Satellite-ground bistatic SAR dual-angle two-dimensional deformation inversion method

Abstract

The invention discloses a satellite-ground bistatic synthetic aperture radar (SAR) dual-angle two-dimensional deformation inversion method, which constructs a dual-angle system configuration based on a satellite-ground bistatic SAR, proposes a two-dimensional deformation inversion method under the satellite-ground bistatic SAR dual-angle configuration, and obtains two-dimensional deformation of a kilometer-level region. The limitation that a traditional satellite-ground SAR can only obtain one-dimensional deformation in the equivalent sight line direction is solved, the method is suitable for refined monitoring of complex deformation fields in geological disasters such as landslides and earthquakes, and multi-dimensional data support is provided for disaster mechanism analysis. Observation data of ground double receiving stations are fused, shielding and low correlation areas are masked, the influence of noise phases and orbit error phases is effectively restrained, the problem that the space-variant property of the equivalent sight included angle of the satellite-ground bistatic SAR is large is solved, and the two-dimensional deformation precision reaches the centimeter level.

Description

Technical Field

[0001] The present invention belongs to the technical field of synthetic aperture radar, and particularly relates to a space-ground bistatic SAR two-angle two-dimensional deformation inversion method. Background Art

[0002] Deformation monitoring technology has very important applications in the fields of geological disaster prevention, urban ground subsidence monitoring, building safety, mineral resource development, etc. Existing deformation monitoring technologies include laser scanning, total station, GPS measurement, etc. They are restricted by bad weather and the observation area, and these factors will affect the accuracy and reliability of deformation measurement, bringing uncertainty to disaster warning and disaster assessment work. The differential interferometric synthetic aperture radar (D-InSAR) technology using spaceborne synthetic aperture radar can solve the above problems. However, most of the remote sensing satellites in orbit at present are polar orbit satellites. When performing D-InSAR, such satellites can only obtain the line-of-sight deformations in the ascending / descending orbit at two angles. Moreover, the north-south components of the line-of-sight at these two angles are very small, which will result in low accuracy of the north-south deformation obtained by inversion.

[0003] To improve the limitations of spaceborne SAR in the direction of deformation inversion, a space-ground bistatic SAR system can be used. The space-ground bistatic SAR system forms an equivalent phase center by setting up a receiving system on the ground to obtain additional deformation observation angles, and then can supplement the deformation inversion results in the north-south dimension and improve the deformation inversion accuracy. At the same time, the space-ground bistatic SAR can make full use of many remote sensing satellite resources. At present, this one-station fixed space-ground bistatic SAR has received the attention of many scholars internationally. Countries such as the United Kingdom, Spain, Germany, and China have all carried out relevant experiments and achieved a series of results in imaging, terrain elevation inversion, deformation inversion, etc.

[0004] To obtain the two-dimensional deformation of a scene, it is usually necessary to supplement the observation angles by combining single-satellite data. However, it is difficult to obtain single-satellite data. Considering the characteristics of the space-ground bistatic SAR, a two-angle receiving system can be considered by arranging two receiving stations on the ground, and then two groups of scene data with different line-of-sight directions can be obtained in one satellite repeat orbit period, and the two-dimensional deformation can be inverted. Therefore, there is an urgent need for a space-ground bistatic SAR two-angle two-dimensional deformation inversion method that can overcome the problem of large spatial variability of the equivalent line-of-sight angle of the space-ground bistatic SAR and invert the two-dimensional deformation of the observation area. Summary of the Invention

[0005] To solve the above problems, the present invention provides a space-ground bistatic SAR two-angle two-dimensional deformation inversion method. This method is applicable to the deformation inversion model of the space-ground bistatic SAR two-angle receiving configuration, can obtain the one-dimensional deformation quantities at two observation angles, and can be fused into the two-dimensional deformation of the scene, and at the same time provides an operation processing flow for obtaining the two-dimensional deformation.

