A SAR image radiation difference correction method, device, equipment and medium

Abstract

The application relates to the technical field of remote sensing monitoring, and provides a SAR image radiation difference correction method, device, equipment and medium. The SAR image radiation difference correction method comprises the following steps: acquiring a plurality of SAR images of a target area, and dividing each SAR image into a plurality of grid images; acquiring the root mean square height corresponding to the grid images; acquiring a diffuse reflection image based on the root mean square heights corresponding to all the grid images; constructing a least square adjustment model based on all the diffuse reflection images; solving the least square adjustment model to obtain the mean correction number and the standard deviation correction number of each SAR image; and correcting the radiation difference of the SAR image based on the mean correction number and the standard deviation correction number of the SAR image. The SAR image radiation difference correction method can improve the quality of radiation difference correction of a large range of SAR images.

Description

Technical Field

[0001] This application relates to the field of remote sensing monitoring technology, and in particular to a method, device, equipment and medium for correcting radiometric differences in SAR images. Background Technology

[0002] Synthetic Aperture Radar (SAR) possesses the capability to conduct all-weather, all-day Earth observation and provide high-resolution images. It is widely used in disaster monitoring, environmental monitoring, marine monitoring, resource exploration, crop yield estimation, surveying and mapping, and military applications. In large-scale studies, the coverage of a single SAR image is insufficient, necessitating image stitching of the SAR images for the study area. Due to seasonal variations, incident angle errors, noise, antenna pattern gain errors, transmit or receive power gain errors, and imaging processor gain errors, significant radiometric differences exist between SAR images from different orbits. This results in poor radiometric continuity and abrupt changes in radiometric brightness between images in the stitched SAR image. To eliminate these radiometric differences and ensure good radiometric consistency in the stitched image, it is necessary to correct for radiometric differences between large-scale SAR images. This is of great significance for the large-scale study and application of SAR images. Existing methods for correcting SAR image radiometric differences include:

[0003] Linear models: These models represent the radiometric relationship between the reference and target images by constructing a linear function, such as the Wallis transform. The Wallis transform adjusts the target image to force its mean and standard deviation to be close to or the same as the reference image. However, this method can lead to "blocking artifacts" in the image, damaging its texture information. Another type of linear transform model is the pseudo-invariant feature method. This method constructs a linear function between the target and reference images using overlapping regions, and then uses least squares to solve for the unknown parameters of the linear function to perform radiometric correction on the target image. This method requires precise image registration and a large number of pseudo-invariant feature points, but some overlapping regions between images lack pseudo-invariant feature points, making it impossible to construct a linear function model.

[0004] Nonlinear models: These models achieve radiometric correction by matching the histograms of the image to be corrected to those of a reference image, making their histogram shapes similar. These models are further divided into local and global histogram matching. Local histogram matching utilizes statistical information from overlapping image regions to match the histograms of the entire image, while global histogram matching relies on the statistical information of the entire image. When there are significant differences in the histogram shapes between images, both local and global histogram matching may damage the radiometric characteristics of the images.

[0005] Other models: The simplest radiometric difference correction is to feather the boundaries of the overlapping areas of the images, but this does not correct the radiometric differences in other areas; another method is to construct polygon curves from the overlapping areas of the images to obtain the radiometric gain correction coefficients of the images, but this method requires strict radiometric calibration of the images and has a low degree of automation; in addition, there is the random cross-observation method, which constructs a linear correction function based on the information of the overlapping areas of the images and uses Gaussian filtering to separate the high and low frequency information of the images, thus preserving the texture information of the images, but when the overlapping areas do not meet the conditions for constructing a linear function, this method cannot achieve the expected results.

[0006] Based on the analysis of existing methods for radiometric difference correction of SAR images, it can be seen that these methods all have inherent defects. They cannot obtain large-scale mosaic SAR images with good radiometric consistency, thus affecting the interpretation and application of large-scale SAR images. There is a problem of low quality in radiometric difference correction of large-scale SAR images. Summary of the Invention

[0007] This application provides a method, apparatus, device, and medium for SAR image radiometric difference correction, which can solve the problem of low quality in radiometric difference correction of large-area SAR images.

