A three-dimensional back-projection imaging method and device for array interferometric SAR

Abstract

The present invention discloses a three-dimensional reverse projection imaging method and device of array interferometric SAR, which belongs to the field of SAR three-dimensional imaging, including: a low-altitude scene three-dimensional array SAR imaging model based on polar coordinate system and Cartesian coordinate system, using a three-dimensional grid interpolation method to replace registration, and reconstructing three-dimensional information by using a compressed sensing (CS) algorithm through an extended sensing matrix to accurately achieve high-quality three-dimensional reconstruction of the observed scene. Under the premise of ensuring computational efficiency, the present invention improves the resolution and accuracy of three-dimensional imaging of array interferometric SAR under low-altitude conditions, significantly improves the reconstruction quality of overlapping areas, and improves the three-dimensional entropy index in measured data by about 0.1.

Description

Technical Field

[0001] The invention belongs to the field of SAR three-dimensional imaging, and in particular relates to a three-dimensional back-projection imaging method and device of array interferometric SAR. Background Art

[0002] As an important observation technology, Synthetic Aperture Radar (SAR) 3D imaging has been widely used on Unmanned Aerial Vehicle (UAV) platforms due to its convenience and versatility. Compared with satellite and airborne platforms, the emergence of UAVs simplifies data collection methods, making SAR 3D imaging more convenient. In particular, UAV array SAR can stably obtain multiple coherent observations with only one channel.

[0003] However, due to the lower flight altitude of UAVs compared to satellite-borne and airborne platforms, many traditional algorithms based on compressed sensing fail. Not only is it necessary to infer the mechanism based on the spherical wave model, but the traditional registration method is also not fully applicable. Registration needs to ensure that all targets within the same distance grid unit are solved, which is difficult to achieve at low altitudes because when targets overlap in one antenna, the channel of the other antenna will inevitably separate, which is called channel migration. In low-altitude SAR three-dimensional imaging, once the offset exceeds one distance grid unit, the traditional registration method will fail and the resolution of the three-dimensional imaging will be greatly affected.

[0004] In order to achieve super-resolution in three-dimensional reconstruction and effectively solve the channel migration problem, the existing literature uses the BP method of raw SAR data, focusing on circular trajectories, and then uses the Compressive Sensing (CS) method in the elevation direction. However, this strategy is mainly aimed at holographic systems and high-altitude platforms, especially in forest environments, and cannot alleviate the low-altitude channel migration problem of the present invention. Alternatively, the existing literature uses a three-dimensional grid to extract the elevation signal, and then uses the CS method to identify the target of each grid point. This method combines 3D interpolation with CS technology, but because the phase information is still not fully utilized, and the estimation of amplitude and phase still mainly relies on the least squares method, this method cannot achieve complete three-dimensional super-resolution. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a three-dimensional backprojection imaging method and device for array interferometric SAR. The method improves the resolution and accuracy of three-dimensional imaging of array interferometric SAR while ensuring the computational efficiency, especially in the overlapped area, greatly improves the three-dimensional imaging quality, and improves the three-dimensional entropy evaluation by about 0.1.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0007] In one aspect, the present invention provides a three-dimensional back-projection imaging method for array interferometric SAR, comprising the following steps:

[0008] Step 1: The UAV performs flight photography to obtain the array interferometric SAR raw echo data of the target scene;

[0009] Step 2: Processing the original echo data of the array interferometric SAR to obtain multiple groups of two-dimensional imaging results of the array channels;

[0010] Step 3: Establish a three-dimensional coordinate system, divide the grid of the main antenna, obtain a three-dimensional grid, and filter out grids with amplitude vectors below the set value;

[0011] Step 4: interpolate the non-low-value grid to obtain the observation vector matrix corresponding to the three-dimensional grid, and use the observation vector matrix as the perception matrix;

[0012] Step 5: Expand the perception matrix to match different angles of different distance units, use the compressed sensing algorithm to perform sparse 3D imaging, and obtain the initial 3D point cloud of the target scene;

[0013] Step 6: Process the initial three-dimensional point cloud to obtain a complete three-dimensional imaging point cloud of the target area.

