SAR cross-radiation calibration method based on offshore wind farm

Abstract

This invention discloses a SAR cross-radiometry calibration method based on offshore wind farms. It acquires calibrated and uncalibrated time-series SAR satellite data of the same offshore wind farm area using the same imaging parameters from the same satellite. The calibrated time-series SAR satellite data is cropped according to the latitude and longitude of the offshore wind farm, and the radar cross-section of the offshore wind farm in the calibrated SAR data is calculated. The uncalibrated time-series SAR satellite data is then cropped, and the response energy of the offshore wind farm in the cropped image is calculated using an integral method. Based on the relationship between the response energy, radar cross-section, and calibration coefficients, the radar cross-section of the offshore wind farm in the uncalibrated time-series SAR satellite data is calculated and compared with the radar cross-section of the offshore wind farm calculated from the calibrated SAR data to obtain the RCS deviation. This invention results in a more concentrated distribution of offshore wind farm clusters, a simpler RCS calculation process, and a more accurate response energy calculation process.

Description

Technical Field

[0001] This invention relates to the field of SAR satellite radiometric calibration technology, and in particular to a SAR cross-radiometric calibration method based on offshore wind farms. Background Technology

[0002] Synthetic Aperture Radar (SAR) is a radar technology that uses radar beams to synthesize a virtual large-aperture antenna, enabling high-resolution imaging even when the radar platform is in motion. It possesses all-weather, all-time observation capabilities. Since the launch of the first SAR satellite, Seasona-A, SAR technology has continuously developed, and the amount of data has steadily increased. SAR data was initially used for qualitative remote sensing of targets based on images, utilizing information such as location, texture, and shape. However, in some monitoring fields (such as water pollution and pest monitoring), traditional qualitative remote sensing methods are insufficient. Therefore, it is necessary to extract the backscattering coefficients of ground objects from SAR data to achieve more accurate quantitative processing, i.e., quantitative remote sensing. SAR radiometric calibration is crucial for SAR quantitative processing. SAR radiometric calibration establishes an accurate relationship between the digital quantization (DN) values ​​recorded in SAR images and the backscattering coefficients of ground objects, unifying various types of SAR data under the same measurement standard and ensuring the accuracy, consistency, and comparability of measurement results.

[0003] Existing SAR radiometric calibration techniques mostly use artificial point targets with known radar cross-sections (RCS) as reference targets. Artificial point targets include corner reflectors and active corner reflectors. For example, the European Space Agency's Sentinel-1 satellite used corner reflectors installed in the Surat Basin, Queensland, Australia for radiometric calibration, while Japan's ALOS PALSAR-2 used corner reflectors deployed in Tomakomai, Hokkaido for calibration. Since the radar cross-section of artificial point targets is known, calibration can be achieved by deploying a certain number of them. However, these artificial point targets are large and inconvenient to transport, and their deployment and maintenance are labor-intensive and costly, resulting in limited use throughout the SAR's lifecycle.

[0004] Besides artificial point targets, natural point targets are increasingly being used in SAR radiometric calibration due to their advantages such as requiring no manual deployment and maintenance and their large quantity. Existing techniques for SAR radiometric calibration using natural point targets primarily focus on communication towers, offshore drilling platforms, parabolic antennas, and power transmission towers, as shown in Table 1. However, techniques for SAR radiometric calibration using natural point targets have three technical drawbacks:

[0005] 1. The distribution of natural point targets is scattered. In radiometric calibration, a greater number of natural point targets in the same image can provide higher calibration accuracy. However, the current distribution of natural point targets is scattered, with a small number within the same image area, resulting in large calibration errors.

[0006] 2. Obtaining the true RCS of natural point targets is difficult. Relying solely on theoretical modeling and calculation ignores errors present in actual measurements. Methods that rely on deploying corner reflectors with known RCS within the same image area require manual alignment, a complex process.