[0008] In a first aspect, embodiments of this application provide a SAR image radiometric difference correction method, which includes:

[0009] Acquire multiple SAR images of the target area and divide each SAR image into multiple grid images;

[0010] For each grid image, obtain the root mean square height corresponding to the grid image;

[0011] For each SAR image, the diffuse reflectance image in the SAR image is obtained based on the root mean square height corresponding to all grid images in the SAR image;

[0012] Based on all diffuse reflectance images, a least squares adjustment model is constructed to solve for the mean correction of SAR images and a least squares adjustment model is constructed to solve for the standard deviation correction of SAR images; the mean correction is used to correct the mean of SAR images and the standard deviation correction is used to correct the standard deviation of SAR images.

[0013] The least squares adjustment model used to solve for the mean correction of SAR images is solved to obtain the mean correction for each SAR image.

[0014] The least squares adjustment model used to solve the standard deviation correction of SAR images is solved to obtain the standard deviation correction of each SAR image.

[0015] For each SAR image, the radiometric differences of the SAR image are corrected based on the mean correction and standard deviation correction corresponding to the SAR image.

[0016] Optionally, obtain the root mean square height corresponding to the grid image, including:

[0017] Through the formula:

[0018]

[0019] Calculate the root mean square height s of the grid image;

[0020] Among them, Z i Let represent the elevation of the i-th pixel in the grid image, where i = 1, 2, ..., N, and N represents the total number of pixels in the grid image. This represents the average elevation of all pixels in the grid image.

[0021] Optionally, based on the root mean square height corresponding to all grid images in the SAR imagery, the diffuse reflectance imagery in the SAR imagery is obtained, including:

[0022] For each grid image, the root mean square height of the grid image is added to the region where the grid image is located in the geographic coordinate system to obtain the surface roughness of the region.

[0023] Determine whether the surface roughness is greater than or equal to the preset surface roughness value;

[0024] If so, the grid image corresponding to the surface roughness is taken as the diffuse grid image;

[0025] All diffuse reflectance grid images are integrated into a single image to obtain diffuse reflectance images from SAR imagery.

[0026] Optionally, based on all diffuse reflectance images, a least squares adjustment model is constructed to solve for the mean correction of SAR images and a least squares adjustment model is constructed to solve for the standard deviation correction of SAR images, including:

[0027] For each diffuse image, iterate through each other diffuse image. If the regions corresponding to the diffuse image and other diffuse images overlap, then the diffuse image and other diffuse images are considered as an overlapping pair corresponding to the diffuse image.

[0028] Based on the overlapping regions corresponding to all overlapping pairs, a least squares adjustment model is constructed to solve for the mean correction of SAR images and a least squares adjustment model is constructed to solve for the standard deviation correction of SAR images.

[0029] Optionally, the least squares adjustment model used to solve for the mean correction of SAR images is:

[0030] C·X u -L u =0

[0031] Where, C = [A T B T ] T , C represents the parameter matrix, X u L represents the mean correction matrix. u Let A be the mean constant matrix, B be the correction coefficient matrix, and L be the reference coefficient matrix. A,u L represents the correction mean constant matrix. B,u Represents the reference mean constant matrix:

[0032]

[0033]

[0034]

[0035]

[0036] L B,u =|0 … 0| T

[0037] Where u'0 represents the mean correction of the 0th SAR image, u' M-1 This represents the mean correction for the (M-1)th SAR image. Elements in matrix A are correction coefficients, taking values ​​of 1, -1, or 0. 0,0 a 0,M-1 a N-1,0 and a N-1,M-1 All elements are from matrix A, and the elements in matrix B are reference coefficients, taking values ​​of 0 or 1. 0,0 b 0,M-1 b M-1,0 and b M-1,M-1 All are elements in matrix B. This represents the mean value of the overlapping region in the first diffuse image of the 0th overlapping pair. This represents the mean value of the overlapping region in the second diffuse image of the 0th overlapping pair. This represents the mean value of the overlapping region in the first diffuse reflection image of the (N-1)th overlapping pair. This represents the mean value of the overlapping region in the second diffuse reflection image of the N-1th overlapping pair.

