A multi-angle imaging method based on MIMO-SAR

Abstract

The application discloses a multi-angle imaging method based on MIMO-SAR, which comprises the following steps: step 1, obtaining echo signals of a point target at multiple different angles in MIMO-SAR; step 2, respectively performing azimuth deblurring on the echo signals at the angles to obtain azimuth non-blurred signals at the angles; step 3, performing azimuth focusing on the azimuth non-blurred signals at the angles; and step 4, using a geographic information scale invariant feature transformation method to fuse the obtained focusing images at the angles to obtain a SAR image without overlapping. The application can obtain an imaging mode of SAR images at different angles under a low pulse repetition frequency.

Description

Technical Field

[0001] This invention relates to the field of radar imaging methods, specifically a multi-angle imaging method based on MIMO-SAR. Background Technology

[0002] Multiple-Input Multiple-Output Synthetic Aperture Radar (MIMO-SAR) is an active Earth observation system capable of achieving high-resolution wide-swath (HRWS) imaging. Its scene imaging quality is comparable to optical images used for similar purposes, and it features all-weather, all-time, wide-swath, and long-range capabilities. Resolution refers to the azimuth and range resolution of the SAR image; "high" refers to the achievable resolution (azimuth and range resolutions can reach 1 meter or even higher); "swath" refers to the azimuth and range bands in the SAR image; and "wide" refers to the achievable size of the swath (azimuth and range bands can reach several kilometers, tens of kilometers, or even wider). However, due to limitations in minimum antenna area, high resolution and wide HRWS in spaceborne SAR present a trade-off, making it impossible to obtain high-resolution images with a wide HRWS for multi-angle imaging.

[0003] Meanwhile, Multi-Angle Synthetic Aperture Radar (MA-SAR) imaging refers to a space-based microwave imaging radar system that observes the same target or scene from different azimuth angles. Compared to conventional single-angle synthetic aperture radar, MA-SAR has the advantage of spatial diversity. By increasing the number of data samples processed by the system, it can effectively expand the spatial spectral support region of the detected target, avoiding target occlusion and overlap problems that exist in conventional synthetic aperture radar imaging detection, and improving the ability to classify and identify targets. However, the implementation methods of multi-angle imaging systems are all mounted on synthetic aperture radars with high pulse repetition frequencies, and none of them can obtain multi-angle wide mapping swathe imaging results. Summary of the Invention

[0004] This invention provides a multi-angle imaging method based on MIMO-SAR to solve the problem that existing technologies such as MIMO-SAR and MA-SAR cannot achieve both wide mapping band and high resolution at low pulse repetition frequency when used for multi-angle imaging.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A multi-angle imaging method based on MIMO-SAR includes the following steps:

[0007] Step 1: In MIMO-SAR (Multiple Incoming and Multiple Outgoing Synthetic Aperture Radar), acquire echo signals from point targets at multiple different angles;

[0008] Step 2: De-ambiguously process the echo signals obtained in Step 1 at each angle to obtain unambiguous azimuth signals at each angle.

[0009] Step 3: Focus the azimuth signal without ambiguity for each angle to obtain the azimuth signal for each angle. - The process of focusing the image in the domain is as follows:

[0010] Step (3.1): Transform the unambiguous azimuth signal at each angle into a dual-frequency domain echo signal. The transformation results are shown below: , ,

[0011] In the formula: This represents the echo signal after transformation into the dual-frequency domain; Indicates the range frequency; Indicates azimuth Doppler; This indicates the shortest distance from a point target in the scene to the radar. Indicates slow time; A window function representing the distance direction; A window function representing azimuth; Indicates the complex unit; Represents the echo phase function; Indicates radar speed; This represents the maximum Doppler echo of a point target; Indicates the Doppler phase center; Indicates frequency modulation; X represents the coordinates of the point target. The x-coordinate;

[0012] Step (3.2): Construct a quadratic range-frequency compensation term to compensate for the quadratic range-frequency term in the dual-frequency domain echo signal obtained in step (3.1). The constructed quadratic range-frequency compensation term is shown in the following formula: ],

[0013] In the formula: This represents the constructor expression for the quadratic distance-frequency term used to compensate for the echo signal;

[0014] The dual-frequency domain echo signal of the compensated distance-frequency quadratic term is: ;

[0015] Step (3.3): The dual-frequency domain echo signal after compensating for the quadratic term of the distance frequency in step (3.2) is transformed into a wavenumber domain echo signal. The transformation result is shown in the following formula:

