Spaceborne SAR Geometric Calibration Method with Additional Control Point Geometric Constraints

Abstract

The present invention relates to a spaceborne SAR geometric calibration method with additional control point geometric constraints, which is characterized by the following steps: collecting control points; fitting the control points based on a fitting method to obtain additional geometric constraints of the control points; constructing an initial SAR geolocation model; establishing a control point pair database with pixel coordinates and geodetic coordinates; calculating the geometric positioning error of the initial SAR geolocation model; establishing a SAR geometric calibration model; and solving the SAR geometric calibration model to obtain geometric calibration parameters, thereby realizing SAR geometric calibration. Compared with the prior art, the present invention has the advantages of reducing the error of collecting control points for DOM and improving the SAR geometric calibration accuracy, etc.

Description

Technical Field

[0001] The present invention relates to the field of remote sensing images, and in particular to a spaceborne SAR geometric calibration method with additional control point geometric constraints. Background Art

[0002] Improving the ground positioning accuracy of Synthetic Aperture Radar (SAR) satellites helps to expand the application scope of SAR images and the benefits of SAR satellites. Common methods for improving the ground positioning accuracy of SAR images include geometric correction methods based on ground control points, block adjustment methods, and geometric calibration methods. Since the geometric correction method can only be carried out scene by scene with low efficiency, the block adjustment method only requires a small number of ground control points, but it takes a large amount of computing time to extract homologous connection points and can only improve the positioning accuracy within the area. The geometric calibration method

[0003] SAR geometric calibration requires the use of ground control points to solve geometric calibration parameters. Therefore, the accuracy of ground control points directly affects the reliability of geometric calibration. In order to obtain reliable geometric calibration parameters, it is often necessary to deploy corner reflectors in a large-scale site to obtain high-precision ground control points. However, there are problems such as limited suitable sites, high manufacturing costs of corner reflector equipment, and large field measurement workloads. Therefore, there is currently a large amount of research on obtaining three-dimensional ground control information based on Digital Orthophoto Map (DOM) and Digital Elevation Model (DEM). This method only needs to find homologous points on the SAR image and DOM respectively, and then obtain the plane geodetic coordinates on the DOM by means of point piercing. Based on this plane geodetic coordinate, the elevation value is obtained by interpolation algorithm in the DEM. Thus, the three-dimensional geodetic coordinates of the control point can be obtained. The pixel coordinates in the image space are obtained by methods such as gray centroid to obtain sub-pixel accuracy coordinates. Combining the geodetic coordinates and pixel coordinates of the homologous points forms a control point pair for the solution of the geometric calibration model.

[0004] Although it is relatively gradual to consider combining DOM with DEM to obtain ground control points, there is a large uncertainty. For example, the DOM piercing point error can usually reach 1 pixel, which will seriously affect the reliability of SAR geometric calibration. At present, some studies have combined optical cameras and lidar devices to obtain the positions of devices in unfamiliar environments and build maps. In this process, the optical camera and lidar devices observe the surrounding environment simultaneously. The straight line features existing in the environment can be used to calibrate the imaging parameters of the optical camera and lidar, and improve the accuracy of positioning and mapping. There are also studies on optimizing the internal and external parameters of satellites based on straight line features such as building shadows and roof ridge lines in optical satellite images. However, there is no relevant research in the field of SAR geometric calibration. Therefore, it is necessary to further explore the research on considering the geometric constraints of ground control points in the process of SAR geometric calibration to weaken the influence of DOM piercing point error on geometric calibration and improve the geometric calibration accuracy. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defects existing in the above-mentioned prior art and provide a spaceborne SAR geometric calibration method with additional control point geometric constraints to weaken the influence of the error of the control points collected in DOM on geometric calibration and improve the SAR geometric calibration accuracy.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A spaceborne SAR geometric calibration method with additional control point geometric constraints includes the following steps:

[0008] Collect control points based on the collected SAR images and DOM, or collect control points according to real prior information;

[0009] Fit the control points based on a fitting method to obtain additional geometric constraints of the control points;

[0010] Construct an initial SAR geolocation model according to the SAR image auxiliary file, SAR imaging parameters, and SAR satellite orbit file of the SAR image;

[0011] Collect homologous image points in the SAR intensity image and DOM after SAR image preprocessing, and establish a control point pair database with pixel coordinates and geodetic coordinates;

[0012] Calculate the geometric positioning error of the initial SAR geolocation model based on the control point pair database;

[0013] Based on analyzing the geometric positioning error and SAR imaging mechanism, determine the SAR image geolocation error, and establish a SAR geometric calibration model based on the initial SAR geolocation model and additional geometric constraints;

[0014] Solve the SAR geometric calibration model to obtain geometric calibration parameters, thereby realizing SAR geometric calibration.