[0006] The technical solution of the present invention is as follows:

[0007] A method for dual-angle two-dimensional deformation inversion of spaceborne-ground bistatic SAR, specifically including:

[0008] S1. Obtain radar data and related parameters; obtain the imaging data of each observation collected by two-angle receivers, as well as the range [x min , x max in the range direction and the range [y min , y max in the azimuth direction of the imaging grid; obtain the coordinates of the direct wave antenna and the satellite ephemeris of each observation; obtain the payload parameters of the satellite, including the wavelength λ and the bandwidth B; obtain the refined frequency points N c required for interferometric phase correction; obtain the external digital elevation model (DEM) of the imaging area; obtain the GoldStein filtering parameters, including the filtering block size A1, the overlapping block size A2, and the filtering coefficient α; obtain the window size W for correlation coefficient calculation and the correlation coefficient threshold C;

[0009] S2. Use the bistatic back-projection algorithm to image according to the observation data and the imaging grid obtained in S1 to obtain the bistatic imaging result, specifically including:

[0010] S21. Calculate the bistatic slant range of each grid point according to the satellite ephemeris and the selected ground grid (where the origin is set to the position of the receiver);

[0011] S22. Find the point corresponding to the slant range in the two-dimensional data, perform coherent accumulation along the range migration line after compensating the bistatic slant range phase to obtain the final bistatic imaging result;

[0012] S23. Since the two receivers form two-track images during the repeat pass, select the main images of the two angles respectively, and then perform interferometric processing to obtain the interferometric phase maps of the two angles respectively.

[0013] S3. According to the refined frequency points N c obtained in S1, use the Chirp-Z transform method to estimate the linear frequency error in the interferogram and perform interferometric phase correction on the two interferometric phase maps; re-image using the corrected ephemeris, and then calculate the correlation coefficient between the main and auxiliary images according to the window size W for correlation coefficient calculation obtained in S1.

[0014] The reason for performing this step is that the publicly available ephemeris used in imaging usually has errors, resulting in serious interference phase fringes in the imaging results. Since the imaging range of spaceborne bistatic SAR is limited and the number of image pixels is also limited, there is a problem of insufficient frequency resolution when using the traditional Fourier transform method to identify the frequency of interference phase fringes. This method uses the Chirp-Z transform to effectively improve the frequency resolution, thereby accurately compensating for the error phase. The essence of this method is to perform frequency domain transformation on the data of the interferogram along the range direction and azimuth direction, extract and eliminate the linear trend in the frequency domain, and finally compensate for the interference error. At the same time, the baseline error can be calculated based on the estimated linear fringe frequency and the corrected ephemeris can be obtained through compensation. Re-imaging the auxiliary image using the corrected ephemeris can improve the correlation coefficient between the main and auxiliary images.

[0015] S4. Geocode the interference results according to the elevation direction of the spaceborne bistatic SAR to project the repeat-pass interference results at two angles into the geodetic coordinate system of longitude, latitude, and altitude. Then register the repeat-pass interference results at two angles according to the geocoding results, specifically including:

[0016] S41. Calculate the elevation direction of the spaceborne bistatic SAR according to the satellite ephemeris and the ground grid of imaging;

[0017] S42. Use the external DEM to project each DEM point along the elevation direction onto the imaging plane, and establish the mapping relationship between the geodetic coordinate system of longitude, latitude, and altitude and each coordinate point on the imaging plane;

[0018] S43. Based on the above mapping relationship, back-project the repeat-pass interferogram into the geodetic coordinate system of longitude, latitude, and altitude;

[0019] S44. Register the repeat-pass interference results at two angles according to the geocoding results.

[0020] S5. Determine the visible area and perform masking processing in combination with the correlation coefficient, specifically including:

[0021] S51. With the help of the external DEM, calculate the elevation angle and azimuth angle of each target point in the receiver's observation scene. Then, at the same elevation angle / azimuth angle, set the target point closest to the receiver as the visible point, and the remaining target points as the occluded points.