[0007] 3. The traditional method of determining the integration region using the center-maximum-value method in the point target response energy calculation process results in an insufficiently concentrated integration region, leading to errors in the extracted response energy. In an ideal point target SAR image, an artificial point target appears as a cross-shaped spot, and the center of the spot is the phase center and also the point with the maximum DN value. Therefore, traditional methods often use the center-maximum-value method to determine the center position of the integration region. However, natural point targets differ from artificial point targets; in SAR images, the maximum value is often not located in the center. Using the traditional integration method results in insufficient energy concentration in the integration region, introducing errors into the extracted response energy.

[0008] Table 1. Advantages and disadvantages of existing SAR radiometric calibration techniques using natural point targets.

[0009] Summary of the Invention

[0010] To address the aforementioned technical issues, this invention proposes a SAR cross-radiative calibration method based on offshore wind farms, which features a more concentrated distribution of offshore wind turbine clusters, simplifies the RCS calculation process, and makes the response energy calculation process more accurate.

[0011] To achieve the above objectives, the technical solution of the present invention is as follows:

[0012] The SAR cross-radiometric calibration method based on offshore wind farms includes the following steps:

[0013] Acquire calibrated time-series SAR satellite data and uncalibrated time-series SAR satellite data of the same offshore wind farm area using the same imaging parameters from the same satellite.

[0014] The scope of the offshore wind farm is determined based on latitude and longitude, and the calibrated time-series SAR satellite data is calibrated based on the row and column values ​​of the calibrated scope. The radar cross-section of the offshore wind turbine in the calibrated SAR data is then calculated.

[0015] Based on the latitude and longitude of offshore wind farms, the time-series SAR satellite data to be calibrated is cropped. The response energy of the offshore wind farm in the cropped image is calculated using the integral method. The radar cross section of the offshore wind farm in the time-series SAR satellite data to be calibrated is calculated based on the relationship between the response energy, radar cross section and calibration coefficient. The radar cross section of the offshore wind farm calculated in the calibrated SAR data is compared with that of the offshore wind farm calculated in the calibrated SAR data to obtain the RCS deviation.

[0016] Preferably, the step of determining the scope of the offshore wind farm and cropping the calibrated time-series SAR satellite data according to the latitude and longitude of the offshore wind farm, and calculating the radar cross-section of the offshore wind turbines in the calibrated SAR data, includes the following steps:

[0017] The calibrated time-series SAR satellite data includes N calibrated SAR satellite images at N time points, with the time span of the N time points being no less than 12 months.

[0018] Based on the latitude and longitude of the offshore wind farm, the calibrated time-series SAR satellite data is cropped, and the signal-to-noise ratio of each offshore wind turbine in each cropped offshore wind farm image is calculated sequentially. If the signal-to-noise ratio of the offshore wind turbine is greater than the preset value, the offshore wind turbine is retained in the current image; otherwise, it is deleted. The radar cross-section of each offshore wind turbine in the current image is calculated based on the retained offshore wind turbines.

[0019] The mean signal-to-noise ratio (SNR) and the sum of radar cross sections of each offshore wind farm image are statistically analyzed. Based on the mean SNR and the sum of radar cross sections of each offshore wind farm image, the calculated mean SNR, the calculated mean radar cross section, and the normalized standard deviation of the radar cross section of the offshore wind farm in N calibrated SAR satellite images are obtained.

[0020] Determine whether the mean signal-to-noise ratio (SNR) and normalized standard deviation of the radar cross-section of offshore wind farms in N calibrated SAR satellite images meet the requirements for engineering applications. If they do, use the mean of the radar cross-section as the true value and compare it with the radar cross-section of the offshore wind farm in the uncalibrated time-series SAR satellite data to obtain the RCS deviation. If they do not meet the requirements, cut off the range of another offshore wind farm and reanalyze.

[0021] Preferably, before cropping the calibrated time-series SAR satellite data based on the latitude and longitude of the offshore wind farm, the following steps are also included:

[0022] Preprocessing of calibrated time-series SAR satellite data.

