A slant-range projection algorithm based on real-time SAR imaging system

Abstract

This invention discloses an oblique ground projection algorithm based on a real-time SAR imaging system, belonging to the field of radar imaging processing technology. The algorithm includes the following steps: a) setting the grid resolution; b) constructing a two-dimensional empty matrix; c) obtaining the line-of-sight coordinates; d) calculating the coordinate rotation factor; e) obtaining the cylindrical coordinate system coordinates; f) calculating the conical coordinate position; g) determining whether the conical coordinate position is within a preset coordinate range; h) selecting the corresponding coordinate position data from the oblique range map and storing it in the ground distance coordinate system; i) completing the oblique ground projection. This invention fully utilizes the geometric projection relationship between the oblique plane and the ground plane. Using an FPGA, a two-dimensional matrix corresponding to the ground distance SAR image is constructed with the beam center as the matrix center. Back-projection is performed, and oblique range SAR image data is directly stored in the ground distance coordinate system, avoiding multiple FFT, IFFT, and transpose operations. This significantly reduces the time required for oblique ground projection and also reduces memory resource consumption.

Description

Technical Field

[0001] This invention relates to the field of radar imaging processing technology, and in particular to an oblique ground projection algorithm based on a real-time SAR imaging system. Background Technology

[0002] For oblique-looking SAR imaging systems, due to their structural characteristics, the echo data acquired by the system, after range and azimuth processing, results in oblique-plane SAR images with rotational and stretched geometric distortions compared to ground-plane SAR images, affecting SAR image interpretation and processing. Furthermore, the SAR images obtained after target imaging are oblique-plane images and need to be converted into more intuitive ground-plane SAR images to obtain the target's latitude, longitude, and altitude information, laying the foundation for subsequent target localization, identification, and image processing. To address these two problems, coordinate projection transformation using an oblique-ground projection algorithm can effectively correct geometric distortions and simultaneously convert oblique-plane SAR images to ground-plane SAR images.

[0003] Oblique projection algorithms are widely used in various SAR imaging fields because they are not constrained by platform conditions. For oblique-looking SAR imaging systems, it is based on the oblique projection of traditional SAR images in the azimuth and time domains. Traditional oblique projection algorithms require two separate FFT, IFFT, and transpose operations, which greatly increases time costs and storage resource requirements. For real-time SAR imaging systems, the time and storage resource usage of these operations makes the system unable to meet real-time requirements. Therefore, traditional oblique projection algorithms are not applicable.

[0004] Chinese patent document CN111722226A, published on September 29, 2020, discloses a method for real-time processing of bistatic forward-looking SAR echo data, characterized by the following steps:

[0005] Step S201: Acquire raw echo data from bistatic forward-looking SAR;

[0006] Step S202: Calculate the oblique projection matrix;

[0007] Step S203: Extract distance data;

[0008] Step S204: Range processing, including range FFT; range pulse compression, range migration correction, azimuth FFT, range curvature correction, range secondary pulse compression residual migration correction, and range FFT;

[0009] Step S205: Extract azimuth data;

[0010] Step S206: Azimuth processing, including azimuth IFFT, higher-order polynomial fitting, azimuth NCS matching, azimuth FFT, higher-order phase filtering, azimuth IFFT, and azimuth pulse compression.

[0011] Step S207: Oblique image mapping.

[0012] The method for real-time processing of bistatic forward-looking SAR echo data disclosed in this patent document improves the data processing efficiency of bistatic forward-looking SAR, enhances the real-time data processing performance of the bistatic forward-looking SAR system, and increases computational efficiency. However, it still requires multiple FFT and IFFT operations, increasing the time required for oblique projection and consuming more memory resources. Summary of the Invention

[0013] To overcome the shortcomings of the prior art, this invention provides an oblique ground projection algorithm based on a real-time SAR imaging system. This invention fully utilizes the geometric projection relationship between the oblique plane and the ground plane, and constructs a two-dimensional matrix corresponding to the ground distance SAR image using an FPGA with the beam center as the matrix center. It performs back projection and selects oblique range SAR image data to be directly stored in the ground distance coordinate system, avoiding multiple FFT, IFFT and transpose operations, greatly reducing the time required for oblique ground projection, and also reducing the consumption of memory resources.