[0022] S52. After determining the visible area, it is necessary to perform masking processing on the interferogram, that is, set the interference phase of the invisible area to 0. At the same time, calculate the correlation coefficient of the repeat-pass data, and also perform masking processing on the phase of the area in the visible area where the correlation coefficient is lower than the threshold C.

[0023] S53. After determining the visible areas at two angles, take the intersection of them to obtain the double-angle visible area.

[0024] S6. Phase unwrapping is performed on the re-orbit interferometry results of the two angles, and GoldStein phase filtering and multi-look processing are performed on the re-orbit interferometry results of the two angles according to the filtering parameters obtained in S1 to reduce phase noise.

[0025] S7. Calculate the terrain phase based on the external DEM and remove the terrain phase from the interference phase map to obtain the deformation phase of the two angles, recorded as Δφ r,d,1 and Δφ r,d,2 .

[0026] S8. Invert the one-dimensional deformation results of the two angles and fuse the two angle deformation measurement results into a two-dimensional deformation, specifically including:

[0027] S81. Invert the one-dimensional deformation results of the two angles. The direction of the deformation variable inverted by the single-angle re-orbit interferometry of the satellite-ground bistatic SAR is the equivalent line of sight direction of each angle, and the magnitudes d1 and d2 are expressed as:

[0028]

[0029] Where β1 and β2 are the bistatic angles of the two angles, respectively, and are calculated as:

[0030]

[0031] in, is the line of sight vector from the satellite to the target, and are the sight vectors from the receiver to the target at two angles.

[0032] S82: Fuse the one-dimensional deformation results of the two angles to obtain a two-dimensional deformation of the scene. First, the deformation of the two angles obtained in S81 is combined into a two-dimensional deformation vector, which is recorded as To quantitatively represent the two-dimensional deformation vector This method sets up a set of orthogonal bases, n1 and n2, such as Figure 4 As shown. Among them, n1 is the direction of the two angles equivalent to the line of sight (i.e. and ), n2 is the unit vector in the direction of the angle bisector of the two-dimensional deformation measurement plane ( and The unit vector in the plane where n lies is perpendicular to n1.

[0033] set up is with Unit vectors with the same direction, is with The unit vectors in the same direction are calculated as follows:

[0034]

[0035] Let \(H\) be the observation matrix, expressed as:

[0036]

[0037] Then the deformation measurement results in two dimensions are expressed as:

[0038]

[0039] where \(a_1\) and \(a_2\) are real numbers.

[0040] Beneficial effects:

[0041] 1. Construct a dual - angle system configuration based on space - borne and ground - based bistatic SAR, propose a two - dimensional deformation inversion method under the dual - angle configuration of space - borne and ground - based bistatic SAR, obtain two - dimensional deformation in a kilometer - level area, solve the limitation that traditional space - borne and ground - based SAR can only obtain one - dimensional deformation along the equivalent line - of - sight direction, and is applicable to the refined monitoring of complex deformation fields in geological disasters such as landslides and earthquakes, providing multi - dimensional data support for disaster mechanism analysis.

[0042] 2. Integrate the observation data of two ground receiving stations, mask the occluded and low - correlation areas, effectively suppress the influence of noise phase and orbit error phase, overcome the problem of large spatial variability of the equivalent line - of - sight angle of space - borne and ground - based bistatic SAR, and the two - dimensional deformation accuracy reaches the centimeter level, which can meet the deformation monitoring requirements of urban buildings, bridges, dams and other infrastructure, and support high - precision engineering safety assessment.

