging methods. Sparse SAR imaging technology has now achieved down-sampled data processing, reduced image calculation volume, focused imaging of moving targets, and de-blurred imaging in multiple fields, and has great application potential.

[0003] Emission signals in different bands can obtain different scattering information of the observed scene. Common bands include the L band, C band, X band, Ku band, P band, etc. The L band has a longer wavelength and stronger penetration ability, can penetrate dense vegetation, and obtain surface information, and is commonly used for biomass measurement and forest monitoring. The C band and X band are widely used in surface monitoring under various environmental and weather conditions, can provide clear images, and are suitable for urban planning, topographic mapping, and natural resource management. The Ku band is usually used in airborne systems at altitudes above clouds and rain, has higher resolution and penetration ability, and is suitable for high-precision imaging in complex environments. The P-band SAR system performs well in forest monitoring, can penetrate deep into the tree canopy to obtain accurate information, and is suitable for forest biomass measurement and the generation of height maps. Applying sparse signal processing technology to the joint processing of multi-band data and exploring the information correlation in multi-band data are expected to use the data of multiple bands to improve the observation ability of the scene and provide an information basis for subsequent image processing. Summary of the Invention

[0004] Object of the Invention: The object of the present invention is to propose a sparse SAR joint imaging method, device, and medium based on multi-band data, and effectively realize multi-band joint imaging of the scene.

[0005] Technical Solution: The sparse SAR joint imaging method based on multi-band data described in the present invention specifically includes the following steps:

[0006] (1) Construct a sparse SAR joint imaging model for multi-band data;

[0007] (2) Based on the consistency of the scene support sets of multi-band data, use the norm to constrain the common part of the image data in multiple bands; based on the scene sparsity, use The norms are used to constrain the multi - band image data respectively;

[0008] (3) Solve the sparse SAR joint imaging model of multi - band data to obtain the optimal solution of the minimization problem of this model;

[0009] (4) Design the corresponding echo simulation operator according to the characteristics of the matched filtering operator;

[0010] (5) Obtain the joint imaging result of multi - band data of the scene through iteration.

[0011] Further, the sparse SAR joint imaging model of the multi - band data in step (1) is:

[0012] ;

[0013] Where, respectively represent the echo data received by non - connected bands, respectively represent the back - scattering coefficients of the scene under the corresponding band observations, respectively represent the corresponding echo simulation operators,

[0014] Further, the implementation process of step (2) is as follows:

[0015] Based on the consistency of the scene support sets of the echo data of different bands, the group sparsity of the common part in the image is constrained; for the sparse characteristics of the scene, the images of the data of the first band and the second band are respectively sparsely constrained; the scene is obtained by solving the following regularization problem:

[0016] ;

[0017] Where, is the regularization parameter, is norm, is

[0018] the joint matrix norm of

[0019] ;

[0020] Where, respectively represent the th row and th column elements of the matrix and

[0021] Furthermore, the implementation process of step (3) is as follows:

[0022] According to the sparse SAR joint imaging model obtained in step (1), use the structural sparse model solving algorithm to update the solution to the problem, and the final scene restoration result is ; the algorithm inputs the echo data and the operator , ; respectively represent the echo simulation operators in the 1st to the th bands ; are respectively matched filtering imaging operators; let the initial value of the sparse reconstruction image be initialized to 1, the sparsity , the iteration step size , the error parameter , and the initial value of the iteration number ; when the residual , execute steps S1 to S7; until , end the iteration;

[0023] S1: Calculate the residual:

[0024] ;

[0025] S2: Calculate the image update value:

[0026] ;

[0027] S3: Calculate the regularization parameter:

[0028] ;

[0029] S4: Apply sparse constraints to the image update value:

[0030] ;

[0031] S5: Calculate the common part:

[0032] ;

[0033] S6: Calculate the regularization parameter of the common part:

[0034] ;

[0035] S7: Apply group sparse constraints to the image update value:

[0036] ;

[0037] Finally, output the imaging result of the scene .