[0016] ,

[0017] , , ,

[0018] In the formula: This represents the expression for the echo signal after transforming the original dual-frequency domain echo signal into the wavenumber domain; This represents a redefinition expression in the wavenumber domain; This represents a redefinition expression in the wavenumber domain; The azimuth window function representing the wavenumber domain echo signal; The range window function representing the wavenumber domain echo signal; This represents a redefinition expression in the wavenumber domain; Indicates the carrier frequency; Represents the speed of light;

[0019] Step (3.4): Construct a uniform compression compensation term to compensate for the uniform compression term in the wavenumber domain echo signal obtained in step (3.3). The constructed uniform compression compensation term is shown in the following formula:

[0020] ,

[0021] In the formula: The constructor expression for the uniform compression term used to compensate for the echo signal; Let represent the shortest distance from the scene center to the radar; then the wavenumber domain echo signal result after compensating for the consistent compression term is shown in the following equation:

[0022]

[0023] ;

[0024] Step (3.5): Using the improved range migration algorithm, the wavenumber domain echo signal after compensating for the uniform compression term in step (3.4) is interpolated to complete the range migration correction, secondary range compression, and azimuth compression of the wavenumber domain echo signal. The improved range migration algorithm is shown in the following formula:

[0025] ,

[0026] In the formula: Represents the interpolation function; This represents a redefinition expression in the wavenumber domain;

[0027] The wavenumber domain echo signal result after interpolation is shown in the following formula:

[0028]

[0029]

[0030] Step (3.6): Perform range-direction IFFT processing on the wavenumber domain echo signal after interpolation in step (3.5), and the result is shown in the following formula:

[0031] ,

[0032] Then, a residual phase compensation term is constructed to compensate for the residual phase in the wavenumber domain echo signal after range-direction IFFT processing due to interpolation. The residual phase compensation term is shown in the following formula:

[0033] ,

[0034] In the formula: The constructor expression for the residual phase compensation term used to compensate for the echo signal; Representing fast time; the wavenumber domain echo signal after compensating for residual phase is shown in the following equation:

[0035] ;

[0036] Step (3.7): Perform azimuth IFFT processing on the wavenumber domain echo signal after residual phase compensation in step (3.6). The processing result is shown below:

[0037] ,

[0038] , ,

[0039] In the formula: This represents a redefinition expression in the wavenumber domain; Indicates the distance traveled by the radar; Indicates location and time.

[0040] Then, after performing a Decirp operation on the wavenumber domain echo signal processed by IFFT, correction is performed to make the Doppler center of the wavenumber domain echo signal after the Decirp operation zero.

[0041] Step (3.8): Perform a direction-position FFT operation on the wavenumber domain echo signal obtained after correction in step (3.7) to obtain the azimuth of each angle. - Domain-focused image;

[0042] Step 4: Take each angle obtained in Step 3. - The domain-focused images are compensated for azimuth differences, deviations, and residual errors to obtain images after compensation for each angle. Then, the images after compensation for each angle are fused using the geographic information scale-invariant feature transformation method to obtain the final image.

[0043] In the further step 1, when the multiple-transmitter multiple-receiver synthetic aperture radar (MIMO-SAR) is mounted on a satellite, the echo signal of the point target received by the MIMO-SAR is divided into several imaging segments, and each segment has a different imaging angle, thereby obtaining echo signals of the point target at multiple different angles.

[0044] In further step 1, when the multi-transmitter multi-receiver synthetic aperture radar (MIMO-SAR) is mounted on an aircraft, it obtains echo signals from multiple different angles of a point target during a single flight based on digital multibeamforming technology.

[0045] Further, step 2 is as follows: First, range compression is performed on the echo signal obtained in step 1 for each angle; then, azimuth FFT processing is performed on the range-compressed echo signal; finally, based on the spatial degrees of freedom of the echo signal after azimuth FFT processing in the azimuth direction, a spatial filtering weight vector function is constructed, and the spatial filtering weight vector function is used to filter the echo signal after azimuth FFT processing to obtain an unambiguous azimuth signal.

[0046] Furthermore, in step 4, during azimuth difference compensation, the frontal side view angle is used as the reference image, and the imaging results from other angles are compensated as the reference image. That is, it is assumed that the angle between the two lines of sight is... The scaling factor is And the rotation center point of the imaging scene is Then, any point in the image to be compensated for azimuth difference... Transformed into the reference image This is the image after azimuth difference compensation, as shown in the following formula:

[0047] .

[0048] In the further step 4, the deviation compensation formula is as follows:

[0049] ,

[0050] In the formula: This represents the function expression used to compensate for the deviation term; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate.