[0043] 3. Through the collaborative reception of two ground stations, synchronously obtain dual - angle observation data during a single satellite pass. Compared with the traditional single - station reception mode, the timeliness is improved, which is applicable to the real - time monitoring of geological disasters, and at the same time provides a theoretical basis for the subsequent research on multi - station distributed space - borne and ground - based SAR. Brief description of the drawings

[0044] Figure 1 is the system configuration diagram for two - dimensional deformation inversion of space - borne and ground - based bistatic SAR with dual - angles;

[0045] Figure 2 is the geometric diagram of repeat - pass interferometry of space - borne and ground - based bistatic SAR;

[0046] Figure 3 is the geometric diagram of single - angle repeat - pass differential interferometry of space - borne and ground - based bistatic SAR;

[0047] Figure 4 is the vector decomposition diagram of two - angle deformation measurement of space - borne and ground - based bistatic SAR;

[0048] Figure 5 is the experimental scene diagram of two - dimensional deformation inversion of space - borne and ground - based bistatic SAR with dual - angles;

[0049] Figure 6is the DEM of the experimental area;

[0050] Figure 7 The satellite-ground bistatic SAR images are at two angles ((a) is angle 1, (b) is angle 2);

[0051] Figure 8 The two angles of the heavy orbit interferogram after phase correction ((a) is angle 1, (b) is angle 2);

[0052] Figure 9 The one-dimensional deformation inversion results for two angles ((a) is angle 1, (b) is angle 2);

[0053] Figure 10 These are the results of dual-angle two-dimensional deformation inversion ((a) is the n1 direction, (b) is the n2 direction). DETAILED DESCRIPTION

[0054] The specific operation process of the present invention is further described in detail below through specific implementation methods and combined with the accompanying drawings.

[0055] Figure 1 The system configuration for dual-angle, two-dimensional deformation inversion using a satellite-ground bistatic SAR was demonstrated. Two receivers on the ground simultaneously receive scene echo signals, forming two equivalent phase centers. The satellites pass over the scene twice, illuminating it separately. These two imaging passes allow for interferometry and inversion of scene deformation.

[0056] S1. Obtain radar data and related parameters. Obtain the imaging data of each observation collected by the two angle receivers, as well as the range of the imaging grid [x min ,x max ] and azimuth range [y min ,y max ]; obtain the direct wave antenna coordinates and the satellite ephemeris of each observation; obtain the satellite payload parameters, including wavelength λ and bandwidth B; obtain the refined frequency points N required for interferometric phase correction c ; Obtain the external digital elevation model (DEM) of the imaging area; Obtain the GoldStein filter parameters, including the filter block size A1, the overlap block size A2 and the filter coefficient α; Obtain the window size W and the correlation coefficient threshold C for correlation coefficient calculation;

[0057] S2: Use the bistatic backprojection algorithm to perform imaging based on the observation data obtained in S1 and the imaging grid to obtain bistatic imaging results. Specifically, it includes:

[0058] S21. Calculate the bistatic slant range for each grid point based on the satellite's ephemeris and a selected ground grid (where the origin is set to the receiver's position);

[0059] S22. Locate the points corresponding to the slant range in the two-dimensional data, perform coherent accumulation along the range migration line after compensating the bistatic slant range phase to obtain the final bistatic imaging result;

[0060] S23. Since the two receivers each form two-track images during the repeat-pass, select the main images for the two respective angles, and then perform interferometric processing to obtain the interferometric phase diagrams for the two respective angles.

[0061] S3. According to the refined number of frequency points N obtained in S1 c , use the Chirp-Z transform method to estimate the linear frequency error in the interferogram and perform interferometric phase correction on the two interferometric phase diagrams; re-image using the corrected ephemeris, and then calculate the correlation coefficient between the master and slave images according to the window size W of the correlation coefficient obtained in S1.

[0062] The reason for performing this step is that the publicly available ephemeris used during imaging usually has errors, resulting in serious interferometric phase fringes in the imaging result. Since the imaging range of spaceborne bistatic SAR is limited and the number of image pixels is also limited, there is a problem of insufficient frequency resolution when using the traditional Fourier transform method to identify the frequency of the interferometric phase fringes. This method can effectively improve the frequency resolution by using the Chirp-Z transform, thereby accurately compensating the error phase. The essence of this method is to perform frequency domain transformation on the data of the interferogram along the range direction and azimuth direction, extract and eliminate the linear trend in the frequency domain, finally compensate the interferometric error, and at the same time, the baseline error can be calculated according to the estimated linear fringe frequency and the corrected ephemeris can be obtained by compensation. Re-imaging the slave image using the corrected ephemeris can improve the correlation coefficient between the master and slave images.

