Multi-angle SAR imaging method and device based on unmanned aerial vehicle group

Abstract

The invention provides a multi-angle SAR imaging method and device based on an unmanned aerial vehicle group, and relates to the technical field of radar imaging. The method comprises the following steps: acquiring a target image of each unmanned aerial vehicle SAR about a target by using an SAR system based on an unmanned aerial vehicle group, and fusing each target image to obtain a fused image; wherein each unmanned aerial vehicle SAR detects a target from different angles, and each unmanned aerial vehicle SAR is in a self-transmitting and self-receiving mode; and for any unmanned aerial vehicle SAR, based on an alternating direction multiplier optimization algorithm and the fused image, performing iterative optimization on the target image of the unmanned aerial vehicle SAR to obtain an optimized target image of the unmanned aerial vehicle SAR, and taking the optimized target image as a multi-angle radar imaging graph of the target. According to the method, on the basis of quickly and flexibly obtaining SAR radar imaging, the influence of side lobes and noise can be reduced by using multi-angle target characteristics, overlay and shadow of the SAR image are weakened, and high-resolution SAR radar imaging is realized.

Description

Technical Field

[0001] The present invention relates to the field of radar imaging technology, and in particular to a multi-angle SAR imaging method and device based on a swarm of unmanned aerial vehicles. Background Art

[0002] Synthetic aperture radar (SAR), with its all-day, all-weather, high-resolution observation capabilities, has seen significant development and application in both military and civilian fields. Unmanned aerial vehicles (UAVs) are widely used as carrier platforms for SAR systems due to their low cost, flexible deployment, ease of operation, and low detection potential. However, single-platform UAV-mounted SAR systems, particularly those on small UAVs, are limited by the size, speed, and payload of the carrier platform, resulting in a relatively fixed geometric configuration and a limited synthetic aperture length. This can lead to a weak Doppler effect and compromised survivability and long-term comprehensive observation capabilities in the event of attack, malfunction, or in adverse weather or complex terrain.

[0003] In recent years, with the development of drone swarm technology, swarm SAR (UAV SAR) using swarm drones as payload platforms has become a new direction. Swarm-based SAR systems are deployed in mission environments using random dispersal or specific topological configuration strategies, forming distributed SAR systems. Through networked processing, collaboration between multiple drone SAR systems can accomplish complex and diverse imaging tasks. Compared to traditional single-platform drones, swarm-based SAR offers significant advantages in intelligence, survivability, and multi-viewpoint, high-resolution observation.

[0004] Multi-angle SAR (Multi-angle SAR) is an imaging mode that detects the same target or scene from different angles. Compared to traditional SAR, multi-angle SAR offers the advantage of spatial diversity, capturing target anisotropy and eliminating inherent shadowing and overlap in SAR images. This provides rich information on target scattering characteristics for SAR data applications such as image enhancement, image super-resolution, and automatic target recognition. Broadly speaking, multi-angle SAR can be divided into two categories. One is single-platform multi-angle SAR imaging, which uses the system to obtain detection results at different times and angles. This system has a simple architecture, but long data acquisition times and limited adaptability to rapidly changing target scenes. The other is multi-platform SAR systems that simultaneously detect different observation angles. While applicable to a wide range of imaging scenarios, common radar systems operate in single-transmitter-multiple-receiver or multi-transmitter-multiple-receiver modes, resulting in relatively complex system architectures and the inevitable occurrence of time-frequency and phase synchronization errors. Therefore, how to quickly and flexibly implement multi-angle SAR observations is a critical issue that needs to be addressed.

[0005] To obtain high-resolution SAR images, there are two main multi-angle SAR imaging methods. One imaging method is based on parameterized modeling, characterizing the regularized scattering behavior of the scatterer. It then extracts the target's electromagnetic scattering characteristics from the SAR observation data and further inverts the target's characteristic information from the scattering characteristics. However, the target's electromagnetic scattering model parameters are highly dimensional, mutually coupled, and computationally complex. The other imaging method is the subaperture method, which divides the full aperture into several subapertures along the azimuth direction. Because the imaging angle range of each subaperture is relatively small, the scattering characteristics of the target in each subaperture can be assumed to be the same. Traditional matched filtering methods or regularized optimization methods are used to image the echo data of each subaperture, and all subimages are then synthesized. This method does not fully utilize the correlation between the observation angles, and the final scene image is simply a fusion of the images reconstructed from the subaperture data. Therefore, how to use multi-angle data to achieve high-resolution SAR imaging is another important issue that needs to be addressed. Summary of the Invention