[0015] Further, the additional geometric constraints include additional line constraints, additional circular constraints, and additional elliptical constraints.

[0016] Further, the specific steps for obtaining the additional geometric constraints of the control points include:

[0017] Classify the control points, establish a set of control points with implicit constraints, and establish fitting models respectively;

[0018] Fit the set of control points based on the fitting model to obtain statistical parameters;

[0019] Calculate the fitting residuals of each control point based on the statistical parameters;

[0020] Based on the fitting residuals, reject the point sets with outliers whose rejection threshold is greater than the preset threshold and the number of control points is less than the preset number, and finally form the additional geometric constraints of the control points.

[0021] Further, the initial SAR geolocation model is:

[0022]

[0023] In the formula, R is the slant range between the observed target and the sensor, (Xs, Ys, Zs) is the spatial rectangular coordinates of the sensor at the imaging time of a certain target, (X g , Y g , Z g ) are the object space coordinates of a certain target on the ground, that is, the collected control points, R0 is the starting slant range in the range direction, x is the image space coordinate in the range direction, (x, y) are the high-precision sub-pixel coordinates collected in the SAR image, P r is the slant range resolution in the range direction; f d is the Doppler center frequency existing between the satellite platform and the observed target at the imaging time, P sc and V sc are the position and velocity vectors of the sensor respectively, P gc and V gc are the position and velocity vectors of the observed target respectively; a and b are the major and minor semi-axes of the reference ellipsoid, h is the elevation of the observed target, and λ is the radar wavelength;

[0024] The conditional equation of the initial SAR geolocation model is:

[0025]

[0026] In the formula, x is the column number of the SAR image, c is the speed of light, f ris the system range sampling frequency.

[0027] Further, the control point pair database is divided into a control point database and a check point database.

[0028] Further, the geometric positioning error includes object space positioning error and image space positioning error. The object space positioning error is the difference between the pixel coordinates of the control point pair converted to geodetic coordinates and the geodetic coordinates of the control point pair. The image space positioning error is the difference between the geodetic coordinates of the control point pair converted to pixel coordinates and the pixel coordinates of the control point pair.

[0029] Further, the expression of the SAR geometric calibration model is:

[0030]

[0031] where i is the column coordinate of the point target in the SAR image, j is the row coordinate of the point target in the SAR image, t r is the imaging time of the i-th column of the SAR image, t r0 is the imaging time of the first column of the SAR image, Δtr is the range imaging time correction value, t delay is the atmospheric delay time, f r is the system range sampling frequency; t a is the imaging time of the j-th row of the SAR image, t a0 is the imaging time of the first row of the SAR image, Δta is the azimuth imaging time correction value, f a is the SAR satellite pulse repetition frequency;

[0032] The conditional equation of the SAR geometric calibration model is:

[0033]

[0034] where t r0 is the imaging time of the first column of the SAR image, c is the speed of light, t delay is the calculated atmospheric delay time, Δt r is the SAR slant range time correction value, x is the column coordinate of the SAR image, f r is the system range sampling frequency, λ is the radar wavelength, R sc and V sc are the SAR satellite orbital position vector and velocity vector respectively, R gc and V gc are the position vector and velocity vector of the ground point target respectively. Dis(.) calculates the distance from each control point participating in geometric calibration to the fitted geometric element.

[0035] Furthermore, the geometric positioning error of the SAR image relative to the ground includes the internal electronic time delay, the time synchronization error between the SAR payload and the GNSS payload, and the radar sampling frequency and pulse repetition frequency errors.

[0036] Furthermore, the specific steps for obtaining the geometric calibration parameters include:

[0037] Expand the geometric calibration parameters of the conditional equation to the first term using the Taylor series for linearization;

[0038] Solve for the geometric calibration parameters using the Newton-Raphson method.