[0023] Preferably, the preprocessing includes orbit correction, radiometric calibration, and terrain correction.

[0024] Preferably, the formula for calculating the signal-to-noise ratio of each offshore wind turbine in each offshore wind farm image is as follows:

[0025]

[0026] In the formula, SCR is the signal-to-noise ratio of each offshore wind turbine; For target RCS; The background average clutter RCS.

[0027] Preferably, the formula for calculating the normalized standard deviation of the radar cross-section is as follows:

[0028]

[0029] In the formula, STD is the normalized standard deviation of the RCS of offshore wind farms in time-series SAR satellite data; x i It is the normalized RCS value of the i-th offshore wind farm in the time-series SAR satellite data; is the normalized RCS mean of offshore wind farms in time-series SAR satellite data; N is the number of offshore wind turbines.

[0030] Preferably, the step of calculating the response energy of the offshore wind farm in the cropped image using the integral method includes the following steps:

[0031] In selecting the integration region, a custom mask is selected based on the shape of the windmill's SAR image to determine the centroid. The integration region is then redefined with the centroid as the center. The specific formula is as follows:

[0032]

[0033] In the formula, ε represents the windmill response energy; N A DN is the number of pixels in the windmill's response energy integral; N is the digital quantization value of a specific pixel. B Δa is the number of pixels in the background area; Δr is the azimuth resolution and Δr is the range resolution.

[0034] Based on the above technical solutions, the beneficial effects of this invention are as follows: To address the technical deficiency of dispersed distribution of natural point targets, this invention selects concentrated offshore wind farms with a large number within the SAR image range as targets for SAR radiometric calibration; to address the technical deficiency of difficulty in calculating the RCS of natural point targets, it calculates the RCS of offshore wind farms by acquiring calibrated time-series SAR satellite data of the same offshore wind farm; and to address the technical deficiency of large selection errors in the integration region of natural point target response energy, it determines the integration region by constructing a mask with a shape similar to that of the offshore wind farm SAR image. Compared with existing methods for radiometric calibration using natural point targets, the offshore wind farm clusters in this invention are more concentrated, the RCS calculation process is simpler, and the response energy calculation process is more accurate. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating a SAR cross-radiometry calibration method based on an offshore wind farm in one embodiment.

[0036] Figure 2 This is a schematic diagram of the mask in a SAR cross-radiometry calibration method based on an offshore wind farm in one embodiment;

[0037] Figure 3This is a schematic diagram illustrating the calculation of response energy using the integral method in a SAR cross-radiometry calibration method based on an offshore wind farm in one embodiment;

[0038] Figure 4 This is a schematic diagram of the location of an offshore wind farm in an embodiment of a SAR cross-radiometry calibration method based on offshore wind farms;

[0039] Figure 5 This is a comparison of the RCS obtained by calculating the response energy using the ordinary integration method and the masking method. Detailed Implementation

[0040] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0041] like Figures 1 to 5 As shown, this embodiment provides a SAR cross-radiometry calibration method based on offshore wind farms, including the following steps:

[0042] Step 100: Obtain calibrated time-series SAR satellite data and uncalibrated time-series SAR satellite data of the same offshore wind farm area using the same imaging parameters from the same satellite.

[0043] In this embodiment, calibrated time-series SAR satellite data is acquired. The calibrated SAR satellite data must originate from the same satellite, have the same ascent and descent orbits, cover the same offshore wind farm area, and span a time period of more than 12 months. The data to be calibrated is also acquired. This data must originate from the same satellite as the calibrated SAR satellite data, have the same as ascent and descent orbits, and cover the same offshore wind farm area. Specifically, the example data used is Sentinel-1 satellite imagery data. Sentinel-1 data acquisition is a collaborative effort between two polar-orbiting satellites equipped with C-band synthetic aperture radar. Compared to other on-orbit radar satellites, Sentinel-1 satellites have higher radiometric accuracy, with an absolute radiometric accuracy of 1 dB. Compared to other spaceborne SAR satellites, Sentinel-1 satellite data better meets the requirements for SAR scattering stability characteristics in absolute radiometric calibration, facilitating further exploration of the radiometric accuracy of SAR data.