[0006] The embodiments of the present invention provide a multi-angle SAR imaging method and device based on a swarm of drones, which, on the basis of quickly and flexibly obtaining SAR radar imaging, utilizes the multi-angle target characteristics to reduce the influence of sidelobes and noise, weaken the overlap and shadow of SAR images, and achieve high-resolution SAR radar imaging.

[0007] In a first aspect, an embodiment of the present invention provides a multi-angle SAR imaging method based on a drone swarm, comprising:

[0008] A SAR system based on a swarm of drones is used to obtain a target image of each drone-borne SAR about the target, and each target image is fused to obtain a fused image; wherein, each drone-borne SAR detects the target from a different angle, and each drone-borne SAR is in a self-transmitting and self-receiving mode.

[0009] For any UAV-mounted SAR, the target image of the UAV-mounted SAR is iteratively optimized based on the alternating direction multiplier optimization algorithm and fused images, and the optimized target image of the UAV-mounted SAR is obtained as the multi-angle radar imaging map of the target.

[0010] In one possible implementation, the target image of the UAV-borne SAR is iteratively optimized based on an alternating direction multiplier optimization algorithm and a fused image to obtain an optimized target image of the UAV-borne SAR, including:

[0011] Based on the fused image, sparse prior constraints and structural prior constraints are constructed.

[0012] An optimization function is established based on the target image, echo matrix, measurement matrix, sparse prior constraints and structural prior constraints of the UAV-borne SAR.

[0013] Based on the fused image and the optimization function, the augmented Lagrangian function is established.

[0014] Based on the augmented Lagrangian function, local variables, sparse variables, structural variables and dual variables are obtained.

[0015] Solving for local variables yields the first result; solving for sparse variables yields the second result; solving for structural variables yields the third result; and solving for dual variables yields the fourth result.

[0016] Based on the first result, the local variables are updated to obtain updated local variables; based on the second result, the sparse variables are updated to obtain updated sparse variables; based on the third result, the structural variables are updated to obtain updated structural variables; based on the fourth result, the dual variables are updated to obtain updated dual variables.

[0017] The residual is calculated based on the updated local variables and the local variables before and after the update.

[0018] If the residual meets the preset conditions or the iteration reaches the maximum number of iterations, the updated local variables are used as the target image after the UAV-borne SAR optimization; otherwise, return to the step "update the local variables based on the first result to obtain the updated local variables" and subsequent steps.

[0019] In one possible implementation, an optimization function is established based on the target image, echo matrix, measurement matrix, sparse prior constraints, and structural prior constraints of the UAV-mounted SAR, including:

[0020] Based on the echo matrix, measurement matrix and the target image of the UAV-borne SAR, consistency constraints are obtained.

[0021] An optimization function is established based on consistency constraints, sparse prior constraints and structural prior constraints.

[0022] In one possible implementation, an augmented Lagrangian function is established based on the fused image and the optimization function, including:

[0023] An augmented Lagrangian function is established based on the optimization function, the fused image and the modulus of the fused image.

[0024] In one possible implementation, solving the local variables to obtain the first result includes:

[0025] The local variables are solved based on the least squares method to obtain the first result.

[0026] Solving for the structural variables yields the third result, including:

[0027] The third result is obtained by solving the structural variables based on the least squares method.

[0028] In one possible implementation, solving the sparse variable to obtain the second result includes:

[0029] The second result is obtained by solving the sparse variables based on the principle of proximal gradient minimization.

[0030] In one possible implementation, solving the dual variables to obtain the fourth result includes:

[0031] Based on the first result, the second result and the third result, the dual variable is solved to obtain the fourth result.

[0032] In one possible implementation, the dual variable includes a first dual variable and a second dual variable; and the dual variable is updated based on the fourth result to obtain an updated dual variable, including:

[0033] Based on the updated local variables, the updated sparse variables, and the fourth result, the first dual variables are updated to obtain updated first dual variables.