[0014] This invention is achieved through the following technical solution:

[0015] An oblique projection algorithm based on a real-time SAR imaging system, characterized by the following steps:

[0016] a. Set the grid resolution according to the resolution requirements of the ground distance map;

[0017] b. Construct a two-dimensional spatial matrix of the ground distance SAR image based on the size and resolution requirements of the ground distance map;

[0018] c. Based on the distance coordinates Projection yields viewing distance coordinates ;

[0019] d. Calculate the coordinate rotation factor;

[0020] e. Adjusting the line-of-sight coordinates using a coordinate rotation factor Rotation yields the corresponding cylindrical coordinates. ;

[0021] f. Calculate the corresponding cone coordinate position. ;

[0022] g. Determine the coordinate position of the cone If the coordinates are within the preset range, proceed to step h; otherwise, return to step c.

[0023] h. Based on the cone coordinate position Select the corresponding coordinate location data from the slant distance map and store it in the ground distance coordinates of the ground distance map. Location;

[0024] i. Repeat steps c through h until the oblique projection is completed.

[0025] In step d, the coordinate rotation factor is calculated using Equation 1;

[0026] Formula 1.

[0027] In step f, the cone coordinate position Calculations are performed using Equation 2;

[0028] Equation 2.

[0029] It also includes simulation and data interaction.

[0030] The simulation refers to using MATLAB to perform oblique projection simulation on a set of airborne measured data.

[0031] The data interaction refers to the data interaction process between the DSP and the FPGA. The eight DSP boards send the slant range map to the DDR space of the two FPGAs via SRIO.

[0032] The two FPGAs are divided into four blocks in the range direction, and two-dimensional matrices of the ground distance SAR image are drawn in each block.

[0033] The two-dimensional matrix data of the ground distance SAR image is stitched together by sending it back to eight DSPs via SRIO.

[0034] The FPGA mentioned in this invention refers to a Field Programmable Gate Array.

[0035] The SAR mentioned in this invention refers to Synthetic Aperture Radar.

[0036] The FFTS mentioned in this invention refers to Fast Fourier Transform.

[0037] The IFFT mentioned in this invention refers to the Inverse Fast Fourier Transform.

[0038] The NCS mentioned in this invention refers to nonlinear frequency modulation scaling.

[0039] The MATLAB mentioned in this invention refers to Matrix Factory.

[0040] The DSP mentioned in this invention refers to Digital Signal Processing.

[0041] The FPGA mentioned in this invention refers to a Field Programmable Gate Array.

[0042] The SRIO mentioned in this invention refers to Serial Fast Input / Output Port.

[0043] The DDR mentioned in this invention refers to Double Data Rate Synchronous Dynamic Random Access Memory.

[0044] The beneficial effects of this invention are mainly reflected in the following aspects:

[0045] 1. This invention fully utilizes the geometric projection relationship between the oblique plane and the ground plane. By using an FPGA to construct a two-dimensional matrix corresponding to the ground distance SAR image with the beam center as the matrix center, back projection is performed and oblique distance SAR image data is selected and directly stored in the ground distance coordinate system. This avoids multiple FFT, IFFT and transpose operations, greatly reducing the time required for oblique ground projection, and also reducing the consumption of memory resources.

[0046] 2. Compared with existing oblique projection methods, this invention enables real-time SAR imaging systems to be more efficient in terms of imaging time and can solve the problem of system resource shortage in terms of storage resource usage.

[0047] 3. The present invention reduces the image output time of the SAR real-time imaging system by 5 times compared with the prior art when acquiring ground distance SAR images. Attached Figure Description

[0048] The present invention will now be further described in detail with reference to the accompanying drawings and specific embodiments:

[0049] Figure 1 A schematic diagram of the target projection from the ground distance coordinate system to the cylindrical coordinate system;

[0050] Figure 2 A schematic diagram of projecting a target from a cylindrical coordinate system onto a conical coordinate system. Detailed Implementation

[0051] Example 1

[0052] See Figure 1 and Figure 2 An oblique projection algorithm based on a real-time SAR imaging system includes the following steps:

[0053] a. Set the grid resolution according to the resolution requirements of the ground distance map;

[0054] b. Construct a two-dimensional spatial matrix of the ground distance SAR image based on the size and resolution requirements of the ground distance map;

[0055] c. Based on the distance coordinates Projection yields viewing distance coordinates ;

[0056] d. Calculate the coordinate rotation factor;

[0057] e. Adjusting the line-of-sight coordinates using a coordinate rotation factor Rotation yields the corresponding cylindrical coordinates. ;

[0058] f. Calculate the corresponding cone coordinate position. ;

[0059] g. Determine the coordinate position of the cone If the coordinates are within the preset range, proceed to step h; otherwise, return to step c.