[0051] In the further step 4, the residual error compensation is as follows:

[0052] ,

[0053] ,

[0054] In the formula: This represents the function expression used to compensate for the deviation term; This represents the function expression used to compensate for the deviation term; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate.

[0055] This invention can obtain imaging modes of SAR images at different angles under low pulse repetition frequencies. Compared with existing technologies, the advantages of this invention are:

[0056] (1) This invention proposes a multi-angle imaging method based on MIMO-SAR, which can achieve high-resolution images at different angles with low pulse repetition frequency, and fully overcomes the contradiction between traditional high pulse repetition frequency and mapping zone.

[0057] (2) This invention improves the traditional Inter-Range Distance Migration (IRMA) algorithm. By improving the interpolation formula of the IRMA algorithm, the spectrum utilization rate can be effectively improved, and finally complete spectrum information can be obtained.

[0058] (3) This invention improves the geographic information scale-invariant feature transformation algorithm. Traditional geographic information scale-invariant feature transformation requires three steps: azimuth difference, deviation elimination, registration, and fusion. Among them, after deviation elimination, there may be residual errors before registration that are not compensated. Ultimately, this will lead to a deterioration in the accuracy of registration and the effect of fusion. This invention eliminates residual errors by using an improved geographic information scale-invariant feature transformation, thereby improving the accuracy of registration and the effect of fusion.

[0059] (4) In this invention, spatial filtering technology is first used to remove the azimuth ambiguity of the echo signal, then the distance migration algorithm is used for imaging processing, and finally the improved geographic information scale invariant feature transformation algorithm is used to fuse the imaging results from various angles, thereby obtaining a target without overlap. Attached Figure Description

[0060] Figure 1 The image shown is the working mode of airborne multi-beam multi-angle MIMO-SAR imaging.

[0061] Figure 2 The image shows the working mode of spaceborne multi-angle MIMO-SAR imaging.

[0062] Figure 3 Simulation results before and after orientation deblurring, where: (a) is the simulation result before orientation deblurring; (b), (c), and (d) are the simulation results after orientation deblurring.

[0063] Figure 4 Band synthesis and point target imaging height maps, where: (a) simulation results after band synthesis; (b) height map after point target imaging.

[0064] Figure 5 Azimuth sampling map and range sampling map of point target imaging, wherein: (a) azimuth sampling map of point target imaging; (b) range sampling map of point target imaging.

[0065] Figure 6 The point target imaging results under the CS algorithm at different angles are as follows: (a) point target imaging results at [-1°, 1°]; (b) point target imaging results at [-2°, 2°]; (c) point target imaging results at [-3°, 3°]; (d) point target imaging results at [-4°, 4°]; (e) point target imaging results at [-5°, 5°]; and (f) point target imaging results at [-6°, 6°].

[0066] Figure 7Imaging results at different angles using the IRMA algorithm, including: (a) point target imaging results at [-1°, 1°]; (b) point target imaging results at [-2°, 2°]; (c) point target imaging results at [-3°, 3°]; (d) point target imaging results at [-4°, 4°]; (e) point target imaging results at [-5°, 5°]; and (f) point target imaging results at [-6°, 6°].

[0067] Figure 8 Imaging results of point targets in the range [-6°, 6°] using CS, RMA, and IRMA algorithms, including: (a) point target imaging results using the CS algorithm in the range [-6°, 6°]; (b) point target imaging results using the IRMA algorithm in the range [-6°, 6°]; (c) a comparison of range sampling between the CS and IRMA algorithms in the range [-6°, 6°]; (d) point target imaging results using the RMA algorithm in the range [-6°, 6°]; (e) point target imaging results using the IRMA algorithm in the range [-6°, 6°]; and (f) a comparison of range sampling between the RMA and IRMA algorithms in the range [-6°, 6°].

[0068] Figure 9 Imaging results of point targets under CS, RMA, and IRMA algorithms, including: (a) Simulation results of point target spectrum after range compression under CS algorithm; (b) Imaging results of point targets under CS algorithm; (c) Height map of point target imaging under CS algorithm; (d) Simulation results of point target spectrum after range compression under RMA algorithm; (e) Imaging results of point targets under RMA algorithm; (f) Height map of point target imaging under RMA algorithm; (g) Simulation results of point target spectrum after range compression under IRMA algorithm; (h) Imaging results of point targets under IRMA algorithm; (i) Height map of point target imaging under IRMA algorithm.

[0069] Figure 10 The fusion results after imaging point targets are: (a) the result after imaging point target at angle 1; (b) the result after imaging point target at angle 2; and (c) the fusion result after imaging point target at angles 1 and 2.