[0034] Based on the updated local variables, the updated structural variables and the fourth result, the second dual variables are updated to obtain updated second dual variables.

[0035] In one possible implementation, the residual is calculated based on the updated local variables and the local variables before and after the update, including:

[0036] The difference between the local variable after updating and the local variable before updating is calculated and recorded as the first difference.

[0037] An L2 norm of the first difference is calculated to obtain a fifth result.

[0038] Calculate the L2 norm of the local variable before updating and record it as the sixth result.

[0039] The ratio of the fifth result to the sixth result is taken as the residual.

[0040] In a second aspect, an embodiment of the present invention provides a multi-angle SAR imaging device based on a drone swarm, comprising:

[0041] The first processing module is used to use the SAR system based on the drone swarm to obtain a target image of the target from each drone-mounted SAR and to fuse each target image to obtain a fused image; wherein each drone-mounted SAR detects the target from a different angle and each drone-mounted SAR is in a self-transmitting and self-receiving mode.

[0042] The second processing module is used to iteratively optimize the target image of any UAV-borne SAR based on the alternating direction multiplier optimization algorithm and the fused image, and obtain the optimized target image of the UAV-borne SAR as the multi-angle radar imaging map of the target.

[0043] In a third aspect, an embodiment of the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method in the first aspect or any possible implementation of the first aspect is implemented.

[0044] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method in the first aspect or any possible implementation of the first aspect.

[0045] In a fifth aspect, an embodiment of the present invention provides a computer program product, including a computer program, which, when executed by a processor, implements the method in the first aspect or any possible implementation of the first aspect.

[0046] In an embodiment of the present invention, SAR based on a swarm of drones and multi-angle SAR are combined. First, the target images of each drone at different angles of the target are fused to obtain a fused image. Then, the target image corresponding to any drone is iteratively optimized based on the alternating direction multiplier algorithm and the fused image to obtain a multi-angle radar imaging map of the target. In this process, the alternating direction multiplier algorithm incorporates the constraints of sparse prior and structural prior. Combined with the target images of each drone at different angles, it can fully utilize the multi-angle characteristics of the target, suppress the influence of sidelobes and overlap, weaken overlap and shadows in the final image, and improve the imaging clarity and resolution. The drone obtains the target image in a self-transmitting and self-receiving mode, which can reduce the influence of synchronization errors during image fusion and reduce the time consumption of aperture synthesis. In other words, it reduces the time of image fusion, thereby reducing the final radar imaging time and quickly obtaining SAR imaging. In addition, the drone of this solution has a shorter trajectory, which means less time is needed to collect data. When the target scene changes rapidly, it can well adapt to the rapidly changing scene, with higher flexibility and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flowchart of an implementation of a multi-angle SAR imaging method based on a swarm of drones provided by an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of the process of multi-angle SAR imaging based on a swarm of drones provided by an embodiment of the present invention;

[0049] Figure 3 1 is a schematic structural diagram of a multi-angle SAR imaging device based on a drone swarm provided by an embodiment of the present invention;

[0050] Figure 4 is a schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0052] See also Figure 1 , which shows a flowchart of the implementation of the multi-angle SAR imaging method based on a drone swarm provided by an embodiment of the present invention, and is detailed as follows:

[0053] Step 101: Utilize a SAR system based on a swarm of drones to obtain a target image of a target from each drone-borne SAR, and fuse each target image to obtain a fused image; wherein each drone-borne SAR detects the target from a different angle, and each drone-borne SAR is in a self-transmitting and self-receiving mode.

[0054] For example, the target here can refer to a target object or target scene. The SAR mode means that the SAR system on the drone is responsible for both transmitting radar waves and receiving reflected radar signals. The target image is the radar image converted from the reflected radar signals.

[0055] For example, each drone-mounted SAR detects the target from a different angle. Specifically, for an object that does not move, the drones are distributed in different areas in the sky. Specifically, Figure 2 As shown in the example, assuming the target is a building on the ground and there are three drones, each drone will fly to a specified height and trace a circular trajectory around the building. All three drones fly simultaneously to complete this trajectory, with each drone's movement angle reaching 120 degrees (360 / 3). If the object is moving, the circular trajectory will become an ellipse. All other settings remain unchanged, but the flight distances of the different drones will differ.