[0039] Furthermore, it also includes updating the initial SAR positioning model relative to the ground based on the geometric calibration parameters to obtain an updated SAR positioning model relative to the ground, and using the checkpoint database to evaluate the positioning error of the updated SAR positioning model relative to the ground before and after the update.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] (1) By considering the information of the additional geometric constraints of the control points and applying it in geometric calibration, the present invention further reduces the error compared with the method that only considers the position information of the control points, achieving a meter-level improvement in the SAR geometric positioning performance.

[0042] (2) SAR geometric calibration requires the use of control points to solve the geometric calibration parameters. Therefore, the accuracy of the control points directly affects the reliability of geometric calibration. The present invention improves the accuracy of geometric calibration by adding geometric constraints to the collected control points.

[0043] (3) In order to obtain reliable geometric calibration parameters, it is often necessary to deploy corner reflectors within a large area to obtain high-precision control points. The present invention obtains three-dimensional ground control information based on DOM and DEM. This method only needs to find homologous points on the SAR intensity image and DOM respectively, and then obtain the plane geodetic coordinates on the DOM by the method of puncturing points. Based on this plane geodetic coordinate, the elevation value is obtained by interpolation algorithm in DEM. This method of obtaining high-precision ground control points can be applied to various sites, reducing the measurement workload and cost.

[0044] (4) By analyzing the geometric positioning error and the SAR imaging mechanism, the present invention determines the geometric positioning error of the SAR image relative to the ground, thereby constructing a SAR geometric calibration model and updating the initial SAR positioning model relative to the ground, significantly improving the positioning accuracy of the updated SAR positioning model relative to the ground. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flowchart of the present invention;

[0046] Figure 2 It is a regional map of the case study for the embodiment;

[0047] Figure 3 It is a schematic diagram for extracting control points with additional geometric constraints;

[0048] Figure 4 It is a comparison of the geometric calibration results with additional linear constraints and the traditional point-based ones;

[0049] Figure 5 It is a comparison of the geometric calibration results with additional circular constraints and the traditional point-based ones. Specific implementation manner

[0050] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and gives the detailed implementation manner and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0051] As Figure 2 shown, this embodiment takes the Yangtze River Delta in China as the region, and takes the GF-3 and Sentinel-1A SAR images as the research objects. The standard atmospheric model is combined with the Saastamoinen model and the Klobuchar model to estimate the atmospheric path delay. The DOM and DEM are used to obtain the three-dimensional geodetic coordinates with linear and circular features as control points, forming a spaceborne SAR geometric calibration method with additional control point geometric constraints, including the following steps:

[0052] S1. Collect control points based on the collected SAR images and DOM, or collect control points according to real prior information.

[0053] In the SAR images, natural features such as exposed rocks in flat ground, buildings and street lamps in urban areas, and wind turbines, etc. have strong and stable backscattering characteristics, so they can be used as stable point targets for collecting control points. Identify such targets and collect the geodetic coordinates of such targets in the DOM. At the same time, control points can also be directly collected by using GNSS (Global Navigation Satellite System) equipment according to real prior information. According to the above research objects, the basic information of the obtained SAR images is shown in Table 1.

[0054] Table 1 Image basic information

[0055]

[0056] S11. Collect SAR images, perform preprocessing steps such as radiometric calibration and complex data conversion on the SAR images to obtain SAR intensity images, and the basic contour information and phase centers of ground targets can be identified in the intensity images;

[0057] S12. Select SAR pairs with good spatio-temporal baselines, estimate coherence, and extract independent point targets with stable scattering intensities, such as man-made targets arranged regularly like wind turbines and street lamp poles on both sides of the road. This type of point target is very conducive to identifying the phase center and is convenient for obtaining geodetic coordinates by means of collection in DOM or field measurement operations.

[0058] S13. According to real prior information such as road planning and landscape design, extract regularly distributed targets, such as circular roads, and directly collect the corresponding control points. In this study, regularly arranged wind turbines, straight roads, and the intersection centers of circular roads in the wind farm are used as control points with additional linear constraints and additional circular constraints respectively.

[0059] S2. Fit the control points based on the fitting method to obtain additional geometric constraints for the control points.