[0038] Furthermore, the implementation process of step (4) is as follows:

[0039] According to the characteristics of the matched filtering imaging operator, design the corresponding echo simulation operator, and select the back projection algorithm as the matched filtering imaging operator, which is expressed as:

[0040] ;

[0041] Among them, represents the range compression operation of the th band, represents the phase to be compensated for the th band, represents the phase compensation operation for the th band, represents the azimuth coherent accumulation operation for the th band, represents the Hadamard product;

[0042] Derive the inverse imaging operator from the matched filtering imaging operator of the th band based on the back projection algorithm, that is, the echo simulation operator of the th band, which is expressed as:

[0043] ;

[0044] Among them, represents the corresponding inverse operation, represents the conjugate transpose operation of the matrix.

[0045] Furthermore, the implementation process of step (5) is as follows:

[0046] Obtain the multi-band joint imaging result of the scene through iteration; use the multi-band echo data as the input, and through the constructed sparse SAR joint imaging model based on echo simulation, realize the multi-band joint imaging of the scene.

[0047] A storage medium according to the present invention, on which a computer program is stored, and when the computer program is executed by at least one processor, the steps of the sparse SAR joint imaging method based on multi-band data as described above are implemented.

[0048] An electronic device according to the present invention, including a memory and a processor, wherein:

[0049] The memory is used to store a computer program that can run on the processor;

[0050] A processor, configured to execute the steps of the sparse SAR joint imaging method based on multi-band data as described above when running the computer program.

[0051] Beneficial effects: Compared with the prior art, the beneficial effects of the present invention are as follows: The sparse SAR joint imaging method based on the norm and the norm takes into account the characteristic that the scene support sets are consistent under different bands, grasps the sparse characteristics of the target, performs sparse constraints on the scene, reduces the sidelobes, and effectively realizes the imaging of the scene through algorithm iteration; the present invention can reduce the data volume and reduce the requirements for the system load; combines and processes the echo data of multiple bands, utilizes the information in the multi-band data, realizes sparse imaging of the scene, and improves the image quality. Description of the Drawings

[0052] Figure 1 is a flowchart of the sparse SAR joint imaging method based on multi-band data;

[0053] Figure 2 is the target detection result of the present invention applied to simulated data; among them, (a) is the imaging result of the back-projection algorithm for X-band data; (b) is the imaging result of the back-projection algorithm for K-band data; (c) is the multi-band joint imaging result of the scene. Detailed Embodiments

[0054] The present invention will be further described in detail below with reference to the drawings.

[0055] As Figure 1 shown, the present invention provides a sparse SAR joint imaging method based on multi-band data, including the following steps:

[0056] Step 1: Construct a sparse SAR joint imaging model for multi-band data:

[0057] ;

[0058] Among them, respectively represent the echo data received by non-connected bands, respectively represent the backscattering coefficients of the scenes observed under their corresponding bands, respectively represent , represents the noise matrix.

[0059] Based on the consistency of the scene support sets of echo data in different bands, impose a group sparsity constraint on the common part in the image; for the sparse characteristics of the scene, respectively impose sparse constraints on the images of the data in the first band and the second band; the scene can be obtained by solving the following regularization problem:

[0060] ;

[0061] where, is the regularization parameter, is the norm, is

[0062] the joint matrix norm of, and its definition is:

[0063] ;

[0064] where, respectively represent the elements of the matrix at the -th row and the -th column, and respectively represent the sizes of the rows and columns of the matrix.

[0065] Step 2: Based on the algorithm framework for solving the structured sparse model, solve the sparse SAR joint imaging model for multi-band data to obtain the optimal solution to the minimization problem of this model.

[0066] According to the sparse SAR joint imaging model obtained in Step 1, use the structured sparse model solving algorithm to update the solution to the problem, and the final scene recovery result is ; the algorithm inputs the echo data and the operator , ; respectively represent the echo simulation operators of under the 1st to the -th bands; are respectively the matched filtering imaging operators of ; let the initial value of the sparse reconstruction image be initialized to 1, the sparsity , the iteration step size , the error parameter , and the initial value of the number of iterations .