[0056] Among them, since the drone's airborne SAR obtains the target image of the complete target (the target image contains the complete target), even if one or several drones are damaged, a multi-angle radar imaging map can be formed. However, the clarity of the target outline will be slightly reduced compared to when there is no drone damage, and the target will not be unable to be imaged.

[0057] In this embodiment, clustered drones are used as payload platforms and distributed in different directions of the target scene. The SAR on each drone adopts a self-transmitting and self-receiving mode, which can reduce the impact of synchronization errors. Then, a clustered distributed short aperture is formed in three-dimensional space. Through the collaboration between the clustered drones, a long aperture is synthesized in a short time, which greatly reduces the time consumption of aperture synthesis and realizes multi-angle real-time imaging of the target scene in a short time.

[0058] Step 102 : For any UAV-borne SAR, the target image of the UAV-borne SAR is iteratively optimized based on the alternating direction multiplier optimization algorithm and the fused image to obtain the optimized target image of the UAV-borne SAR as the multi-angle radar imaging image of the target.

[0059] In one possible implementation, iteratively optimizing the target image of the UAV-borne SAR based on an alternating direction multiplier optimization algorithm and a fused image to obtain the optimized target image of the UAV-borne SAR may include:

[0060] Based on the fused image, sparse prior constraints and structural prior constraints are constructed.

[0061] An optimization function is established based on the target image, echo matrix, measurement matrix, sparse prior constraints and structural prior constraints of the UAV-borne SAR.

[0062] Based on the fused image and the optimization function, the augmented Lagrangian function is established.

[0063] Based on the augmented Lagrangian function, local variables, sparse variables, structural variables and dual variables are obtained.

[0064] Solving for local variables yields the first result; solving for sparse variables yields the second result; solving for structural variables yields the third result; and solving for dual variables yields the fourth result.

[0065] Based on the first result, the local variables are updated to obtain updated local variables; based on the second result, the sparse variables are updated to obtain updated sparse variables; based on the third result, the structural variables are updated to obtain updated structural variables; based on the fourth result, the dual variables are updated to obtain updated dual variables.

[0066] The residual is calculated based on the updated local variables and the local variables before and after the update.

[0067] If the residual meets the preset conditions or the iteration reaches the maximum number of iterations, the updated local variables are used as the target image after the UAV-borne SAR optimization; otherwise, return to the step "update the local variables based on the first result to obtain the updated local variables" and subsequent steps.

[0068] Specifically, the imaging algorithm based on the consistency alternating direction multiplier optimization algorithm (ADMM distributed optimization framework) transforms the UAV swarm SAR imaging problem into a global solution of distributed optimization. Each single platform imaging is used as a local image (local variable), and the UAV swarm imaging is used as a global image (fused image). Then, using the consistency constraint criterion, the global variable is imposed as a consistency constraint in the local variable optimization to better retain the common features between the local variables. Finally, through the collaboration between the local variables, high-quality global variables are obtained in a data-driven manner.

[0069] For example, after obtaining the first result, the second result, the third result and the fourth result, the iteration is performed. At this time, the hyperparameters need to be set and initialized first. Specifically, the iteration count k = 0, the maximum number of iterations k max , parameters λ1, λ2 and β, tolerance ε, variables Z1 0 =0,Z2 0 =0, After initialization, the loop begins to obtain updated local variables, updated sparse variables, updated structural variables, updated dual variables and updated residuals. When the residual does not meet the preset conditions or the iteration does not reach the maximum number of iterations, the number of iterations is increased by 1, and the "update the local variables based on the first result to obtain the updated local variables" and subsequent steps are executed again.

[0070] Exemplarily, the residual meeting the preset condition specifically refers to the residual being less than or equal to the tolerance ε.

[0071] For example, the echo matrix and the measurement matrix are for each drone, and each drone corresponds to an echo matrix and a measurement matrix.

[0072] In one possible implementation, an optimization function is established based on the target image, echo matrix, measurement matrix, sparse prior constraints, and structural prior constraints of the UAV-mounted SAR, which may include:

[0073] Based on the echo matrix, measurement matrix and the target image of the UAV-borne SAR, consistency constraints are obtained.