[0063] S4. Perform geocoding on the interferometric result according to the elevation direction of the spaceborne bistatic SAR to project the repeat-pass interferometric results of the two angles into the geodetic coordinate system of longitude, latitude, and altitude. Then register the repeat-pass interferometric results of the two angles according to the geocoding results, specifically including:

[0064] S41. Calculate the elevation direction of the spaceborne bistatic SAR according to the satellite ephemeris and the imaging ground grid;

[0065] S42. Use the external DEM to project each DEM point along the elevation direction onto the imaging plane to establish the mapping relationship between the geodetic coordinate system of longitude, latitude, and altitude and each coordinate point on the imaging plane;

[0066] S43. Based on the above mapping relationship, back-project the repeat-pass interferogram into the geodetic coordinate system of longitude, latitude, and altitude;

[0067] S44. Register the repeat-pass interferometric results of the two angles according to the geocoding results.

[0068] The reason for performing this step is that after obtaining the absolute phase of the scene terrain, it is necessary to establish the correspondence between the scene terrain phase and the scene terrain to facilitate registration and subsequent processing of judging the visible area.

[0069] S5. Judge the visible area and perform masking processing in combination with the correlation coefficient. The specific steps are as follows:

[0070] S51. With the help of an external DEM, calculate the elevation angle and azimuth angle of each target point in the receiver observation scene. Then, at the same elevation angle / azimuth angle, set the target point closest to the receiver as the visible point, and the remaining target points as the occluded points.

[0071] After judging the visible area in S52, it is necessary to perform masking processing on the interferogram, that is, set the interferometric phase of the invisible area to 0. At the same time, calculate the correlation coefficient of the repeat-pass data, and also perform masking processing on the phase of the area in the visible area where the correlation coefficient is lower than the threshold C.

[0072] After judging the visible areas of the two angles in S53, take the intersection of them to obtain the double-angle visible area.

[0073] The reason for performing this step is that due to the relatively low elevation of the receiving end of the space-ground bistatic SAR system, occlusion areas will appear near the targets with higher elevation in the experimental scene. The interferometric results corresponding to these occlusion areas have relatively low signal-to-noise ratios and poor interferometric quality, which will affect the subsequent phase unwrapping processing. These areas need to be masked. At the same time, the low-coherence part in the visible area will also lead to a decrease in the accuracy of deformation inversion, so masking is also required.

[0074] S6. Perform phase unwrapping on the repeat-pass interferometric results of the two angles, and perform GoldStein phase filtering and multi-look processing on the repeat-pass interferometric results of the two angles according to the filtering parameters obtained in S1 to reduce phase noise.

[0075] The reason for performing this step is to obtain the actual terrain phase, reduce phase noise, and smooth the interferometric phase.

[0076] S7. Calculate the terrain phase according to the external DEM, and remove the terrain phase in the interferometric phase diagram to obtain the deformation phases of the two angles, denoted as Δφ r,d,1 and Δφ r,d,2 .

[0077] The following derives the calculation method of the terrain phase in repeat-pass space-ground bistatic SAR. The interferometric geometry of repeat-pass space-ground bistatic SAR is as Figure 2 shown. In the repeat-pass space-ground bistatic SAR image after removing the flat ground, the phase of the target T can be respectively expressed as:

[0078]

[0079] Furthermore, the topographic phase after the interference between orbit S1 and orbit S2 can be expressed as:

[0080]

[0081] Wherein, and are unit vectors in the same direction as and respectively, is the unit vector in the same direction as is, R is approximately the slant range from the transmitting end to the projection point T0 of the target elevation, that is, the modulus of, θ is the angle between the target elevation direction and the horizontal plane, α is the angle between the vertical baseline and the target elevation, h is the target elevation, which can be obtained through external DEM.