[0044] Sentinel-1 satellite data is freely available, has high absolute radiometric calibration accuracy, a large volume of publicly available data, and a long time span, meeting the data selection requirements for this technology. The location of offshore wind farms was queried using the 4COffshore website (https: / / map.4coffshore.com / ). 55 wind turbines from a specific offshore wind farm were selected to further determine its specific latitude and longitude, and SAR satellite imagery data containing the wind farm was downloaded. The calibrated SAR satellite data used was Sentinel-1A satellite data, covering the period from April 2021 to April 2022, with one image per month. The SAR satellite data to be calibrated used was Sentinel-1A satellite data, covering the period from November 2022 to January 2023, with one image per month. Specific data selection parameters are shown in Table 2.

[0045] Table 2. Detailed information on calibrated and uncalibrated SAR satellite data.

[0046]

[0047] Step 200: Determine the range of the offshore wind farm based on latitude and longitude, and cut the calibrated time-series SAR satellite data according to the row and column values ​​of the cut range, and calculate the radar cross-section of the offshore wind turbine in the calibrated SAR data.

[0048] Step 201: Obtain the backscattering coefficients of calibrated SAR satellite data using software. Preprocess the pixel values ​​of the obtained calibrated satellite data to acquire the backscattering coefficients. Preprocessing includes: orbit correction, radiometric calibration, and terrain correction; 1) Orbit correction: The accuracy of satellite orbital status information in the metadata files of synthetic aperture radar products is not very high. If precise orbital information is required, it takes two weeks to obtain it. For processing requiring high registration accuracy, orbit correction is necessary; 2) Radiometric correction: To reduce storage, the data before calibration is encoded and does not display backscattering coefficients. To obtain the backscattering coefficients, radiometric calibration converts the received backscattered signal (energy) into a physical quantity with units (some are unitless proportional values), such as the backscattering coefficients; 3) Terrain correction: Improving the image with actual coordinate information and correcting the image terrain allows for better location of the corresponding target in the SAR satellite data using latitude and longitude coordinates.

[0049] In this embodiment, Sentinel-1 radiometric calibration is a mature technology that can be performed directly using SNAP software. Due to the cloud penetration of SAR data, Sentinel-1 data does not require atmospheric correction of optical images; therefore, radiometric calibration is performed directly via SNAP. Range-Doppler topographic correction is applied to the calibrated SAR satellite data image. Geocoding and topographic radiometric correction are then performed to obtain the backscattering coefficients of the calibrated SAR satellite data.

[0050] Step 202: Crop data slices containing offshore wind farms. Open the calibrated SAR satellite imagery data, preliminarily determine the location of a certain offshore wind farm using latitude and longitude, select the area containing the offshore wind farm on the calibrated SAR satellite imagery, record the row and column values ​​of the area containing the offshore wind farm, and crop it using SNAP software.

[0051] Step 203: Comparison of signal-to-noise ratio (SNR) of offshore windmills in calibrated SAR satellite data. Not all offshore windmills are suitable for cross-radiometry calibration. To ensure effective extraction of offshore windmills from the background, sufficient visibility of the windmills relative to the background is required. SNR is used to measure the visibility of offshore windmills relative to the background; the specific calculation formula is as follows:

[0052]

[0053] In the formula, SCR is the signal-to-noise ratio of each offshore wind turbine; For target RCS; The background average clutter RCS is used. Typically, the ratio of the peak power of the offshore wind turbine to the average background clutter power is used. The SCR is calculated for each turbine in an offshore wind farm; a turbine with a signal-to-noise ratio greater than 20 dB can be considered a valid reference point.