[0074] An optimization function is established based on consistency constraints, sparse prior constraints and structural prior constraints.

[0075] For example, a non-convex but continuous MC function is selected as the sparse prior constraint, and an RTV with an adaptive image structure protection function is selected as the structural prior constraint. A composite regularizer is constructed to achieve the joint enhancement of the sparse and structural features of the SAR imaging target. The obtained optimization function can be:

[0076]

[0077] in, represents the consistency constraint, P MC (X global ) represents the sparse prior constraint, P RTV (|X global |) represents the structural prior constraint, λ1 represents the MC regularization penalty coefficient, λ2 represents the RTV regularization penalty coefficient, and Y q represents the echo matrix, A q represents the measurement matrix, X q represents the target image of the UAV-mounted SAR, X global Represents the fused image.

[0078] In one possible implementation, establishing an augmented Lagrangian function based on the fused image and the optimization function may include:

[0079] An augmented Lagrangian function is established based on the optimization function, the fused image and the modulus of the fused image.

[0080] For example, in order to establish the augmented Lagrangian function, first determine the sparse variable Z1 and the structure variable Z2, Z1 = X global , Z2=|X global Specifically, the augmented Lagrangian function can be:

[0081]

[0082] Where X=[X1,X2,…X Q ], is the target image acquired by Q UAV airborne SAR, also known as local variable, U1=[U 11 ,U 12 ,…,U 1Q ] and U2=[U 21 ,U 22 ,…,U 2Q ] represents the dual variable, and β>0 represents the augmented Lagrange multiplier coefficient.

[0083] The strategy of ADMM optimization solution is to update the local variable X alternately. q , sparse variables Z1, structural variables Z2 and dual variables U 1q 、U 2q The update process is as follows:

[0084]

[0085] Where k is the number of iterations.

[0086] In one possible implementation, solving the local variable to obtain the first result may include:

[0087] The local variables are solved based on the least squares method to obtain the first result.

[0088] For example, the local variable X q The solution is a typical least squares problem. To X q is differentiable, through Obtaining the closed form, we get:

[0089] X q k+1 =(A q H A q +βI) -1 (A q H Y q +β(Z1 k +Z2 k )-(U 1q k +U 2q k ))

[0090] Among them, the measurement matrix A q It can be obtained according to the reprojection operation, It can be obtained according to the back-projection operation, where I represents a unit vector, and the unit vector is a matrix whose number of rows and columns are the same as the number of rows and columns of the measurement matrix, respectively.

[0091] Solving for the structural variables to obtain the third result may include:

[0092] The third result is obtained by solving the structural variables based on the least squares method.

[0093] For example, the solution of the above structural variable Z2 is an RTV regularization problem, which can be transformed into an iterative weighted least squares algorithm to obtain its closed-form solution:

[0094]

[0095] ψ RTV (X,λ)=X(1+λL k ) -1

[0096] in, For local variable X q k+1 The global consistency variable can be obtained by pixel-wise maximization over Q images.

[0097] In one possible implementation, solving the sparse variable to obtain the second result may include:

[0098] The second result is obtained by solving the sparse variables based on the principle of proximal gradient minimization.

[0099] For example, solving the above sparse variable Z1 is a non-convex optimization problem, and its corresponding closed-form solution can be solved using the proximal gradient minimum principle:

[0100]

[0101] Here, θ represents the concavity of the control penalty function.

[0102] In one possible implementation, solving the dual variable to obtain the fourth result may include:

[0103] Based on the first result, the second result and the third result, the dual variable is solved to obtain the fourth result.

[0104] In one possible implementation, the dual variable includes a first dual variable and a second dual variable; and updating the dual variable based on the fourth result to obtain an updated dual variable may include:

[0105] Based on the updated local variables, the updated sparse variables, and the fourth result, the first dual variables are updated to obtain updated first dual variables.

[0106] Based on the updated local variables, the updated structural variables and the fourth result, the second dual variables are updated to obtain updated second dual variables.

[0107] For example, in the local variable X q After the sparse variables Z1 and structure variables Z2 are updated, the dual variables U 1q 、U 2q Solve:

[0108]

[0109] In one possible implementation, calculating the residual based on the updated local variables and the local variables before and after the update may include:

[0110] The difference between the local variable after updating and the local variable before updating is calculated and recorded as the first difference.