[0070] Figure 11 The fusion results of ground target imaging, including: (a) the result of ground target imaging at angle 1; (b) the result of ground target imaging at angle 2; and (c) the fusion result of ground target imaging at angles 1 and 2.

[0071] Figure 12Image fusion results of the scale-invariant feature transformation algorithm and the improved scale-invariant feature transformation algorithm, wherein: (a) the image fusion result of the scale-invariant feature transformation algorithm; (b) the image fusion result of the improved scale-invariant feature transformation algorithm. Detailed Implementation

[0072] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0073] A multi-angle imaging method based on MIMO-SAR includes the following steps:

[0074] Step 1: In MIMO-SAR (Multiple Incoming and Multiple Outgoing Synthetic Aperture Radar), acquire echo signals from multiple different angles of a point target.

[0075] like Figure 1 As shown, when a multi-transmitter multi-receiver synthetic aperture radar (MIMO-SAR) is mounted on an aircraft, it can obtain echo signals from point targets at multiple different angles during a single flight based on digital multibeamforming technology.

[0076] like Figure 2 As shown, when multiple-transmitter multiple-receiver synthetic aperture radar (MIMO-SAR) is mounted on a satellite, the echo signal of the point target received by the MIMO-SAR is divided into several imaging segments, and each segment has a different imaging angle, thus obtaining echo signals of the point target from multiple different angles.

[0077] Step 2: De-ambiguously process the echo signals obtained in Step 1 at each angle to obtain unambiguous azimuth signals at each angle. The process is as follows:

[0078] First, range compression is performed on the echo signals obtained in step 1 at each angle. Then, azimuth FFT processing is performed on the range-compressed echo signals. Finally, based on the spatial degrees of freedom of the azimuth-directed echo signals after azimuth FFT processing, a spatial filtering weight vector function W(m) is constructed. This spatial filtering weight vector function W(m) is used to filter the azimuth-directed FFT-processed echo signals to obtain an unambiguous azimuth signal. The constructed spatial filtering weight vector function W(m) is shown in the following equation:

[0079] ,

[0080] In the formula: Indicates the pseudo-inverse of a matrix; Let V be the number of azimuth channel array vector matrices (number of sub-aperture rows of the area array antenna). for The pseudo-inverse of the matrix; .

[0081] Step 3: Focus the azimuth signal without ambiguity for each angle to obtain the azimuth signal for each angle. - The process of focusing the image in the domain is as follows:

[0082] Step (3.1): Transform the unambiguous azimuth signal at each angle into a dual-frequency domain echo signal. The transformation results are shown below:

[0083] ,

[0084] ,

[0085] In the formula: This represents the echo signal after transformation into the dual-frequency domain; Indicates the range frequency; Indicates azimuth Doppler; This indicates the shortest distance from a point target in the scene to the radar. Indicates slow time; A window function representing the distance direction; A window function representing azimuth; Indicates the complex unit; Represents the echo phase function; Indicates radar speed; This represents the maximum Doppler echo of a point target; Indicates the Doppler phase center; Indicates frequency modulation; X represents the coordinates of the point target. The x-coordinate.

[0086] Step (3.2): Construct a quadratic range-frequency compensation term to compensate for the quadratic range-frequency term in the dual-frequency domain echo signal obtained in step (3.1). The constructed quadratic range-frequency compensation term is shown in the following formula:

[0087] ],

[0088] In the formula: This represents the constructor expression for the quadratic distance-frequency term used to compensate for the echo signal.

[0089] The dual-frequency domain echo signal of the compensated distance-frequency quadratic term is:

[0090] .

[0091] Step (3.3): The dual-frequency domain echo signal after compensating for the quadratic term of the distance frequency in step (3.2) is transformed into a wavenumber domain echo signal. The transformation result is shown in the following formula:

[0092] ,

[0093] , , ,

[0094] In the formula: This represents the expression for the echo signal after transforming the original dual-frequency domain echo signal into the wavenumber domain; Represents the redefinition expression for the wavenumber domain; Represents the redefinition expression for the wavenumber domain; The azimuth window function representing the wavenumber domain echo signal; The range window function representing the wavenumber domain echo signal; This represents a redefinition expression in the wavenumber domain; Indicates the carrier frequency; It represents the speed of light.

[0095] Step (3.4): Construct a uniform compression compensation term to compensate for the uniform compression term in the wavenumber domain echo signal obtained in step (3.3). The constructed uniform compression compensation term is shown in the following formula:

[0096] ,

[0097] In the formula: The constructor expression for the uniform compression term used to compensate for the echo signal; This indicates the shortest distance from the center of the scene to the radar.