[0060] h. Based on the cone coordinate position Select the corresponding coordinate location data from the slant distance map and store it in the ground distance coordinates of the ground distance map. Location;

[0061] i. Repeat steps c through h until the oblique projection is completed.

[0062] This embodiment is the most basic implementation method. It makes full use of the geometric projection relationship between the oblique plane and the ground plane. The FPGA constructs a two-dimensional matrix corresponding to the ground distance SAR image with the beam center as the matrix center. Back projection is performed and the oblique distance SAR image data is selected and directly stored in the ground distance coordinate system. This avoids multiple FFT, IFFT and transpose operations, greatly reducing the time required for oblique ground projection, and also reducing the consumption of memory resources.

[0063] Example 2

[0064] See Figure 1 and Figure 2 An oblique projection algorithm based on a real-time SAR imaging system includes the following steps:

[0065] a. Set the grid resolution according to the resolution requirements of the ground distance map;

[0066] b. Construct a two-dimensional spatial matrix of the ground distance SAR image based on the size and resolution requirements of the ground distance map;

[0067] c. Based on the distance coordinates Projection yields viewing distance coordinates ;

[0068] d. Calculate the coordinate rotation factor;

[0069] e. Adjusting the line-of-sight coordinates using a coordinate rotation factor Rotation yields the corresponding cylindrical coordinates. ;

[0070] f. Calculate the corresponding cone coordinate position. ;

[0071] g. Determine the coordinate position of the cone If the coordinates are within the preset range, proceed to step h; otherwise, return to step c.

[0072] h. Based on the cone coordinate position Select the corresponding coordinate location data from the slant distance map and store it in the ground distance coordinates of the ground distance map. Location;

[0073] i. Repeat steps c through h until the oblique projection is completed.

[0074] Furthermore, in step d, the coordinate rotation factor is calculated using Equation 1;

[0075] Formula 1.

[0076] In step f, the coordinate position of the cone is calculated using Equation 2;

[0077] Equation 2.

[0078] This embodiment is a preferred implementation method. Compared with the existing oblique projection method, it can make the real-time SAR imaging system more efficient in terms of imaging time and solve the problem of system resource shortage in terms of storage resource usage.

[0079] Example 3

[0080] See Figure 1 and Figure 2 An oblique projection algorithm based on a real-time SAR imaging system includes the following steps:

[0081] a. Set the grid resolution according to the resolution requirements of the ground distance map;

[0082] b. Construct a two-dimensional spatial matrix of the ground distance SAR image based on the size and resolution requirements of the ground distance map;

[0083] c. Based on the distance coordinates Projection yields viewing distance coordinates ;

[0084] d. Calculate the coordinate rotation factor;

[0085] e. Adjusting the line-of-sight coordinates using a coordinate rotation factor Rotation yields the corresponding cylindrical coordinates. ;

[0086] f. Calculate the corresponding cone coordinate position. ;

[0087] g. Determine the coordinate position of the cone If the coordinates are within the preset range, proceed to step h; otherwise, return to step c.

[0088] h. Based on the cone coordinate position Select the corresponding coordinate location data from the slant distance map and store it in the ground distance coordinates of the ground distance map. Location;

[0089] i. Repeat steps c through h until the oblique projection is completed.

[0090] In step d, the coordinate rotation factor is calculated using Equation 1;

[0091] Formula 1.

[0092] In step f, the coordinate position of the cone is calculated using Equation 2;

[0093] Equation 2.

[0094] Furthermore, this also includes simulation and data interaction.

[0095] The simulation refers to using MATLAB to perform oblique projection simulation on a set of airborne measured data.

[0096] The data interaction refers to the data interaction process between the DSP and the FPGA. The eight DSP boards send the slant range map to the DDR space of the two FPGAs via SRIO.

[0097] Example 4

[0098] See Figure 1 and Figure 2 An oblique projection algorithm based on a real-time SAR imaging system includes the following steps:

[0099] a. Set the grid resolution according to the resolution requirements of the ground distance map;

[0100] b. Construct a two-dimensional spatial matrix of the ground distance SAR image based on the size and resolution requirements of the ground distance map;

[0101] c. Based on the distance coordinates Projection yields viewing distance coordinates ;

[0102] d. Calculate the coordinate rotation factor;

[0103] e. Adjusting the line-of-sight coordinates using a coordinate rotation factor Rotation yields the corresponding cylindrical coordinates. ;

[0104] f. Calculate the corresponding cone coordinate position. ;

[0105] g. Determine the coordinate position of the cone If the coordinates are within the preset range, proceed to step h; otherwise, return to step c.