[0111] An L2 norm of the first difference is calculated to obtain a fifth result.

[0112] Calculate the L2 norm of the local variable before updating and record it as the sixth result.

[0113] The ratio of the fifth result to the sixth result is taken as the residual.

[0114] According to the definition of residual and stopping criterion, define res = ||Xq k+1 -X q k ||2 / ||X q k ||2 is used as the residual. When the residual is less than the tolerance ε, the iteration is stopped and the global optimization result is output.

[0115] The multi-angle SAR imaging method based on a swarm of drones combines SAR based on a swarm of drones with multi-angle SAR. First, the target images of each drone at different angles are fused to obtain a fused image. Then, based on the alternating direction multiplier algorithm and the fused image, the target image corresponding to any drone is iteratively optimized to obtain a multi-angle radar image of the target. In this process, the alternating direction multiplier algorithm incorporates sparse prior and structural prior constraints. Combining the target images from each drone at different angles can fully utilize the multi-angle characteristics of the target, suppress the influence of sidelobes and overlap, and reduce overlap and shadows in the final image, thereby improving the imaging clarity and resolution. The drones obtain target images in a self-transmitting and self-receiving mode, which can reduce the influence of synchronization errors during image fusion and the time consumed by aperture synthesis. In other words, the image fusion time is reduced, thereby reducing the final radar imaging time and quickly obtaining SAR images. In addition, the drones in this scheme have shorter trajectory, which reduces the time required for data collection. When the target scene changes rapidly, the drone can well adapt to the rapidly changing scene, and has higher flexibility and adaptability.

[0116] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0117] The following are device embodiments of the present invention. For details not fully described therein, reference may be made to the corresponding method embodiments described above.

[0118] Figure 3 The following is a schematic diagram showing the structure of a multi-angle SAR imaging device based on a drone swarm according to an embodiment of the present invention. For ease of explanation, only the parts related to the embodiment of the present invention are shown, which are described in detail as follows:

[0119] like Figure 3 As shown, the multi-angle SAR imaging device based on the drone swarm includes:

[0120] The first processing module 201 is used to use the SAR system based on the drone swarm to obtain a target image of the target from each drone-mounted SAR and to fuse each target image to obtain a fused image; wherein each drone-mounted SAR detects the target from a different angle and each drone-mounted SAR is in a self-transmitting and self-receiving mode.

[0121] The second processing module 202 is used to iteratively optimize the target image of any UAV-borne SAR based on the alternating direction multiplier optimization algorithm and the fused image, and obtain the optimized target image of the UAV-borne SAR as the multi-angle radar imaging image of the target.

[0122] In a possible implementation, the second processing module 202 may be configured to:

[0123] Based on the fused image, sparse prior constraints and structural prior constraints are constructed.

[0124] An optimization function is established based on the target image, echo matrix, measurement matrix, sparse prior constraints and structural prior constraints of the UAV-borne SAR.

[0125] Based on the fused image and the optimization function, the augmented Lagrangian function is established.

[0126] Based on the augmented Lagrangian function, local variables, sparse variables, structural variables and dual variables are obtained.

[0127] Solving for local variables yields the first result; solving for sparse variables yields the second result; solving for structural variables yields the third result; and solving for dual variables yields the fourth result.

[0128] Based on the first result, the local variables are updated to obtain updated local variables; based on the second result, the sparse variables are updated to obtain updated sparse variables; based on the third result, the structural variables are updated to obtain updated structural variables; based on the fourth result, the dual variables are updated to obtain updated dual variables.

[0129] The residual is calculated based on the updated local variables and the local variables before and after the update.

[0130] If the residual meets the preset conditions or the iteration reaches the maximum number of iterations, the updated local variables are used as the target image after the UAV-borne SAR optimization; otherwise, return to the step "update the local variables based on the first result to obtain the updated local variables" and subsequent steps.

[0131] In a possible implementation, the second processing module 202 may be configured to:

[0132] Based on the echo matrix, measurement matrix and the target image of the UAV-borne SAR, consistency constraints are obtained.