[0054] Step 204: Obtain the RCS of offshore wind turbines from the calibrated SAR satellite data. Calculate the RCS of each offshore wind turbine based on the backscattering coefficient. The specific formula is as follows:

[0055]

[0056] In the formula I P It is the energy of the main lobe of the offshore windmill; C F It is the relative power of the side lobes of the offshore wind turbine; P A It is the pixel area of ​​calibrated SAR satellite data; K 已定标 These are the scaling coefficients; θ is the angle of incidence; A dn It is the scaling factor, which is a constant of 474.

[0057] During RCS calculation, the main lobe energy I PThe calculation process involves first extracting a 256*256 sub-image centered on the offshore windmill; then converting the image pixel values ​​into intensity values ​​through squaring; and finally selecting four background regions on the sub-image as representative samples of background intensity, such as... Figure 2 As shown; the pixel values ​​in the background area are summed and then divided by the pixel value to calculate the average background pixel intensity value; then, the image is interpolated based on experience; the average background pixel intensity value is subtracted from each pixel value of the sub-image; finally, the main lobe energy is calculated within a 3dB range with the offshore windmill as the center; the relative power C of the offshore windmill's side lobes is also calculated. F The calculation formula is:

[0058]

[0059] In the formula, ISLR is the integral sidelobe ratio.

[0060] Calculate the RCS of the same offshore wind farm in 13 calibrated SAR satellite images.

[0061] Step 205: Compare the time-series stability of RCS of offshore wind turbines in calibrated SAR satellite data. Analyze the changes in RCS of offshore wind turbines in the time-series calibrated SAR satellite data. Since the wind turbines are constantly running, the RCS values ​​of the offshore wind turbines obtained when the SAR passes overhead are different. According to the law of large numbers, when there are enough offshore wind turbines, the RCS of the offshore wind farm is stable to a certain extent. The smaller the change in RCS of the offshore wind farm in the calibrated SAR satellite data, the stronger the time-series stability of the offshore wind farm.

[0062] To assess the temporal stability of RCS, the sum of RCS is normalized, and the standard deviation (STD) is used as the evaluation criterion. The specific calculation formula is as follows:

[0063]

[0064] In the formula, STD is the normalized standard deviation of the sum of the RCS of 50 wind turbines in a certain offshore wind farm in time-series SAR satellite data; x i It is the normalized RCS value of the i-th wind turbine in a certain offshore wind farm in time-series SAR satellite data; It is the mean of the sum of normalized values ​​of the RCS of a certain offshore wind farm in time-series SAR satellite data; N is the number of wind turbines, 50.

[0065] Observing the number of offshore wind turbines with a SCR less than 20dB in 13 image scenes, the RCS value is relatively small. In a certain offshore wind farm, there are approximately 3 turbines with an SCR less than 20dB in each image scene. In addition, there is also interference from ships at sea, which is reflected in an abnormally large RCS value. In the 13 image scenes, there are approximately 2 wind turbines affected by interference. In subsequent calculations, the 3 smallest and 2 largest outliers were removed, and the remaining 50 wind turbines were summed. The mean SCR of a certain offshore wind farm in the 13 image scenes was calculated, and the normalized standard deviation of the RCS sum was calculated. The results are shown in Table 3.

[0066] Table 3 shows the SCR and RCS results of a certain offshore wind farm based on calibrated SAR satellite data.

[0067]

[0068] The average SCR of offshore wind turbines in a certain offshore wind farm among the 13 calibrated images is 24.62 dB, which is greater than 20 dB, indicating that the background noise has little impact on the offshore wind turbines and meets the requirements for engineering applications. The standard deviation of RCS is 0.36, indicating good stability.

[0069] If the measured RCS standard deviation is too large, add influencing factors such as wind speed and temperature to the analysis. If it still cannot be reasonably explained, replace it with another offshore wind farm.