[0082] After calculating the topographic phase according to equation (7), removing this phase from the interferometric phase diagram can obtain the deformation phase.

[0083] S8. Invert the one-dimensional deformation results of two angles, and fuse the deformation measurement results of the two angles into two-dimensional deformation. The specific steps include:

[0084] S81. Invert the one-dimensional deformation results of two angles.

[0085] Next, the deformation inversion model of spaceborne-ground bistatic SAR is derived. First, establish the deformation inversion model of single-angle spaceborne-ground bistatic SAR with single transmit and single receive. On this basis, establish the two-angle deformation inversion model of spaceborne-ground bistatic SAR.

[0086] Figure 3 is the geometric diagram of single-angle repeat-pass differential interferometry of spaceborne-ground bistatic SAR. The repeat-pass interferometric phase can be expressed as:

[0087]

[0088] Since the deformation is usually on the centimeter scale, the repeat-pass baseline of the satellite is usually on the order of hundreds of meters, while the slant range from the satellite to the target is usually on the order of hundreds of kilometers. The slant range difference term in equation (8) can be expanded using Taylor series and simplified to obtain:

[0089]

[0090] Wherein, is the unit vector in the same direction as is the unit vector in the same direction as is the unit vector in the same direction as is the unit vector in the same direction as.

[0091] Based on equation (9), the repeat-pass interferometric phase can be divided into the deformation component Δφ related to the deformation vector r,d ​​and the topographic component Δφ related to the spatial baseline r,t , are respectively expressed as:

[0092]

[0093] where, is the unit vector along the equivalent line-of-sight direction. It can be seen from equation (10) that the deformation measured by this system is the scalar of the true deformation vector (three-dimensional vector ) projected onto the angular bisector direction of the bistatic angle, that is, the deformation direction measured by this system is the bistatic angle bisector direction. In summary, the direction of the deformation quantity inversed by spaceborne-ground bistatic SAR single-angle repeat-pass interferometry is the equivalent line-of-sight direction of each angle, and the magnitudes d1 and d2 can be expressed as:

[0094]

[0095] where, β1 and β2 are the magnitudes of the bistatic angles of the two angles respectively.

[0096] S82. Fuse the one-dimensional deformation results of the two angles to obtain the two-dimensional deformation of the scene.

[0097] Next, the spaceborne-ground bistatic SAR double-angle deformation inversion model is derived, and its geometric schematic diagram is as Figure 1 shown. Based on the above single-angle deformation inversion model, it can be known that the spaceborne-ground bistatic SAR double-angle system can obtain the deformation vectors in two equivalent line-of-sight directions (i.e., and ), which are respectively expressed as and The deformations of the two angles can be synthesized into a two-dimensional deformation vector, denoted as To quantitatively represent the two-dimensional deformation vector This method sets a set of orthogonal bases, n1 and n2, as Figure 4 shown. Among them, n1 is the unit vector in the angular bisector direction of the two-angle equivalent line-of-sight directions (i.e., and ), and n2 is the unit vector perpendicular to n1 in the two-dimensional deformation measurement plane (the plane where and are located).

[0098] Since the two-dimensional deformation vector and the orthogonal bases n1 and n2 are all in the two-dimensional deformation measurement plane, and and are respectively projected onto the two-angle equivalent line-of-sight directions, the deformation quantities d1 and d2 can be expressed in matrix form as:

[0099]

[0100] where \(a_1\) and \(a_2\) are real numbers, is a unit vector in the same direction as and is a unit vector in the same direction as .

[0101] Let the observation matrix \(H\) be:

[0102]

[0103] Based on the least squares estimation and combined with equations (12) and (13), the coefficients \(a_1\) and \(a_2\) can be expressed as:

[0104]

[0105] Substitute equation (11) into equation (14) to obtain the deformation measurement results in the directions of \(n_1\) and \(n_2\).