[0133] An optimization function is established based on consistency constraints, sparse prior constraints and structural prior constraints.

[0134] In a possible implementation, the second processing module 202 may be configured to:

[0135] An augmented Lagrangian function is established based on the optimization function, the fused image and the modulus of the fused image.

[0136] In a possible implementation, the second processing module 202 may be configured to:

[0137] The local variables are solved based on the least squares method to obtain the first result.

[0138] The second processing module 202 may be used to:

[0139] The third result is obtained by solving the structural variables based on the least squares method.

[0140] In a possible implementation, the second processing module 202 may be configured to:

[0141] The second result is obtained by solving the sparse variables based on the principle of proximal gradient minimization.

[0142] In a possible implementation, the second processing module 202 may be configured to:

[0143] Based on the first, second, and third results, solve for the dual variable.

[0144] In a possible implementation, the second processing module 202 may be configured to:

[0145] Based on the updated local variables, the updated sparse variables, and the fourth result, the first dual variables are updated to obtain updated first dual variables.

[0146] Based on the updated local variables, the updated structural variables and the fourth result, the second dual variables are updated to obtain updated second dual variables.

[0147] In a possible implementation, the second processing module 202 may be configured to:

[0148] The difference between the local variable after updating and the local variable before updating is calculated and recorded as the first difference.

[0149] An L2 norm of the first difference is calculated to obtain a first result.

[0150] Calculate the L2 norm of the local variable before updating and record it as the second result.

[0151] The ratio of the first result to the second result is taken as the residual.

[0152] Figure 4 Schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 4As shown, the electronic device 5 of this embodiment includes: a processor 50 and a memory 51. The memory 51 stores a computer program 52. When the processor 50 executes the computer program 52, the steps of the above-mentioned method embodiments are implemented. Alternatively, when the processor 50 executes the computer program 52, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0153] For example, the computer program 52 may be divided into one or more modules / units, which are stored in the memory 51 and executed by the processor 50 to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program 52 in the electronic device 5.

[0154] The electronic device 5 may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art will appreciate that Figure 4 It is only an example of the electronic device 5 and does not constitute a limitation of the electronic device 5. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device 5 may also include input and output devices, network access devices, buses, etc.

[0155] The processor 50 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0156] The memory 51 can be an internal storage unit of the electronic device 5, such as a hard drive or memory of the electronic device 5. The memory 51 can also be an external storage device of the electronic device 5, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. equipped on the electronic device 5. Furthermore, the memory 51 can include both an internal storage unit of the electronic device 5 and an external storage device. The memory 51 is used to store the computer program 52 and other programs and data required by the electronic device 5. The memory 51 can also be used to temporarily store data that has been output or is about to be output.

[0157] For the sake of convenience and brevity, the division of the above functional modules / units is only used as an example. In actual applications, the above functions can be assigned to different functional modules / units as needed. The above modules / units can be implemented in the form of hardware, software, or a combination of hardware and software.

[0158] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods in the above-mentioned method embodiments.

[0159] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the methods in the above-mentioned method embodiments.

[0160] The computer program includes computer program code, which may be in source code form, object code form, executable file, or some intermediate form. Computer-readable media may include any entity or device capable of carrying computer program code, recording media, USB flash drives, mobile hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunications signals, and software distribution media.

[0161] In the above embodiments, the descriptions of each embodiment have their own focus. For parts not described or recorded in detail in one embodiment, please refer to the relevant descriptions of other embodiments. Unless otherwise specified or there is a logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced to each other. The technical features of different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0162] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A multi-angle SAR imaging method based on drone swarms, characterized in that: include: Using a SAR system based on a swarm of drones, obtaining a target image of a target from each drone-borne SAR, and fusing each of the target images to obtain a fused image; wherein each drone-borne SAR detects the target from a different angle, and each drone-borne SAR is in a self-transmitting and self-receiving mode; For any unmanned aerial vehicle (UAV) SAR, the target image of the UAV SAR is iteratively optimized based on the alternating direction multiplier optimization algorithm and the fused image to obtain the optimized target image of the UAV SAR as the multi-angle radar imaging image of the target.