[0060] Table 2 Fitting parameters of linear and circular geometric elements

[0061]

[0062] As Figure 3 shown in the schematic diagram of the extraction of control points with additional geometric constraints and the fitting parameters of linear and circular elements shown in Table 2, this step performs linear fitting, circular fitting, etc. on some or all of the collected control points to estimate whether there are geometric constraints such as lines or circles for the collected target control points. Specifically:

[0063] S21. Classify the control points, establish a set of control points with implicit linear and circular constraints, and establish fitting models respectively.

[0064] S22. Based on the fitting models, fit the linear and circular control point sets to obtain fitting parameters of geometric features and statistical parameters such as R2 for evaluating the quality of feature fitting.

[0065] S23. Calculate the fitting residuals of each control point based on the statistical parameters.

[0066] S24. Based on the fitting residuals, reject the point sets with outliers greater than the preset threshold and the number of control points less than the preset number, and finally obtain the additional geometric constraints for the control points.

[0067] S3. Construct an initial SAR geolocation model according to the SAR image auxiliary file, SAR imaging parameters, and SAR satellite orbit file of the SAR image to realize the mutual mapping between pixel coordinates and geodetic coordinates.

[0068] This step is specifically as follows:

[0069] S31. Fit the position and velocity during satellite imaging using a cubic polynomial according to the ephemeris data in the auxiliary file:

[0070]

[0071] where (a0, a1, a2, a3, b0, b1, b2, b3, c0, c1, c2, c3) are the coefficients of the fitted position equation, and t a is the imaging time corresponding to the y-th row pixel in the SAR image, t a0 is the starting imaging time in the azimuth direction, f a is the pulse repetition frequency, and (X, Y, Z) are the object space coordinates of the ground target point.

[0072] S32. Construct a SAR geolocation model according to the SAR satellite imaging principle:

[0073]

[0074] where R is the slant range between the observed target and the sensor, (Xs, Ys, Zs) are the space rectangular coordinates of the sensor at the imaging time of a certain target, (X g , Y g , Z g ) are the object space coordinates of a certain ground target, that is, the collected control points, R0 is the starting slant range in the range direction, x is the image space coordinate in the range direction, (x, y) are the high-precision sub-pixel coordinates collected in the SAR image, P r is the range direction slant range resolution; f d is the Doppler center frequency existing between the satellite platform and the observed target at the imaging time, P sc and V sc are the position and velocity vectors of the sensor respectively, P gc and V gc are the position and velocity vectors of the observed target respectively; a and b are the major and minor semi-axes of the reference ellipsoid respectively, h is the elevation of the observed target, and λ is the radar wavelength.

[0075] S33. Then list the conditional equations:

[0076]

[0077] where x is the SAR image column number, c is the speed of light, and f r is the system range direction sampling frequency.

[0078] S34. Establish the above equation for each image point. Using the object space coordinates of the center point as the initial value, take partial derivatives of the object space coordinates of the ground target point according to the condition equation (3) during iteration, continuously update the partial derivative matrix. When the coordinate correction value is less than the threshold, it can be considered that the model converges, end the iteration, and the object space coordinates of the ground point can be obtained.

[0079] S4. Collect corresponding image points in the SAR intensity image and DOM after SAR image preprocessing, and establish a control point pair database with pixel coordinates and geodetic coordinates.

[0080] This step collects corresponding image points in the SAR intensity image and DOM after SAR image preprocessing, records the pixel coordinates of the corresponding image points in the SAR image and the geodetic coordinates in the DOM, extracts the elevation values of the corresponding points by interpolation algorithm in the DEM according to the geodetic coordinates, and forms a control point pair database with pixel coordinates and geodetic coordinates. Specifically:

[0081] S41. Convert the geodetic coordinates of the control points into pixel coordinates in the SAR image according to the SAR geolocation model; using this coordinate as the initial value, search for potential corresponding points in the SAR intensity image, and establish a 10×10 window by the gray centroid method. Interpolate the gray values within this window to obtain a 100×100 window, calculate the centroid coordinates of the image pixels row by row, and then calculate the centroid coordinates in the column direction to obtain high-precision sub-pixel coordinates. This sub-pixel coordinate is a more accurate pixel coordinate. Form a corresponding point pair by combining this pixel coordinate and the geodetic coordinates of the collected control points for SAR geometric calibration;

[0082] S42. Perform the above operations on each collected control point, and check the recognition accuracy of each point. If the collected control points cannot be clearly recognized simultaneously in multiple images, it means that this point in the SAR image cannot be reliably recognized due to factors such as environmental changes or changes in the SAR incident angle. The corresponding point pair of this point needs to be removed from the control point database;

[0083] S43. Obtain the pixel coordinates and geodetic coordinates of reliable corresponding point pairs in all SAR images, and form a control point pair database with pixel coordinates and geodetic coordinates.