[0106] So far, all steps are completed.

[0107] Next, an implementation example is given in combination with specific parameters.

[0108] In this example, we collected the space-ground bistatic SAR data of the LuTan-1 satellite. The collection times were August 15, 2023 and August 19, 2023, and the location was Jinan City, Shandong Province. The experimental scenario and the settings of the receiver are as Figure 5 shown. In step S1, the parameters of data processing are shown in Table 1. In addition, the satellite ephemeris data can be downloaded from the Celestrak website (current data) or applied for (historical data). The DEM data of this area is as Figure 6 shown.

[0109] Table 1

[0110]

[0111]

[0112] After executing step S2, the obtained bistatic SAR images (the main images for each of the two angles) are as Figure 7 shown.

[0113] After executing step S3, the obtained corrected multi-track interferometric phase diagrams (for two angles) are as Figure 8 shown.

[0114] After executing steps S4 - S81, the obtained results are as Figure 9The one-dimensional deformation inversion results of the two angles shown. After evaluation, the measured means of the single-angle deformation inversion results of the two angles are -0.19 cm and -0.43 cm respectively, and the standard deviations of the deformation are 1.03 cm and 1.19 cm respectively. Since the time baseline of the repeat-pass interferometry is 4 days, it can be approximately considered that there is no deformation in the scene. This verifies that the single-angle deformation inversion result of the space-ground bistatic SAR has centimeter-level accuracy.

[0115] After performing step S82, the two-angle two-dimensional deformation inversion result as shown in Figure 10 is obtained. After evaluation, the measured deformation means in the n1 direction and the n2 direction are -0.42 cm and 0.1 cm respectively, and the standard deviations are 0.78 cm and 3.39 cm respectively. The deformation accuracy in the n2 direction is worse than that in the n1 direction because the included angle between the equivalent line-of-sight directions of the two angles is small, that is, the component of the deformation result in the direction perpendicular to the bisector of the equivalent line-of-sight angle is small.

[0116] From Figure 9 and Figure 10 and the evaluation results, it can be seen that the proposed method can accurately obtain the deformation amount of the observation area. The above results show that this patent can effectively realize the two-angle two-dimensional deformation inversion of the space-ground bistatic SAR, proving the effectiveness of this method.

[0117] Of course, the present invention may also have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and deformations according to the present invention, but these corresponding changes and deformations should all fall within the protection scope of the appended claims of the present invention.

Claims

1. A space-ground bistatic SAR double-angle two-dimensional deformation inversion method, characterized in that It includes the following steps: S1. Obtain radar data and related parameters; obtain the imaging data of each observation collected by two angle receivers, as well as the range [x min , x max in the range direction and the range [y min , y max in the azimuth direction of the imaging grid; obtain the coordinates of the direct wave antenna and the satellite ephemeris of each observation; obtain the payload parameters of the satellite, including the wavelength λ and the bandwidth B; obtain the number of refined frequency points N required for interferometric phase correction c ; obtain the external digital elevation model (DEM) of the imaging area; obtain the Goldstein filtering parameters, including the filter block size A1, the overlapping block size A2 and the filtering coefficient α; obtain the window size W for correlation coefficient calculation and the correlation coefficient threshold C; S2. Use the bistatic back projection algorithm to perform imaging on the observation data obtained in S1 and the imaging grid to obtain the bistatic imaging result; select the main images of the two angles respectively, and then perform interference processing to obtain the interference phase diagrams of the two angles respectively; S3. The refined frequency point number N obtained according to S1 c , use the Chirp-Z transform method to estimate the linear frequency error in the interferogram and perform interferometric phase correction on the two interferometric phase diagrams; re-image using the corrected ephemeris, and then calculate the correlation coefficient of the master and slave images according to the window size W calculated by the correlation coefficient obtained according to S1; S4. Geocode the interference result in the elevation direction of the space-ground bistatic SAR to project the repeat-pass interference results of the two angles into the geodetic coordinate system; Then register the repeat-pass interference results of the two angles according to the geocoding result; S5. Determine the visible area and perform masking processing in combination with the correlation coefficient; S6. Unwrap the phase of the repeat-pass interference results of the two angles, and perform GoldStein phase filtering and multi-look processing on the repeat-pass interference results of the two angles according to the filtering parameters obtained in S1 to reduce the phase noise; S7. Calculate the topographic phase based on the external DEM, and remove the topographic phase from the interferometric phase diagram to obtain the deformation phases at two angles, denoted as Δφ r,d,1 and Δφ r,d,2 ; S8. Invert the one-dimensional deformation results of the two angles, and fuse the deformation measurement results of the two angles into two-dimensional deformation.