2. The multi-angle SAR imaging method based on drone swarm according to claim 1, characterized in that: The iterative optimization of the target image of the UAV-borne SAR based on the alternating direction multiplier optimization algorithm and the fused image to obtain the optimized target image of the UAV-borne SAR includes: Based on the fused image, constructing a sparse prior constraint and a structural prior constraint; Establishing an optimization function based on the target image, the echo matrix, the measurement matrix, the sparse prior constraint and the structural prior constraint of the UAV-borne SAR; Establishing an augmented Lagrangian function based on the fused image and the optimization function; Based on the augmented Lagrangian function, local variables, sparse variables, structural variables and dual variables are obtained; Solving the local variables to obtain a first result; solving the sparse variables to obtain a second result; solving the structural variables to obtain a third result; solving the dual variables to obtain a fourth result; The local variables are updated based on the first result to obtain updated local variables; the sparse variables are updated based on the second result to obtain updated sparse variables; the structural variables are updated based on the third result to obtain updated structural variables; the dual variables are updated based on the fourth result to obtain updated dual variables; Based on the updated local variables and the local variables before and after the update, the residual is calculated; If the residual satisfies the preset conditions or the iteration reaches the maximum number of iterations, the updated local variables are used as the target image after optimization of the UAV-borne SAR; otherwise, the process returns to step "updating the local variables based on the first result to obtain updated local variables" and subsequent steps.

3. The multi-angle SAR imaging method based on drone swarm according to claim 2, characterized in that: The optimization function is established based on the target image, the echo matrix, the measurement matrix, the sparse prior constraint and the structural prior constraint of the unmanned aerial vehicle SAR, including: Obtaining consistency constraints based on the echo matrix, the measurement matrix, and the target image of the UAV-borne SAR; The optimization function is established based on the consistency constraint, the sparsity prior constraint and the structural prior constraint.

4. The multi-angle SAR imaging method based on drone swarm according to claim 2, characterized in that: The step of establishing an augmented Lagrangian function based on the fused image and the optimization function includes: The augmented Lagrangian function is established based on the optimization function, the fused image and the modulus of the fused image.

5. The multi-angle SAR imaging method based on drone swarm according to claim 2, characterized in that: Solving the local variable to obtain a first result includes: Solving the local variables based on the least squares method to obtain a first result; Solving the structural variables to obtain a third result includes: The structural variables are solved based on the least square method to obtain a third result.

6. The multi-angle SAR imaging method based on drone swarm according to claim 5, characterized in that: Solving the sparse variable to obtain a second result includes: The sparse variable is solved based on the proximal gradient minimum principle to obtain the second result.

7. The multi-angle SAR imaging method based on drone swarm according to claim 6, characterized in that: Solving the dual variable to obtain a fourth result includes: Based on the first result, the second result and the third result, the dual variable is solved to obtain a fourth result.

8. The multi-angle SAR imaging method based on drone swarm according to claim 2, characterized in that: The dual variables include a first dual variable and a second dual variable; and updating the dual variables based on the fourth result to obtain updated dual variables includes: updating the first dual variable based on the updated local variable, the updated sparse variable, and the fourth result to obtain an updated first dual variable; Based on the updated local variables, the updated structural variables and the fourth result, the second dual variables are updated to obtain updated second dual variables.

9. The multi-angle SAR imaging method based on drone swarm according to claim 2, characterized in that: The residual is calculated based on the updated local variables and the local variables before the update, including: Calculating a difference between the updated local variable and the local variable before the update, and recording the difference as a first difference; Calculating an L2 norm of the first difference to obtain a fifth result; Calculate the L2 norm of the local variable before the update, and record it as a sixth result; The ratio of the fifth result to the sixth result is used as the residual.

10. A multi-angle SAR imaging device based on a swarm of drones, characterized in that: include: a first processing module configured to utilize a SAR system based on a swarm of drones to obtain a target image of a target from each drone-borne SAR, and to fuse each of the target images to obtain a fused image; wherein each drone-borne SAR detects the target from a different angle, and each drone-borne SAR is in a self-transmitting and self-receiving mode; The second processing module is used to iteratively optimize the target image of any unmanned aerial vehicle SAR based on the alternating direction multiplier optimization algorithm and the fused image, and obtain the optimized target image of the unmanned aerial vehicle SAR as the multi-angle radar imaging image of the target.

Images