[0070] Step 300: Based on the latitude and longitude of the offshore wind farm, crop the time-series SAR satellite data to be calibrated, calculate the response energy of the offshore wind farm in the cropped image using the integral method, calculate the radar cross-section of the offshore wind farm in the time-series SAR satellite data to be calibrated based on the relationship between the response energy, radar cross-section, and calibration coefficient, and compare it with the radar cross-section of the offshore wind farm calculated in the calibrated SAR data to obtain the RCS deviation.

[0071] Step 301: Open the SAR satellite image data to be calibrated, preliminarily determine the location of a certain offshore wind farm through latitude and longitude, select the area containing the offshore wind farm on the calibrated SAR satellite image, record the row and column values ​​of the area containing the offshore wind farm, and crop it using SNAP software.

[0072] Step 302: Calculate the response energy of offshore wind farms in the SAR satellite data slices to be calibrated.

[0073] The response energy of offshore wind farms in SAR satellite data slices to be calibrated is obtained by integration. The integration region is selected using a customized 7x5 mask based on the shape of the wind turbine's SAR image, with the centroid chosen as the center. The integration region is then redefined using this centroid as the center. Figure 3 As shown, the specific formula is as follows:

[0074]

[0075] In the formula, ε represents the windmill response energy; N A DN is the number of pixels in the windmill's response energy integral; N is the digital quantization value of a specific pixel. B Δa is the number of pixels in the background region; Δa is the azimuth resolution, and Δr is the range resolution. Corresponding to RCS data filtering, after removing the 3 smallest and 2 largest values, calculate the sum of the response energy of the remaining 50 offshore wind turbines.

[0076] The integral method for calculating the response energy in this invention can also employ the peak method.

[0077] Step 303: Calculate the RCS deviation of offshore wind farms in the SAR satellite data slice to be calibrated.

[0078] The formula for calculating the RCS of offshore windmills in SAR satellite data slices to be calibrated is as follows:

[0079]

[0080] In the formula σ 待定标 This is the RCS of offshore windmills in a calibrated SAR satellite data slice calculated in B4; K 待定标 is the calibration coefficient of the offshore wind farm in the SAR satellite data to be calibrated; E is the response energy of the offshore wind farm in the SAR satellite data slice to be calibrated calculated in C2; dividing the calculated response energy by the system calibration constant gives the RCS of the SAR satellite data to be calibrated, which is compared with the RCS of the offshore wind farm obtained from the time-series SAR satellite data to obtain the RCS deviation. The magnitude of the RCS deviation value is used to measure the effectiveness of using offshore wind farms as natural point targets for SAR radiometric calibration. Specific results are as follows: Figure 5 As shown.

[0081] Depend on Figure 5 It can be seen that the average deviation between the RCS obtained by calculating the response energy of offshore wind farms using the ordinary integration method and the RCS obtained by time-series images is 0.31 dB. The calibration effect using offshore wind farms is good, with an RCS deviation of less than 1 dB, which meets the engineering requirements. The average deviation between the RCS obtained by calculating the response energy of offshore wind farms using the masking method and the RCS obtained by time-series images is 0.17 dB. The masking method calculates the response energy of offshore wind farms more accurately, and the RCS deviation is reduced by an average of 0.14 dB.

[0082] It should be understood that although the steps in the flowchart above are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart above may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0083] The above are merely preferred embodiments of the present application and are not intended to limit the embodiments of the present application. For those skilled in the art, the embodiments of the present application can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of the present application should be included within the protection scope of the embodiments of the present application.

Claims

1. A SAR cross-radiometric calibration method based on offshore wind farms, characterized in that, Includes the following steps: Acquire calibrated and uncalibrated time-series SAR satellite data of the same offshore wind farm area using the same imaging parameters from the same satellite. The scope of the offshore wind farm is determined based on latitude and longitude, and the calibrated time-series SAR satellite data is calibrated based on the row and column values ​​of the calibrated scope. The radar cross-section of the offshore wind turbine in the calibrated SAR data is then calculated. Based on the latitude and longitude of offshore wind farms, the time-series SAR satellite data to be calibrated is clipped. The response energy of the offshore wind farm in the clipped image is calculated using the integral method. The radar cross section of the offshore wind farm in the time-series SAR satellite data to be calibrated is calculated based on the relationship between the response energy, radar cross section and calibration coefficient. The radar cross section of the offshore wind farm calculated in the calibrated SAR data is then compared with that of the offshore wind farm calculated in the calibrated SAR data to obtain the RCS deviation.