[0084] In addition, the control point pair database is divided into a control point database and an inspection point database according to the requirement of uniform spatial distribution.

[0085] S5. Calculate the geometric positioning error of the initial SAR geolocation model based on the control point pair database.

[0086] This step mainly includes: using the above SAR geolocation model to convert the pixel coordinates of the control point pairs to geodetic coordinates respectively and taking the difference with the geodetic coordinates of the control points, so as to obtain the object space positioning error. At the same time, converting the geodetic coordinates of the control point pairs to pixel coordinates and taking the difference with the pixel coordinates of the control points, so as to obtain the image space positioning error.

[0087] S6. Based on the analysis of the geometric positioning error and SAR imaging mechanism, determine the geometric positioning error of the SAR image to the ground, and establish a SAR geometric calibration model based on the initial SAR geolocation model and additional geometric constraints.

[0088] This step mainly analyzes the sources of SAR geometric positioning error to the ground. For external errors, correction models are established respectively, and for internal errors, a SAR geometric calibration model is established according to the SAR imaging mechanism. Specifically:

[0089] S61. By analyzing the satellite-ground geometric relationship and imaging mechanism of SAR, it can be determined that SAR orbit error, SAR ranging error, SAR equipment error, atmospheric path delay error, ground elevation error, etc. are the main error sources of SAR geometric positioning to the ground.

[0090] S62. Using precise orbit data, the SAR orbit error can reach the centimeter level, and the SAR positioning mechanism determines that the vertical orbit error is related to the elevation direction positioning error, and the radial error affects the plane accuracy. Currently, the centimeter-level orbit accuracy can meet the requirements of high-precision calibration.

[0091] S63. The SAR ranging error and equipment error are the main factors of the SAR plane positioning error and are also the main error items to be considered in geometric calibration.

[0092] S64. The atmospheric path delay error can be divided into tropospheric atmospheric delay and ionospheric delay. Among them, the tropospheric atmospheric delay is eliminated by relying on tropospheric atmospheric parameters and the zenith tropospheric delay correction model; the ionospheric delay is eliminated by relying on ionospheric electron concentration data and the ionospheric delay correction model. Among them, the standard atmospheric model can provide atmospheric parameters for the SAR geometric calibration model and input them into the atmospheric path delay calculation model. The atmospheric parameters at any location and time can be expressed as:

[0093]

[0094] where Press is the atmospheric pressure, Temp is the Kelvin temperature, h is the altitude, w is the relative humidity, and W press is the water vapor partial pressure.

[0095] S65. According to the standard atmospheric parameters provided above, use the Saastamoinen model to calculate the tropospheric atmospheric path delay in the line-of-sight direction. The method is as follows:

[0096]

[0097] Among them, φ is the incident angle of the SAR satellite, and D h is the path delay caused by the neutral atmosphere, and D w is the path delay caused by water vapor in the atmosphere. ZTD is the total tropospheric delay.

[0098] S66. The ionospheric path delay can be estimated using the single-frequency Klobuchar model. The method is as follows:

[0099]

[0100] Among them, TEC is the ionospheric electron concentration, which can be provided by the daily 2-hour ionospheric electron concentration product provided by the Center for Orbit Determination in Europe (CODE); f is the center frequency of the microwave signal transmitted by the radar; ZID is the ionospheric delay.

[0101] S67. The ground elevation error mainly relies on obtaining higher-precision DEM data products and field measurements using GNSS equipment.