[0098] The wavenumber domain echo signal result after compensation for the uniform compression term is shown in the following equation:

[0099]

[0100] ;

[0101] Step (3.5): Using the improved range migration algorithm, the wavenumber domain echo signal after compensating for the uniform compression term in step (3.4) is interpolated to complete the range migration correction, secondary range compression, and azimuth compression of the wavenumber domain echo signal. The improved range migration algorithm is shown in the following formula:

[0102] ,

[0103] In the formula: Let represent the interpolation function. The wavenumber domain echo signal result after interpolation is shown in the following equation:

[0104]

[0105]

[0106] Step (3.6): Perform range-direction IFFT processing on the wavenumber domain echo signal after interpolation in step (3.5), and the result is shown in the following formula:

[0107] ,

[0108] Then, a residual phase compensation term is constructed to compensate for the residual phase in the wavenumber domain echo signal after range-direction IFFT processing due to interpolation. The residual phase compensation term is shown in the following formula:

[0109] ,

[0110] In the formula: The constructor expression for the residual phase compensation term used to compensate for the echo signal; Indicates a fast time; This represents a redefinition expression in the wavenumber domain.

[0111] The wavenumber domain echo signal after compensating for the residual phase is shown in the following equation:

[0112] .

[0113] Step (3.7): Perform azimuth IFFT processing on the wavenumber domain echo signal after residual phase compensation in step (3.6). The processing result is shown below:

[0114] ,

[0115] , ,

[0116] In the formula: This represents a redefinition expression in the wavenumber domain; Indicates the distance traveled by the radar; Indicates location and time.

[0117] Then, after performing a Decirp operation on the wavenumber domain echo signal processed by IFFT in the azimuth direction, a correction is performed to make the Doppler center of the wavenumber domain echo signal after the Decirp operation zero.

[0118] Step (3.8): Perform a direction-position FFT operation on the wavenumber domain echo signal obtained after correction in step (3.7) to obtain the azimuth of each angle. - Domain-focused image.

[0119] Step 4: Take each angle obtained in Step 3. - The domain-focused image is subjected to azimuth difference compensation, deviation compensation, and residual error compensation to obtain the image after compensation for each angle.

[0120] In specific azimuth difference compensation, the frontal side view angle is used as the reference image, and the imaging results from other angles are compensated as the reference image. That is, it is assumed that the angle between the two lines of sight is... The scaling factor is And the rotation center point of the imaging scene is Then, any point in the image to be compensated for azimuth difference... Transformed into the reference image This is the image after azimuth difference compensation, as shown in the following formula:

[0121] .

[0122] The deviation compensation formula is shown below:

[0123] ,

[0124] In the formula: This represents the function expression used to compensate for the deviation term; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate.

[0125] The residual error compensation is shown in the following formula:

[0126] ,

[0127] ,

[0128] In the formula: This represents the function expression used to compensate for the deviation term; This represents the function expression used to compensate for the deviation term; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate.

[0129] Finally, the images compensated from each angle are fused using the geographic information scale-invariant feature transformation method to obtain the final image.

[0130] The effectiveness of the present invention will be further illustrated by the target simulation experiment below.

[0131] Simulation experiment:

[0132] (1) Simulation conditions:

[0133] Simulation experiments were used to verify the imaging performance and fusion effect of the imaging mode presented in this paper, as well as the effectiveness of the improved distance migration algorithm and geographic information scale-invariant feature transformation. The simulation experiments employed... Figure 2 The satellite-borne MIMO-SAR multi-angle imaging mode is shown. Simulation parameters are shown in Table 1.

[0134] Table 1. Multi-angle simulation parameters of spaceborne MIMO-SAR

[0135]

[0136] (2) Simulation content:

[0137] Figure 3 (a) Doppler spectrum of the echo before azimuth ambiguity is removed. Figure 3 (b), (c), and (d) are the Doppler spectra of the echoes after removing azimuth ambiguity. Figure 4 (a) is the bandwidth synthesis result after removing orientation ambiguity. Figure 4 (b) is the height map after the point target is imaged. Figure 5 (a) and (b) are the azimuth and range sampling maps of the point target imaging, respectively, with the first sidelobe around -13.40 dB in both cases. Therefore, the simulation results show that the imaging effect of MIMO-SAR combined with multi-angle imaging is good, further verifying the effectiveness of the multi-angle MIMO-SAR imaging system proposed in this paper.