[0067] The algorithm process is as follows. When the residual , execute the following steps; until , end the iteration:

[0068] 1) Calculate the residual:

[0069] ;

[0070] 2) Calculate the image update value:

[0071] ;

[0072] 3) Calculate the regularization parameter:

[0073] ;

[0074] 4) Apply sparse constraint to the image update value:

[0075] ;

[0076] 5) Calculate the common part:

[0077] ;

[0078] 6) Calculate the regularization parameter of the common part:

[0079] ;

[0080] 7) Apply group sparse constraint to the image update value:

[0081] ;

[0082] Finally, output the scene imaging result .

[0083] Step 3: Construct operator iteration implementation.

[0084] According to the characteristics of the matched filtering imaging operator, design the corresponding echo simulation operator, and select the back-projection algorithm as the matched filtering imaging operator, expressed as:

[0085] ;

[0086] Among them, represents the range-direction pulse compression operation corresponding to the th band, represents the phase to be compensated corresponding to the th band, represents the phase compensation operation corresponding to the th band, represents the azimuth-direction coherent integration operation corresponding to the th band, represents the Hadamard product.

[0087] The corresponding to the back-projection algorithmDerive the inverse imaging operator of the matched filtering imaging operator for each band, that is, the echo simulation operator corresponding to the th band, expressed as:

[0088] ;

[0089] where represents the corresponding inverse operation, represents the conjugate transpose operation of the matrix.

[0090] Step 4: Multi-band data joint imaging.

[0091] Obtain the multi-band joint imaging result of the scene through iteration. Take the multi-band echo data as the input, and through the constructed sparse SAR joint imaging model based on echo simulation, realize the multi-band joint imaging of the scene and improve the image quality.

[0092] The present invention also provides a storage medium, on which a computer program is stored. When the computer program is executed by at least one processor, the steps of the sparse SAR joint imaging method based on multi-band data as described above are implemented.

[0093] The present invention also provides an electronic device, including a memory and a processor, where: the memory is used to store a computer program that can run on the processor; the processor is used to execute the steps of the sparse SAR joint imaging method based on multi-band data as described above when running the computer program.

[0094] To verify the effectiveness of the method of the present invention, according to the parameters shown in Table 1, take , simulate the echo data of the X-band and K-band, and conduct target simulation verification on the present invention.

[0095] Table 1 Simulation parameters

[0096] Parameter Value X-band operating frequency 9.5 GHz K-band operating frequency 20.3 GHz Azimuth antenna length 0.1 m Signal bandwidth 4.4 GHz Transmitted pulse width 10 ns Equivalent radar velocity 250 m / s Slant range to scene center 500 m

[0097] Figure 2 is the point target detection result of the present invention applied to simulated data, and the echo data is randomly downsampled by 50%. Figure 2 In (a) is the imaging result of the back projection algorithm for X-band data; Figure 2 In (b) is the imaging result of the back projection algorithm for K-band data; Figure 2 In (c) is the multi-band joint imaging result of the scene. From Figure 2 The results given can be seen that the multi-band joint imaging method can reduce the influence of energy dispersion caused by data downsampling on the image and improve the imaging quality of the scene. Therefore, the method proposed by the present invention can realize the sparse SAR joint imaging of multi-band data of the scene.

[0098] The specific embodiments described herein are merely illustrative of the spirit of the present invention. Those skilled in the art to which the present invention pertains may make various modifications or supplements to the described specific embodiments or use similar means for substitution, but will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A sparse SAR joint imaging method based on multi-band data, characterized in that, It includes the following steps: (1) Construct a sparse SAR joint imaging model for multi-band data; (2) Based on the consistency of the scene support set for multi-band data, use the L 2，1 norm to constrain the common part of the image data in multiple bands; Based on the scene sparsity, use the L1 norm to respectively constrain the image data of multiple bands; (3) Solve the sparse SAR joint imaging model for multi-band data to obtain the optimal solution to the minimization problem of this model; (4) Design a corresponding echo simulation operator according to the characteristics of the matched filtering operator; (5) Obtain the joint imaging result of multi-band data of the scene through iteration.