2. The SAR cross-radiometric calibration method based on offshore wind farms according to claim 1, characterized in that, The process of determining the scope of the offshore wind farm based on latitude and longitude, cropping the calibrated time-series SAR satellite data based on the row and column values ​​of the cropping scope, and calculating the radar cross-section of the offshore wind turbines in the calibrated SAR data includes the following steps: The calibrated time-series SAR satellite data includes N calibrated SAR satellite images at N time points, with the time span of the N time points being no less than 12 months. Based on the latitude and longitude of the offshore wind farm, the calibrated time-series SAR satellite data is cropped, and the signal-to-noise ratio of each offshore wind turbine in each cropped offshore wind farm image is calculated sequentially. If the signal-to-noise ratio of the offshore wind turbine is greater than the preset value, the offshore wind turbine is retained in the current image; otherwise, it is deleted. The radar cross-section of each offshore wind turbine in the current image is calculated based on the retained offshore wind turbines. The mean signal-to-noise ratio (SNR) and the sum of radar cross sections of each offshore wind farm image are statistically analyzed. Based on the mean SNR and the sum of radar cross sections of each offshore wind farm image, the calculated mean SNR, the calculated mean radar cross section, and the normalized standard deviation of the radar cross section of the offshore wind farm in N calibrated SAR satellite images are obtained. Determine whether the mean signal-to-noise ratio and the normalized standard deviation of the radar cross-section of the offshore wind farm in N calibrated SAR satellite images meet the requirements for engineering applications. If they do, the mean value calculated by the radar cross-section is used as the true value; if they do not, the range of another offshore wind farm is cropped and reanalyzed.

3. The SAR cross-radiometric calibration method based on offshore wind farms according to claim 2, characterized in that, Before cropping the calibrated time-series SAR satellite data based on the latitude and longitude of the offshore wind farm, the following steps are also included: Preprocessing of calibrated time-series SAR satellite data.

4. The SAR cross-radiometric calibration method based on offshore wind farms according to claim 3, characterized in that, The preprocessing includes orbit correction, radiometric calibration, and terrain correction.

5. The SAR cross-radiometry calibration method based on offshore wind farms according to claim 2, characterized in that, The formula for calculating the signal-to-noise ratio of each offshore wind turbine in each offshore wind farm image is as follows: In the formula, SCR is the signal-to-noise ratio of each offshore wind turbine; For target RCS; The background average clutter RCS.

6. The SAR cross-radiometric calibration method based on offshore wind farms according to claim 2, characterized in that, The formula for calculating the normalized standard deviation of the radar cross-section is as follows: In the formula, STD is the normalized standard deviation of the RCS of offshore wind farms in time-series SAR satellite data; x i It is the normalized RCS value of the i-th offshore wind farm in the time-series SAR satellite data; is the normalized RCS mean of offshore wind farms in time-series SAR satellite data; N is the number of offshore wind turbines.

7. The SAR cross-radiometric calibration method based on offshore wind farms according to claim 1, characterized in that, The calculation of the response energy of the offshore wind farm in the cropped image using the integral method includes the following steps: In selecting the integration region, a mask is customized based on the shape of the SAR image of the windmill to select the centroid, and the integration region is redefined with the centroid as the center. The specific formula is as follows: In the formula, ε represents the windmill response energy; N A DN is the number of pixels in the windmill's response energy integral; N is the digital quantization value of a specific pixel. B Δa is the number of pixels in the background area; Δr is the azimuth resolution and Δr is the range resolution.

Images