[0106] h. Based on the cone coordinate position Select the corresponding coordinate location data from the slant distance map and store it in the ground distance coordinates of the ground distance map. Location;

[0107] i. Repeat steps c through h until the oblique projection is completed.

[0108] In step d, the coordinate rotation factor is calculated using Equation 1;

[0109] Formula 1.

[0110] In step f, the coordinate position of the cone is calculated using Equation 2;

[0111] Equation 2.

[0112] It also includes simulation and data interaction.

[0113] The simulation refers to using MATLAB to perform oblique projection simulation on a set of airborne measured data.

[0114] The data interaction refers to the data interaction process between the DSP and the FPGA. The eight DSP boards send the slant range map to the DDR space of the two FPGAs via SRIO.

[0115] The two FPGAs are divided into four blocks in the range direction, and two-dimensional matrices of the ground distance SAR image are drawn in each block.

[0116] The two-dimensional matrix data of the ground distance SAR image is stitched together by sending it back to eight DSPs via SRIO.

[0117] This embodiment is the best implementation method. When the SAR real-time imaging system acquires ground-distance SAR images, the image output time is reduced by 5 times compared with the prior art.

[0118] The following section uses FPGA to implement the oblique ground projection algorithm of this invention and the traditional oblique ground projection algorithm, and compares and analyzes the processing time and FPGA storage resources occupied by the same data source.

[0119] The FPGA storage resources required for the experiment are shown in Table 1:

[0120] Table 1

[0121]

[0122] The time required for the two sets of experiments is shown in Table 2:

[0123] Table 2

[0124]

[0125] As shown in Table 1, the present invention is superior to the traditional oblique projection algorithm in terms of memory resource requirements.

[0126] As shown in Table 2, the traditional oblique ground projection algorithm requires zero-padding of the data to meet the requirements of subsequent FFT operations. Therefore, the ground distance image size output by the traditional oblique ground projection algorithm after data processing is [32768, 32768], which needs to be further truncated to [8192, 16384]. Compared with the traditional oblique ground projection algorithm, the present invention is not constrained by the above situation and can directly obtain a ground distance SAR image that meets the requirements for subsequent processing, thus achieving higher processing efficiency than the traditional oblique ground projection algorithm.

[0127] In conclusion, the experiments have verified the effectiveness and reliability of the present invention.

Claims

1. An oblique ground projection algorithm based on a real-time SAR imaging system, characterized in that, Includes the following steps: a. Set the grid resolution according to the resolution requirements of the ground distance map; b. Construct a two-dimensional spatial matrix of the ground distance SAR image based on the size and resolution requirements of the ground distance map; c. Based on the distance coordinates Projection yields view distance coordinates ; d. Calculate the coordinate rotation factor; e. Adjusting the line-of-sight coordinates using a coordinate rotation factor Rotation yields the corresponding cylindrical coordinates. ; f. Calculate the corresponding cone coordinate position. ; g. Determine the coordinate position of the cone If the coordinates are within the preset range, proceed to step h; otherwise, return to step c. h. Based on the cone coordinate position Select the corresponding coordinate location data from the slant distance map and store it in the ground distance coordinates of the ground distance map. Location; i. Repeat steps c to h until the oblique projection is completed; In step d, the coordinate rotation factor is calculated using Equation 1; ； In step f, the cone coordinate position Calculations are performed using Equation 2; 。 2. The oblique ground projection algorithm based on a real-time SAR imaging system according to claim 1, characterized in that: It also includes simulation and data interaction.

3. The oblique ground projection algorithm based on a real-time SAR imaging system according to claim 2, characterized in that: The simulation refers to using MATLAB to perform oblique projection simulation on a set of airborne measured data.

4. The oblique ground projection algorithm based on a real-time SAR imaging system according to claim 2, characterized in that: The data interaction refers to the data interaction process between the DSP and the FPGA. The eight DSP boards send the slant range map to the DDR space of the two FPGAs via SRIO.

5. The oblique ground projection algorithm based on a real-time SAR imaging system according to claim 4, characterized in that: The two FPGAs are divided into four blocks in the range direction, and two-dimensional matrices of the ground distance SAR image are drawn in each block.

6. The oblique ground projection algorithm based on a real-time SAR imaging system according to claim 4, characterized in that: The two-dimensional matrix data of the ground distance SAR image is stitched together by sending it back to eight DSPs via SRIO.

Images