[0014] On the other hand, the present invention provides a three-dimensional back projection imaging device of array interferometric SAR, which realizes three-dimensional imaging of array interferometric SAR based on the three-dimensional back projection imaging method of array interferometric SAR, and includes the following modules:

[0015] The data acquisition module is used to acquire the raw echo data of the array interferometric SAR of the target scene during flight shooting of the UAV;

[0016] The imaging result acquisition module processes the raw echo data of the array interferometric SAR to obtain multiple groups of two-dimensional imaging results of the array channels;

[0017] A grid division module establishes a three-dimensional coordinate system, divides the grid of the main antenna, obtains a three-dimensional grid, and filters out grids with a vector amplitude below a set value;

[0018] An observation vector acquisition module interpolates non-low-value grids to obtain an observation vector matrix corresponding to the three-dimensional grid, and the observation vector matrix is ​​used as a perception matrix;

[0019] The initial 3D point cloud acquisition module expands the perception matrix to match different angles of different distance units, uses the compressed sensing algorithm for sparse 3D imaging, and obtains the initial 3D point cloud of the target scene;

[0020] The imaging module processes the initial three-dimensional point cloud to obtain a complete three-dimensional imaging point cloud of the target area.

[0021] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned three-dimensional reverse projection imaging method of array interferometric SAR when executing the program.

[0022] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned three-dimensional backprojection imaging method of array interferometric SAR are performed.

[0023] The beneficial effects of the present invention are:

[0024] The present invention designs a three-dimensional back-projection imaging method and device based on array interferometric SAR observations under low-altitude conditions, combines the precise positioning capability of three-dimensional back-projection with the super-resolution of a compressed sensing (CS) algorithm, solves the problem of difficulty in registration caused by channel migration caused by low-altitude observations, effectively reduces the energy dispersion caused by channel migration, thereby accurately reconstructing weak overlapping points, and achieving accurate super-resolution three-dimensional reconstruction effects while ensuring computational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a flow chart of the three-dimensional back-projection imaging method of the array interferometric SAR of the present invention;

[0026] Figure 2 It is a schematic diagram of the geometric model of array SAR three-dimensional interpolation imaging;

[0027] Figure 3 is a schematic diagram of an extended sensing matrix of the present invention;

[0028] Figure 4 SAR two-dimensional imaging results and selected target areas;

[0029] Figure 5 is a comparison diagram of SAR three-dimensional imaging results of the first area; wherein a, b, c are the results obtained by the traditional registration and CS methods, d, e, f are the results obtained by the method of the present invention; a, d are three-dimensional point cloud images, b, e are azimuth direction side views, c, f are billboard target slice images;

[0030] Figure 6a , Figure 6b is a comparison chart of SAR three-dimensional imaging results of the second area; among them, Figure 6a To use the traditional registration and CS method for the 3D point cloud, Figure 6bA three-dimensional point cloud image using the method of the present invention;

[0031] Figure 7a , Figure 7b is a comparison chart of the overall SAR three-dimensional imaging results; among them, Figure 7a The three-dimensional reconstruction image using the traditional registration and CS method, Figure 7b This is a three-dimensional reconstruction image using the method of the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0033] In the description of the present invention, it should be understood that the terms "upper", "lower", "top", "bottom", "inner", "outer", etc. used in the present invention to indicate the orientation or position relationship is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0034] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0035] like Figure 1 As shown, the present invention provides a three-dimensional back-projection imaging method of an array interferometric SAR, comprising the following steps:

[0036] Step 1: The UAV performs flight photography to obtain the array interferometric SAR raw echo data of the target scene, including:

[0037] The UAV conducts flight photography of the target scene and obtains the original SAR echo data of the target scene.

[0038] Step 2: Process the original echo data of the array interferometric SAR to obtain multiple groups of two-dimensional imaging results of the array channels:

[0039] That is, the array interferometric SAR raw echo data obtained in step 1 is processed into a two-dimensional imaging result.