[0138] When the imaging angle reaches a certain level, an excessively large imaging angle will cause varying degrees of distortion in the side lobes of the imaging result. Simulation experiments define the error caused by this distortion as the imaging side lobe distortion error, as follows:

[0139]

[0140] In the formula: The functional expression representing the definition of distortion error; This represents the integral sidelobe ratio in the range direction of the point target imaging result; This represents the integral sidelobe ratio in the azimuth direction of the point target imaging result.

[0141] Table 2. Integral sidelobe ratio and peak sidelobe ratio within different imaging angle ranges.

[0142]

[0143] Er represents the side lobe distortion error, and DF represents the occurrence of defocusing.

[0144] Figure 6 Table 2 presents the multi-angle imaging results of spaceborne MIMO-SAR at different angles. Figure 6 The integrated sidelobe ratio (ISLR) and peak sidelobe ratio (PSLR) of the midpoint target at different angles. From Figure 6 As shown in Table 2, the sidelobe distortion of the point target imaging results becomes more severe with different angles. Simulation results indicate that traditional imaging algorithms cause sidelobe distortion when the illumination time is long or the imaging angle is large. To address this problem, an IRMA algorithm is proposed. Figure 7 This is a multi-angle imaging result of spaceborne MIMO-SAR at different angles.

[0145] Figure 8 Comparison results of CS, RMA and IRMA simulation experiments. Figure 8 (a) shows the imaging results of CS in the range of [-6°, 6°]. Figure 8 (d) shows the imaging results of RMA in the range of [-6°, 6°]. Figure 8 (b) and (e) show the imaging results of IRMA in the range of [-6°, 6°]. Figure 8(c) and (f) show the comparison results of the range sampling maps of the CS and IRMA algorithms and the RMA and IRMA algorithms, respectively. Table 3 analyzes the integral sidelobe ratio (ISLR) and peak sidelobe ratio (PSLR) of the imaging results of the three algorithms. To address the problems of the CS algorithm's inability to focus on large-angle imaging and the spectral information loss in the RMA algorithm, an improved RMA algorithm is proposed. This algorithm uses the tangent (i.e., the tangent value) of each arc to correct the distorted spectrum and designs the imaging process in the wavenumber domain, which can effectively improve the spectral utilization in large-angle SAR imaging. The simulation results show that the proposed algorithm can effectively solve the problems of sidelobe distortion and low spectral utilization.

[0146] Table 3 Comparison results of three imaging algorithms

[0147]

[0148] Er represents the side lobe distortion error, and DF represents the occurrence of defocusing.

[0149] After verifying that the improved range migration algorithm is suitable for spaceborne MIMO-SAR multi-angle imaging, the airborne multi-beam multi-angle MIMO-SAR imaging mode was validated. When the imaging angle of the airborne multi-beam multi-angle MIMO-SAR imaging mode is greater than 20°, the excessively large imaging angle will cause varying degrees of defocusing in the side lobes of the imaging results. The simulation experiment used... Figure 1 The airborne MIMO-SAR multi-angle imaging mode is shown in Table 4. Simulation parameters for airborne MIMO-SAR are shown in Table 4. IRMA utilizes the tangent of each arc to correct distorted spectra and designs the imaging process in the wavenumber domain, which can effectively improve spectral utilization in large-angle SAR imaging.

[0150] Table 4. Simulation parameters of airborne multibeam MIMO-SAR from multiple angles

[0151]

[0152] Figure 9 (a) shows the simulation results of the point target spectrum after distance compression using the CS algorithm. Figure 9 (b) shows the imaging results of the point target under the CS algorithm. Figure 9 (c) is the height map of the point target image under the CS algorithm. Figure 13 (d) is the simulation result of the point target spectrum after distance compression under the RMA algorithm. Figure 9 (e) shows the imaging results of the point target under the RMA algorithm. Figure 9 (f) is the height map of the point target image under the RMA algorithm. Figure 9 (g) shows the simulation results of the point target spectrum after distance compression using the IRMA algorithm. Figure 9(h) represents the imaging results of the point target using the IRMA algorithm. Figure 9 (i) shows the height map of the point target image under the IRMA algorithm. Table 5 compares the imaging results of the three algorithms in the range direction. The results show that both CS and RMA exhibit defocusing in the imaging results, while IRMA demonstrates good imaging performance. Simulation results indicate that IRMA can be effectively applied to MIMO-SAR multi-angle imaging modes.

[0153] Table 5 Comparison results of three imaging algorithms

[0154]

[0155] DF indicates that defocusing has occurred.