2. The sparse SAR joint imaging method based on multi-band data according to claim 1, wherein The sparse SAR joint imaging model for multi-band data in step (1) is: Among them, Y1, Y2, …, Y n respectively represent the echo data received by n non - connected bands, and X1, X2, …, X n respectively represent the backscattering coefficients of the scenes observed in their corresponding bands, respectively represent X1, X2, …, X n corresponding echo simulation operators, and N represents the noise matrix.

3. The sparse SAR joint imaging method based on multi-band data according to claim 2, wherein The implementation process of step (2) is as follows: Based on the consistency of the scene support sets of echo data in different bands, perform group sparsity constraint on the common part in the image; for the sparse characteristics of the scene, respectively perform sparse constraint on the images of the data of the first band and the second band; the scene is obtained by solving the following regularization problem: where λ1, λ2, …, λ n+1 are regularization parameters, ||·||1 is the L1 norm, and ||X com || 2，1 is the joint matrix norm of X1, X2, …, X n and is defined as: Among them, X1(i, j), X2(i, j), …, X n (i, j) respectively represent the elements in the i-th row and j-th column of matrices X1, X2, …, X n , and M and N respectively represent the sizes of the rows and columns of the matrix.

4. The sparse SAR joint imaging method based on multi-band data according to claim 2, characterized in that The implementation process of step (3) is as follows: Based on the sparse SAR joint imaging model obtained in step (1), use the structural sparse model solving algorithm to update X1, X2, …, X n Solve the problem, and the final scene restoration result is X = X1 + X2 + … + X n + X com ; The algorithm inputs the echo data Y1, Y2, … Y n and the operators respectively represent the echo simulation operators of X1, X2, …, X n under the 1st to the nth bands; are the matched filtering imaging operators of X1, X2, …, X n respectively; Let the initial value of the sparse reconstruction image be initialized to 1, the sparsity degrees K1, K2, …, K n+1 , the iteration step size μ, the error parameter ε, and the initial value of the iteration number k = 1; When the residual ρ k > ε, execute steps S1 to S7; Until ρ k > ε, end the iteration; S1: Calculate the residual; S2: Calculate the image update value; S3: Calculate the regularization parameter; … S4: Perform sparse constraint on the image update value; … S5: Calculate the common part; S6: Calculate the regularization parameter of the common part; S7: Perform group sparse constraint on the image update value; Finally output the scene imaging result 5. The sparse SAR joint imaging method based on multi-band data according to claim 1, characterized in that The implementation process of step (4) is as follows: According to the characteristics of the matched filtering imaging operator, design a corresponding echo simulation operator, and select the back-projection algorithm as the matched filtering imaging operator, which is expressed as: Among them, C ri (·) represents the range - direction pulse compression operation corresponding to the i - th band, Ψ i represents the phase to be compensated corresponding to the - th band, represents the phase compensation operation corresponding to the i - th band, S ai (·) represents the azimuth - direction coherent integration operation corresponding to the i - th band, represents the Hadamard product; Derive its inverse imaging operator from the matched filtering imaging operator corresponding to the i-th band based on the back-projection algorithm, that is, the echo simulation operator corresponding to the i-th band, which is expressed as: Among them, (·) -1 represents the corresponding inverse operation, and (·) * represents the conjugate transpose operation of the matrix.

6. The sparse SAR joint imaging method based on multi-band data according to claim 1, wherein The implementation process of step (5) is as follows: Obtain the joint imaging result of multi-band of the scene through iteration; use the multi-band echo data as the input, and through the constructed sparse SAR joint imaging model based on echo simulation, realize the joint imaging of multi-bands of the scene.

7. A storage medium, characterized in that, A computer program is stored on the storage medium, and when the computer program is executed by at least one processor, it realizes the steps of the sparse SAR joint imaging method based on multi-band data according to any one of claims 1 to 6.

8. An electronic device, characterized in that, It includes a memory and a processor, wherein: The memory is used to store a computer program that can run on the processor; The processor is used to execute the steps of the sparse SAR joint imaging method based on multi-band data according to any one of claims 1 to 6 when running the computer program.

Images