[0040] Step 3: Establish a three-dimensional coordinate system, divide the grid of the main antenna, obtain a three-dimensional grid, and filter out grids with smaller amplitude vectors, such as filtering out grids with amplitudes below 0.1 times the average of all grids, including:

[0041] First, under the assumption of spherical waves, an x ​​(azimuth) - r (distance) - θ (downward viewing angle) coordinate system is established to divide the grid of the main antenna, and the hard threshold method is used to quickly filter out the strongest point. The image of the reference antenna does not need to be interpolated, and the SAR two-dimensional images of other antennas only need to be interpolated into the non-low-value distance grids in the image of the reference antenna. At a certain downward viewing angle, the values ​​in the grids with longer distances are more likely to represent multipath signals than the values ​​in the grids with shorter distances. Therefore, after obtaining a grid with a certain threshold, not interpolating the subsequent grids with larger distances can effectively alleviate multipath interference and significantly speed up the three-dimensional interpolation speed.

[0042] Step 4: interpolate the non-low-value grid to obtain the observation vector corresponding to the three-dimensional grid, including:

[0043] In the three-dimensional grid obtained in step 3, an interpolation operation similar to the three-dimensional elevation back projection is used to obtain the signal of each three-dimensional grid, such as Figure 2 shown. Figure 2 In the above example, point P is a reference point in the reference grid. represents the coordinates of the reference antenna, Indicates the angle value of the reference point under the viewing angle. Using the shortest distance strategy of visual imaging, the value of the distance grid is calculated from near to far under the viewing angle, and the interpolation is calculated according to the following formula.

[0044] ;

[0045] Among them, R P Indicates the distance between the drone and a given position cell (such as Figure 2 The distance of the target in point P, N r represents the total number of distance grid units, Δr represents the distance grid unit; n r The counting symbol used for cumulative operations, represents the observed vector, represents the vector obtained after interpolation; ∑ represents the accumulation operation, and sinc(x) represents the sin(πx) / πx function. The azimuth unit is the resolution unit divided by the SAR system in the azimuth direction, that is, along the flight path. The range grid unit is usually expressed by the range resolution, which refers to the minimum range difference that the radar can distinguish in the target range direction, which can be calculated as follows.

[0046] ;

[0047] in, represents the distance grid unit, represents the speed of light, Represents the radar signal bandwidth.

[0048] The method terminates when the observed vector value reaches a threshold, and grids with smaller magnitude vectors can be further filtered out.

[0049] Step 5: Expand the perception matrix to match different angles of different distance units, and use the compressed sensing (CS) algorithm for sparse 3D imaging to obtain the initial 3D point cloud of the target scene, including:

[0050] For each orientation unit obtained in step 4, an expansion matrix is ​​constructed for CS imaging, which expands the matching matrix for a certain distance grid unit in traditional CS to a matrix containing all possible distance elements in the sliding window, such as Figure 3 shown. Figure 3 In the equation, G is the joint observation matrix of the observation vectors of the same distance history under the reference antenna, and Γ is the joint coefficient vector of the coefficient vectors of the same distance history, and the windowed distance grid unit is obtained. The extended sensing matrix is ​​to match the different angles corresponding to different distance grid units.

[0051] A distance grid unit is arranged according to the following formula:

[0052] ;

[0053] ;

[0054] ;

[0055] Among them, g represents the observation matrix, which consists of Q dimensions, represents Q array observations; γ represents the value in the grid to be reconstructed, L is the number of grids divided by the lower viewing angle, W(Nr) is the number of adjacent distance grid units after windowing, Indicates that after windowing, L W(Nr) dimensional space reconstructs grid coefficient values; represents the accurate distance matching perception matrix, Represents the distance value under the accurate grid, ε is the noise vector, exp represents the exponential function, and j represents an imaginary number. Represents electromagnetic wavelength.

[0056] By using the CS algorithm to perform sparse 3D imaging on the extended matrix, the initial 3D point cloud of the target scene under super resolution can be obtained. The compressed sensing theory can be used to estimate this equation:

[0057] ;

[0058] in, is a hyperparameter adjusted according to the noise situation, which can be decomposed into a signal plus a penalty function. The degree of the penalty function is given by to control. Indicates that the minimum value of the entire objective function is returned. , Indicated in The distance perception matrix under the condition, express norm, express In general, when the signal-to-noise ratio is low, a relatively small , and when the signal-to-noise ratio is high, a larger ,Too low signal-to-noise ratio will seriously affect the performance of super-resolution algorithm in reconstructing scattering points in elevation.