[0156] After verifying the imaging performance of the imaging mode presented in this paper and the applicability of the improved range migration algorithm to both spaceborne MIMO-SAR multi-angle imaging and airborne multi-beam multi-angle MIMO-SAR imaging modes, the improved geographic information scale-invariant feature transformation algorithm was validated. Taking the spaceborne multi-angle MIMO-SAR imaging mode as an example, the improved geographic information scale-invariant feature transformation was used to perform a fusion experiment on the imaging results. Figure 10 (a) shows the imaging result of a single target at angle 1. Figure 10 (b) shows the imaging results of a single target at angle 2. Figure 10 (c) shows the imaging fusion results for a single target. It can be seen that the first sidelobe in both the range and azimuth directions of the fused point target is approximately -13.40 dB, indicating a good fusion effect. Figure 11 (a) shows the imaging result of the surface target at angle 1. Figure 11 (b) shows the imaging result of the surface target at angle 2. Figure 11 (c) shows the imaging fusion result for area targets. (From...) Figure 10 and Figure 11 This indicates that, on the one hand, multi-angle SAR can provide more details and features than traditional single-view SAR systems, and on the other hand, more target details can be obtained in fused images, which can provide more information for SAR target recognition applications.

[0157] Table 6. Distance resolution and signal-to-noise ratio of the two fusion algorithms.

[0158]

[0159] Figure 12 (a) and (b) show the imaging fusion results of the Geographic Information Scale-Invariant Feature Transform (GIS) algorithm and the improved GIS Scale-Invariant Feature Transform (GIS) algorithm, respectively. Table 6 shows the range resolution and signal-to-noise ratio after fusion by the two algorithms. Figure 12A comparison with Table 6 shows that the improved geographic information scale-invariant feature transformation method outperforms the original method in terms of detail fusion. Simulation experiments verified the effectiveness of the improved geographic information scale-invariant feature transformation method.

[0160] The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings. These embodiments are merely descriptions of preferred embodiments and are not intended to limit the scope or concept of the invention. The specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. Such combinations, as long as they do not violate the spirit of the present invention, should also be considered as part of this disclosure. To avoid unnecessary repetition, the present invention will not further describe the various possible combinations.

[0161] This invention is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this invention and without departing from the design idea of ​​this invention, all modifications and improvements made by those skilled in the art to the technical solutions of this invention should fall within the protection scope of this invention. The technical content for which protection is sought in this invention has been fully described in the claims.

Claims

1. A method for multi-angle imaging based on MIMO-SAR, characterized in that, Includes the following steps: Step 1: In MIMO-SAR (Multiple Incoming and Multiple Outgoing Synthetic Aperture Radar), acquire echo signals from point targets at multiple different angles; Step 2: De-ambiguously process the echo signals obtained in Step 1 at each angle to obtain unambiguous azimuth signals at each angle. Step 3, azimuth focusing on the azimuth unambiguous signals of each angle to obtain the azimuth focused image of each angle - domain focused image, Step 4: Take each angle obtained in Step 3. - The domain-focused images are subjected to azimuth difference compensation, deviation compensation, and residual error compensation to obtain images after each angle compensation; then, the images after each angle compensation are fused using the geographic information scale-invariant feature transformation method to obtain the final image. In the azimuth angle difference compensation in Step 4, the normal side view angle is taken as a reference image, and imaging results of other angles are compensated to the reference image, that is, assuming that the angle between the two sight lines is , the scale factor is , and the rotation center point of the imaging scene is , then any point in the image to be compensated for the azimuth angle difference is transformed into in the reference image, that is, the image after the azimuth angle difference compensation, as shown in the following formula: ； In step 4, the deviation compensation formula is as follows: ， In the formula: This represents the function expression used to compensate for the deviation term; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate; In step 4, the residual error compensation is as follows: ， ， In the formula: This represents the function expression used to compensate for the deviation term; This represents the function expression used to compensate for the deviation term; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate; Represents the coordinates of a point in the reference image. The x-coordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The x-coordinate; Represents the coordinates of a point in the reference image. The ordinate; Represents the coordinates of any point in the image to be compensated for azimuth difference. The ordinate.