[0102] S68. According to the above error analysis and the SAR imaging mechanism, it can be determined that the main errors in the geometric positioning of the SAR image on the ground are the internal electronic time delay of the system, the time synchronization error between the SAR payload and the GNSS payload, and the radar sampling frequency and pulse repetition frequency errors. Since Δf r and Δf a have little impact on geometric positioning, the spaceborne SAR geometric calibration models at home and abroad can be simplified as:

[0103]

[0104] S69. Based on the above SAR geometric calibration model, according to the SAR geolocation model and the additional geometric constraint conditions that need to consider the control points, the following conditional equations can be listed:

[0105]

[0106] Among them, Dis(.) can calculate the distance from each control point participating in geometric calibration to the fitted line or circle. Line and Circle respectively represent the above-mentioned fitted line and circle elements. By considering the information of the additional geometric constraints of the control points and applying it in geometric calibration, compared with the method that only considers the position information of the control points, the error is further reduced, and the geometric positioning performance of the SAR is improved by the meter level.

[0107] S7. Solve the SAR geometric calibration model to obtain geometric calibration parameters, thereby realizing SAR geometric calibration.

[0108] S71. Since the above conditional equation is non-linear, the Taylor series is used to expand the conditional equation to the first term with respect to the calibration parameters for linearization.

[0109] S72. At this time, the geometric calibration problem can be transformed into a problem of solving a linear equation system, and the Newton-Raphson method can be used to solve the unknown geometric calibration parameters.

[0110] By adding geometric constraints to the collected control points, the obtained geometric calibration parameters are more reliable, thereby improving the accuracy of geometric calibration.

[0111] S8. Update the initial SAR geolocation model based on the geometric calibration parameters to obtain an updated SAR geolocation model, and use the checkpoint database with additional geometric constraints to evaluate the geolocation errors of the updated SAR geolocation model before and after the update.

[0112] This step is specifically as follows:

[0113] S81. Apply the calculated geometric calibration parameters to the initial SAR geolocation model to obtain an updated SAR geolocation model.

[0114] S82. Evaluate the geolocation error of the updated SAR geolocation model: Use the checkpoint database to calculate the image-space and object-space errors of the updated SAR geolocation model respectively.

[0115] So far, this is the entire process of this embodiment. By comparing the geometric calibration results with additional linear constraints shown in Table 3 and the traditional point-based geometric calibration results, and comparing the geometric calibration results with additional circular constraints shown in Table 4 and the traditional point-based geometric calibration results, Figure 4 and Figure 5 corresponding to Table 3 and Table 4 respectively, it shows that this method has a meter-level accuracy improvement compared with the geometric calibration method that only considers the position information of control points.

[0116] Table 3 Comparison of geometric calibration results with additional linear constraints and traditional point-based geometric calibration

[0117]

[0118] Table 4 Comparison of geometric calibration results with additional circular constraints and traditional point-based geometric calibration

[0119]

[0120] If the above functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0121] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) that contain computer-usable program codes. The solutions in the embodiments of the present invention can be implemented using various computer languages. For example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript.

[0122] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0123] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device that implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable apparatus to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process or a plurality of processes and / or one block or a plurality of blocks. Figure 1 one process or a plurality of processes and / or Figure 1 steps for implementing the functions specified in one block or a plurality of blocks.

[0125] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to cover the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0126] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

[0127] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative efforts. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present invention through logical analysis, reasoning, or limited experiments based on the concept of the present invention on the basis of the prior art should fall within the protection scope determined by the claims.

Claims

1. A spaceborne SAR geometric calibration method with additional geometric constraints on control points, characterized in that, It includes the following steps: Collect control points based on the collected SAR images and DOM, or collect control points according to real prior information; Fit the control points based on a fitting method to obtain additional geometric constraints of the control points; Construct an initial SAR geolocation model according to the SAR image auxiliary file, SAR imaging parameters, and SAR satellite orbit file of the SAR image; Collect homologous image points in the SAR intensity image and DOM after SAR image preprocessing, and establish a control point pair database with pixel coordinates and geodetic coordinates; Calculate the geometric positioning error of the initial SAR geolocation model based on the control point pair database; Based on analyzing the geometric positioning error and SAR imaging mechanism, determine the SAR image geolocation error on the ground, and establish a SAR geometric calibration model based on the initial SAR geolocation model and additional geometric constraints; Solve the SAR geometric calibration model to obtain geometric calibration parameters, thereby realizing SAR geometric calibration.

2. The spaceborne SAR geometric calibration method with additional geometric constraints for control points according to claim 1, characterized in that The additional geometric constraints include additional linear constraints, additional circular constraints, and additional elliptical constraints.