[0059] The penalty function is a constraint or penalty term introduced in the optimization problem. Its purpose is to impose restrictions on certain attributes of the model to prevent overfitting, reduce the impact of noise, or guide the optimization process in a more reasonable direction. .

[0060] Step 6: Process the initial 3D point cloud to obtain a complete 3D imaging point cloud of the target area, including:

[0061] After obtaining the initial 3D point cloud, the threshold is set according to the scattering intensity of each pixel, the inverted residual value, etc., and the point cloud is filtered to remove the noise and dark points to obtain the final 3D point cloud.

[0062] Example:

[0063] This embodiment uses a miniature Ku-band SAR system carried by a drone to conduct a sparse array three-dimensional imaging processing experiment in Tianjin, China. The drone's flight altitude is 400 meters, and the maximum target altitude is 69.2 meters. This low altitude situation will cause registration errors.

[0064] In order to fully evaluate the ability of the present invention to reconstruct overlapping parts, two areas containing prominent features were selected. The first area 1 is a building structure about 60 meters high that is partially covered by a billboard. The billboard consists of four steel pipes parallel to the ground and multiple vertical steel pipes connected. The second area 2 shows the bottom of the building overlapping with the edge of the road. This overlap can also be seen in the 2D image, such as Figure 4 shown.

[0065] 3D reconstruction of the first area 1: First, the 3D reconstruction results of the present invention and the traditional CS method in the first area 1 are compared, wherein the point cloud is represented by the xrs coordinate system. The height of the building exceeds the maximum clear height, resulting in phase entanglement. Figure 5 The upper middle image is obtained by the traditional CS method. Figure 5 As shown in Figure a, the main features of the building structure are clearly reconstructed, but the structure of the ground billboard is still unclear. Figure 5 As shown in the side view in the middle b, the reconstruction points of the billboard are relatively scattered. This is because the low-value points in the overlapping layer are affected by channel migration, resulting in energy defocusing. Figure 5 In the figure c, the slice of the reconstructed ground structure is shown. The structure of the steel pipe is not continuous, and the reconstruction effect of the supporting steel pipe is not good. Figure 5 As shown in (d), the noise clutter is significantly reduced compared to the traditional CS method. Figure 5 This is further confirmed by the side view in Figure e, where the internal structure of the ground billboard is better focused and can be reconstructed more accurately. Similarly, the slices of the reconstructed ground structure are shown in Figure 5. Figure 5 As shown in middle f, the top view clearly shows the overall picture of the four steel pipe structure, and the supporting steel pipes are clearly visible, providing accurate elevation information.

[0066] The second region 2 is 3D reconstructed: This part contains a large part of the building facade, especially the area near the bottom corner. The present invention uses the CS-based method and the method of the present invention to evaluate the reconstruction quality of the overlapping ground edge and building facade points. In order to enhance the visualization of the building and the ground, the present invention converts the elevation into the ground height, such as Figure 6a , Figure 6b The 3D point cloud shown. Figure 6a It can be seen that in terms of the structural continuity of the building, the CS-based method has certain discontinuities, especially at the edge of the road and the bottom of the building, the point cloud of the overlapping part of the ground is relatively scattered and sparse. This is because the traditional registration method will produce errors due to channel migration when reconstructing the overlapping part, resulting in weak building points being filtered out by the threshold and calculation errors. However, using the method of the present invention, the overlapping part is correctly reconstructed, and the reconstruction result clearly reflects the floor structure, and the edge of the road is also clearly visible.

[0067] 3D reconstruction and quality evaluation of the entire scene: For a more intuitive comparison, the present invention performs 3D reconstruction of the entire scene. Figure 7a , Figure 7b As shown, for the reconstruction of the entire scene, the present invention adopts the same threshold screening method and unwrapping phase variance to obtain a three-dimensional image result. Figure 7a Using the CS method for imaging, Figure 7b The imaging method proposed by the present invention can better restore the structural details of the bottom area of ​​the building, enhance the consistency of the floor structure and the height consistency when evaluating the area with overlapping ground points, and more accurately restore the road edge and billboards.