2. The multi-angle imaging method based on MIMO-SAR according to claim 1, characterized in that, The specific process of step 3 is as follows: Step (3.1): Transform the unambiguous azimuth signal at each angle into a dual-frequency domain echo signal. The transformation results are shown below: ， ， In the formula: This represents the echo signal after transformation into the dual-frequency domain; Indicates the range frequency; Indicates azimuth Doppler; This indicates the shortest distance from a point target in the scene to the radar. Indicates slow time; A window function representing the distance direction; A window function representing azimuth; Indicates the complex unit; Represents the echo phase function; Indicates radar speed; This represents the maximum Doppler echo of a point target; Indicates the Doppler phase center; Indicates frequency modulation; X represents the coordinates of the point target. The x-coordinate; Step (3.2): Construct a quadratic range-frequency compensation term to compensate for the quadratic range-frequency term in the dual-frequency domain echo signal obtained in step (3.1). The constructed quadratic range-frequency compensation term is shown in the following formula: ]， In the formula: This represents the constructor expression for the quadratic distance-frequency term used to compensate for the echo signal; The dual-frequency domain echo signal of the compensated distance-frequency quadratic term is: ； Step (3.3): The dual-frequency domain echo signal after compensating for the quadratic term of the distance frequency in step (3.2) is transformed into a wavenumber domain echo signal. The transformation result is shown in the following formula: ， ， ， ， In the formula: This represents the expression for the echo signal after transforming the original dual-frequency domain echo signal into the wavenumber domain; This represents a redefinition expression in the wavenumber domain; This represents a redefinition expression in the wavenumber domain; The range window function representing the wavenumber domain echo signal; The range window function representing the wavenumber domain echo signal; This represents a redefinition expression in the wavenumber domain; Indicates the carrier frequency; Represents the speed of light; Step (3.4): Construct a uniform compression compensation term to compensate for the uniform compression term in the wavenumber domain echo signal obtained in step (3.3). The constructed uniform compression compensation term is shown in the following formula: ， In the formula: The constructor expression for the uniform compression term used to compensate for the echo signal; Indicates the shortest distance from the center of the scene to the radar; The wavenumber domain echo signal result after compensation for the uniform compression term is shown in the following equation: ； Step (3.5): Using the improved range migration algorithm, the wavenumber domain echo signal after compensating for the uniform compression term in step (3.4) is interpolated to complete the range migration correction, secondary range compression, and azimuth compression of the wavenumber domain echo signal. The improved range migration algorithm is shown in the following formula: ， In the formula: Represents the interpolation function; This represents a redefinition expression in the wavenumber domain; The wavenumber domain echo signal result after interpolation is shown in the following formula: Step (3.6): Perform range-direction IFFT processing on the wavenumber domain echo signal after interpolation in step (3.5), and the result is shown in the following formula: ， Then, a residual phase compensation term is constructed to compensate for the residual phase in the wavenumber domain echo signal after range-direction IFFT processing due to interpolation. The residual phase compensation term is shown in the following formula: ， In the formula: The constructor expression for the residual phase compensation term used to compensate for the echo signal; Indicates a fast time; The wavenumber domain echo signal after compensating for the residual phase is shown in the following equation: ； Step (3.7): Perform azimuth IFFT processing on the wavenumber domain echo signal after residual phase compensation in step (3.6). The processing result is shown below: ， , , In the formula: This represents a redefinition expression in the wavenumber domain; Indicates the distance traveled by the radar; Indicates location and time. Then, after performing a Decirp operation on the wavenumber domain echo signal processed by IFFT, correction is performed to make the Doppler center of the wavenumber domain echo signal after the Decirp operation zero. Step (3.8): Perform a direction-position FFT operation on the wavenumber domain echo signal obtained after correction in step (3.7) to obtain the azimuth of each angle. - Domain-focused image.

3. The multi-angle imaging method based on MIMO-SAR according to claim 1, characterized in that, In step 1, when the multiple-transmitter multiple-receiver synthetic aperture radar (MIMO-SAR) is mounted on a satellite, the echo signal of the point target received by the MIMO-SAR is divided into several imaging segments, and each segment has a different imaging angle, thereby obtaining echo signals of the point target at multiple different angles.

4. The multi-angle imaging method based on MIMO-SAR according to claim 1, characterized in that, In step 1, when the multi-transmitter multi-receiver synthetic aperture radar (MIMO-SAR) is mounted on an aircraft, it obtains echo signals from multiple different angles of a point target during a single flight based on digital multibeamforming technology.

5. The multi-angle imaging method based on MIMO-SAR according to claim 1, characterized in that, Step 2 is as follows: First, range compression is performed on the echo signal obtained in Step 1 at each angle; then, azimuth FFT processing is performed on the range-compressed echo signal; finally, based on the spatial degrees of freedom of the echo signal after azimuth FFT processing in the azimuth direction, a spatial filtering weight vector function is constructed, and the spatial filtering weight vector function is used to filter the echo signal after azimuth FFT processing to obtain an unambiguous azimuth signal.

Images