2. The bistatic SAR dual-angle two-dimensional deformation inversion method according to claim 1, characterized in that, The step S2 includes: S21. Calculate the bistatic slant range of each grid point according to the satellite ephemeris and the selected ground grid (where the origin is set to the position of the receiver); S22. Find the point corresponding to the slant range in the two-dimensional data, compensate the bistatic slant range phase, and perform coherent accumulation along the range migration line to obtain the final bistatic imaging result; S23. Since the two receivers each form two-track images during the repeat pass, select the main images of the two angles respectively, and then perform interference processing to obtain the interference phase diagrams of the two angles respectively.

3. The bistatic SAR two-angle two-dimensional deformation inversion method according to claim 1, characterized in that The step S4 includes: S41. Calculate the elevation direction of the space-ground bistatic SAR according to the satellite ephemeris and the imaging ground grid; S42. Use the external DEM to project each DEM point along the elevation direction onto the imaging plane, and establish the mapping relationship between the geodetic coordinate system and each coordinate point on the imaging plane; S43. Based on the above mapping relationship, back-project the repeat-pass interference map into the geodetic coordinate system; S44. Register the repeat-pass interference results of the two angles according to the geocoding result.

4. The bistatic SAR two-angle two-dimensional deformation inversion method according to claim 1, characterized in that The step S5 includes: S51. With the help of the external DEM, calculate the elevation angle and azimuth angle of each target point in the receiver observation scene; then, at the same elevation angle / azimuth angle, set the target point closest to the receiver as the visible point, and the remaining target points as the occluded points; S52. After determining the visible area, it is necessary to perform masking processing on the interference map, that is, set the interference phase of the invisible area to 0; at the same time, calculate the correlation coefficient of the repeat-pass data, and also perform masking processing on the phase of the area in the visible area where the correlation coefficient is lower than the threshold C. S53. After determining the visible areas of the two angles, take the intersection of them to obtain the double-angle visible area.

5. A space-ground bistatic SAR dual-angle two-dimensional deformation inversion method according to claim 1, characterized in that The step S8 includes: S81. Invert the one-dimensional deformation results of the two angles; S82. Fuse the one-dimensional deformation results of the two angles to obtain the two-dimensional deformation of the scene.

6. A space-ground bistatic SAR double-angle two-dimensional deformation inversion method according to claim 5, characterized in that In the step S82, the method of fusing the one-dimensional deformation results of the two angles to obtain the two-dimensional deformation of the scene is: The deformations at two angles obtained through step S81 are synthesized into a two-dimensional deformation vector, denoted as To quantitatively represent the two-dimensional deformation vector This method sets a set of orthogonal bases, n1 and n2; where n1 is the unit vector in the angular bisector direction of the equivalent line-of-sight directions at two angles (i.e., and ), and n2 is the unit vector perpendicular to n1 in the two-dimensional deformation measurement plane ( and 's plane); Let be the unit vector in the same direction as , and be the unit vector in the same direction as . The calculation method is as follows: Let H be the observation matrix, which is expressed as: Then the deformation measurement results of the two dimensions are expressed as: where a1 and a2 are real numbers.

Images