[0068] In addition to visual evaluation, the present invention also introduces three-dimensional entropy to evaluate the texture and focus quality of the image. Entropy reflects the information contained in the data. For three-dimensional data sets with high noise and uneven distribution, the entropy value is higher because the information is unpredictable; in contrast, for data sets with stronger consistency, the entropy value is lower. The present invention calculates the three-dimensional entropy of the ground and floor structures respectively. For the floor structure, the three-dimensional entropy obtained by the traditional method is 0.4051, while the method of the present invention can reduce the entropy to 0.3385, indicating that the restoration of the floor is more accurate; for the ground structure, the three-dimensional entropy obtained by the traditional method is 0.2951, while the method of the present invention reduces it to 0.1870, indicating that the coverage between the ground and floor structures is more accurate. This is because the method of the present invention avoids the energy dispersion caused by channel migration and has less noise.

[0069] It can be seen that the three-dimensional back-projection imaging method and device of array interferometric SAR proposed in the present invention has superior performance in the position and amplitude reconstruction of three-dimensional images compared with the existing methods. In actual experiments, the improvement in quality is obvious, especially in the overlapping area. The method of the present invention effectively reduces the energy dispersion caused by channel migration, thereby accurately reconstructing weak overlapping points and achieving an enhancement of nearly 0.1 in 3D entropy analysis.

[0070] The present invention also provides a three-dimensional back-projection imaging device for array interferometric SAR, comprising the following modules:

[0071] The data acquisition module is used to acquire the raw echo data of the array interferometric SAR of the target scene during flight shooting of the UAV;

[0072] The imaging result acquisition module processes the raw echo data of the array interferometric SAR to obtain multiple groups of two-dimensional imaging results of the array channels;

[0073] A grid division module establishes a three-dimensional coordinate system, divides the grid of the main antenna, obtains a three-dimensional grid, and filters out grids with a vector amplitude below a set value;

[0074] The observation vector acquisition module interpolates the non-low-value grid to obtain the observation vector corresponding to the three-dimensional grid;

[0075] The initial 3D point cloud acquisition module expands the perception matrix to match different angles of different distance units, uses the compressed sensing algorithm for sparse 3D imaging, and obtains the initial 3D point cloud of the target scene;

[0076] The imaging module processes the initial three-dimensional point cloud to obtain a complete three-dimensional imaging point cloud of the target area.

[0077] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned three-dimensional reverse projection imaging method of array interferometric SAR when executing the program.

[0078] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned three-dimensional backprojection imaging method of array interferometric SAR are implemented.

[0079] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The solutions in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and interpreted scripting language JavaScript, etc.

[0080] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

[0081] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0082] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0083] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0084] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A three-dimensional backprojection imaging method for array interferometric SAR, characterized in that: The steps include: Step 1: The UAV performs flight photography to obtain the array interferometric SAR raw echo data of the target scene; Step 2: Processing the original echo data of the array interferometric SAR to obtain multiple groups of two-dimensional imaging results of the array channels; Step 3: Establish a three-dimensional coordinate system, divide the grid of the main antenna, obtain a three-dimensional grid, and filter out grids with amplitude vectors below the set value; Step 4: interpolate the non-low-value grid to obtain the observation vector matrix corresponding to the three-dimensional grid, and use the observation vector matrix as the perception matrix; Step 5: Expand the perception matrix to match different angles of different distance units, use the compressed sensing algorithm to perform sparse 3D imaging, and obtain the initial 3D point cloud of the target scene; Step 6: Process the initial three-dimensional point cloud to obtain a complete three-dimensional imaging point cloud of the target area.

2. The three-dimensional back-projection imaging method of array interferometric SAR according to claim 1, characterized in that: The step 3 comprises: Under the spherical wave assumption, an x-r-θ coordinate system is established to divide the grid of the main antenna, and the hard threshold method is used to quickly filter out the strongest point; the grids below the amplitude vector are set to be grids below 0.1 times the amplitude mean of all grids; x represents the azimuth, r represents the distance, and θ represents the downward viewing angle.