3. The spaceborne SAR geometric calibration method with additional control point geometric constraints according to claim 1, characterized in that The specific steps for obtaining the additional geometric constraints of the control points include: Classify the control points, establish a control point set with implicit constraints, and establish fitting models respectively; Fit the control point set based on the fitting model to obtain statistical parameters; Calculate the fitting residual of each control point based on the statistical parameters; Based on the fitting residual, reject the point set with a gross error rejection threshold greater than the preset threshold and the number of control points less than the preset number, and finally form the additional geometric constraints of the control points.

4. The spaceborne SAR geometric calibration method with additional control point geometric constraints according to claim 1, characterized in that, The initial SAR geolocation model is: Wherein, R is the slant range between the observed target and the sensor, (Xs, Ys, Zs) are the spatial rectangular coordinates of the sensor at the imaging moment of a certain target, (X g , Y g , Z g ) are the object space coordinates of a certain target on the ground, that is, the collected control points, R0 is the starting slant range in the range direction, x is the image space coordinate in the range direction, (x, y) are the high-precision sub-pixel coordinates collected in the SAR image, P r is the slant range resolution in the range direction; f d is the Doppler center frequency existing between the satellite platform and the observed target at the imaging moment, P sc and V sc are the position and velocity vectors of the sensor respectively, P gc and V gc are the position and velocity vectors of the observed target respectively; a and b are the major and minor semi-axes of the reference ellipsoid respectively, h is the elevation of the observed target, and λ is the radar wavelength; The conditional equation of the initial SAR geolocation model is: Where x is the column number of the SAR image, c is the speed of light, and f r is the system sampling frequency in the range direction.

5. The spaceborne SAR geometric calibration method with additional control point geometric constraints according to claim 1, characterized in that The control point pair database is divided into a control point database and a check point database.

6. The spaceborne SAR geometric calibration method with additional geometric constraints for control points according to claim 5, characterized in that, The geometric positioning error includes object space positioning error and image space positioning error. The object space positioning error is the difference between the pixel coordinates of the control point pair converted to geodetic coordinates and the geodetic coordinates of the control point pair, and the image space positioning error is the difference between the geodetic coordinates of the control point pair converted to pixel coordinates and the pixel coordinates of the control point pair.

7. The spaceborne SAR geometric calibration method with additional control point geometric constraints according to claim 1, characterized in that, The expression of the SAR geometric calibration model is: Where, i is the column coordinate of the point target in the SAR image, j is the row coordinate of the point target in the SAR image, and t r is the imaging time of the i-th column of the SAR image, and t r0 is the imaging time of the first column of the SAR image, Δtr is the correction value of the slant range imaging time, and t delay is the atmospheric delay time, and f r is the system range sampling frequency, and t a is the imaging time of the j-th row of the SAR image, Δta is the correction value of the azimuth imaging time, and f a is the pulse repetition frequency of the SAR satellite; The conditional equation of the SAR geometric calibration model is: where t r0 is the imaging time of the first column of the SAR image, c is the speed of light, t delay is the calculated atmospheric delay time, Δt r is the SAR slant range time correction value, x is the SAR image column coordinate, f r is the system range sampling frequency, λ is the radar wavelength, R sc and V sc are the SAR satellite orbital position vector and velocity vector respectively, R gc and V gc are the position vector and velocity vector of the ground point target respectively. Dis(.) calculates the distance from each control point participating in geometric calibration to the fitted geometric element.

8. The spaceborne SAR geometric calibration method with additional control point geometric constraints according to claim 1, characterized in that The SAR image geolocation error on the ground includes internal electronic time delay, time synchronization error between the SAR payload and the GNSS payload, radar sampling frequency, and pulse repetition frequency error.

9. The spaceborne SAR geometric calibration method with additional control point geometric constraints according to claim 7, characterized in that, The specific steps for obtaining geometric calibration parameters include: Expand the geometric calibration parameters of the conditional equation to the first term using Taylor series for linearization; Solve the geometric calibration parameters using the Newton-Raphson method.

10. The spaceborne SAR geometric calibration method with additional geometric constraints on control points according to claim 5, characterized in that, It also includes updating the initial SAR geolocation model based on the geometric calibration parameters, obtaining an updated SAR geolocation model, and evaluating the geolocation error before and after the update of the updated SAR geolocation model using the check point database.

Images