3. The three-dimensional back-projection imaging method of array interferometric SAR according to claim 1, characterized in that: The step 4 comprises: The interpolation operation of 3D elevation back-projection is used to obtain the signal of each 3D grid.

4. The three-dimensional backprojection imaging method of array interferometric SAR according to claim 3, characterized in that: Using the shortest distance strategy of line-of-sight imaging, the value of the distance three-dimensional grid is calculated from near to far under the downward viewing angle. The calculation is performed according to the following formula and terminated when the threshold is reached: ； Among them, R P Indicates the distance between the drone and the target in a given azimuth unit, N r represents the total number of distance grid units, Δr represents the distance grid unit; n r The counting symbol used for cumulative operations, represents the observed vector, represents the vector obtained after interpolation; ∑ represents the accumulation operation, sinc(x) represents the sin(πx) / πx function, and x represents the variable of the function; Wherein, the azimuth unit is a resolution unit divided by the SAR system in the azimuth direction; the azimuth direction represents the direction along the flight path; The range grid unit is represented by the range resolution, which is the minimum range difference that the radar can distinguish in the target range direction, calculated as follows: ； in, represents the distance grid unit, represents the speed of light, Represents the radar signal bandwidth.

5. The three-dimensional back-projection imaging method of array interferometric SAR according to claim 4, characterized in that: The step 5 comprises: For each orientation unit, an expanded matching matrix is ​​constructed for compressed sensing imaging, that is, the sensing matrix is ​​expanded to a matrix containing all possible distance elements within the sliding window to match different angles of different distance units.

6. The three-dimensional back-projection imaging method of array interferometric SAR according to claim 5, characterized in that: A distance grid unit is arranged according to the following formula: ； ； ； Among them, g represents the observation matrix, which consists of Q dimensions, represents Q array observations; γ represents the value in the grid to be reconstructed, L is the number of grids divided by the lower viewing angle, W(Nr) is the number of adjacent distance grid units after windowing, Indicates that L is formed after windowing W(Nr) dimensional space reconstructs grid values; represents the accurate distance matching perception matrix, represents the distance value under the accurate grid, ε is the noise vector, exp represents the exponential function, and j is the imaginary number symbol; Represents electromagnetic wavelength.

7. The three-dimensional back-projection imaging method of array interferometric SAR according to claim 6, characterized in that: The compressed sensing algorithm is used to perform sparse 3D imaging on the expanded matrix to obtain the initial 3D point cloud of the target scene under super resolution. The compressed sensing theory is used to estimate the following formula: ； in, is a hyperparameter adjusted according to the noise situation, which can be decomposed into a signal plus a penalty function; the degree of the penalty function is given by to control; Indicates that the minimum value of the entire objective function is returned. , Indicated in The distance perception matrix under the condition, express norm, express Norm.

8. A three-dimensional back-projection imaging device for array interferometric SAR, characterized in that: The three-dimensional back-projection imaging method of array interferometric SAR according to any one of claims 1 to 7 is used to realize three-dimensional imaging of array interferometric SAR, comprising the following modules: The data acquisition module is used to acquire the raw echo data of the array interferometric SAR of the target scene during flight shooting of the UAV; The imaging result acquisition module processes the raw echo data of the array interferometric SAR to obtain multiple groups of two-dimensional imaging results of the array channels; A grid division module establishes a three-dimensional coordinate system, divides the grid of the main antenna, obtains a three-dimensional grid, and filters out grids with a vector amplitude below a set value; An observation vector acquisition module interpolates non-low-value grids to obtain an observation vector matrix corresponding to the three-dimensional grid, and the observation vector matrix is ​​used as a perception matrix; The initial 3D point cloud acquisition module expands the perception matrix to match different angles of different distance units, uses the compressed sensing algorithm for sparse 3D imaging, and obtains the initial 3D point cloud of the target scene; The imaging module processes the initial three-dimensional point cloud to obtain a complete three-dimensional imaging point cloud of the target area.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the three-dimensional backprojection imaging method of array interferometric SAR described in any one of claims 1-7 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the three-dimensional backprojection imaging method of array interferometric SAR described in any one of claims 1 to 7 are implemented.

Images