[0038] Optionally, the least squares adjustment model used to solve for the standard deviation correction of SAR images is:

[0039] C·Xδ -L δ =0

[0040] Where, C = [A T B T ] T , C represents the parameter matrix, X δ L represents the standard deviation correction matrix. δ L represents the standard deviation constant matrix. A,δ L represents the correction standard deviation constant matrix. B,δ Represents the reference standard deviation constant matrix:

[0041]

[0042]

[0043] L B,δ =[0 … 0] T

[0044] Where δ'0 represents the standard deviation correction of the 0th SAR image, δ' M-1 This represents the standard deviation correction for the (M-1)th SAR image. This represents the standard deviation of the overlapping region in the first diffuse image of the 0th overlapping pair. This represents the standard deviation of the overlapping region in the second diffuse image of the 0th overlapping pair. This represents the standard deviation of the overlapping region in the first diffuse image of the (N-1)th overlapping pair. It represents the standard deviation of the overlapping area in the second diffuse reflection image of the N-1th overlapping pair.

[0045] Optionally, the least squares adjustment model used to solve for the mean correction of SAR images is solved to obtain the mean correction for each SAR image, including:

[0046] Through the formula:

[0047] X u =(C T PC) -1 C R PL u

[0048] Calculate the mean correction matrix;

[0049] Where P represents the weight matrix:

[0050]

[0051] Where W represents the identity matrix, O represents the zero matrix, and W represents the weight matrix of the overlapping pairs:

[0052]

[0053] Where w0 represents the weight of the 0th overlapping pair, w n w represents the weight of the nth overlapping pair. N-1 The weight of the (N-1)th overlapping pair is:

[0054]

[0055] Among them, P n P represents the number of pixels in the overlapping region of the nth overlapping pair. j This represents the number of pixels in the overlapping region of the j-th overlapping pair, where n,j = 0, 1, ..., N-1, and N represents the total number of overlapping pairs.

[0056] The least squares adjustment model used to solve for the standard deviation correction of SAR images is solved to obtain the standard deviation correction for each SAR image, including:

[0057] Through the formula:

[0058] X δ =(C T PC) -1 C T PL δ

[0059] Calculate the standard deviation correction matrix.

[0060] Optionally, the radiometric differences of the SAR image are corrected based on the mean correction and standard deviation correction corresponding to the SAR image, including:

[0061] Through the formula:

[0062]

[0063] Correct the mean of all SAR images;

[0064] in, This represents the mean value of the 0th SAR image after correction. This represents the mean value of the (M-1)th SAR image after correction. This represents the mean of the 0th SAR image. This represents the mean of the (M-1)th SAR image;

[0065] Through the formula:

[0066]

[0067] Correct the standard deviation of all SAR images;

[0068] in, This represents the standard deviation of the 0th SAR image after correction. This represents the standard deviation of the (M-1)th SAR image after correction. This represents the standard deviation of the 0th SAR image. This represents the standard deviation of the (M-1)th SAR image;

[0069] Through the formula:

[0070]

[0071] Correct the radiance value of the r-th SAR image at (x,y);

[0072] Where, f(x,y) r Let g(x,y) be the corrected radiance value of the r-th SAR image at (x,y), where x represents the horizontal axis, y represents the vertical axis, r = 0, 1, ..., M-1, and M represents the total number of SAR images. r Let represent the radiance value of the i-th SAR image at (x, y). Let represent the mean value of the r-th SAR image. This represents the standard deviation of the r-th SAR image after correction. This represents the standard deviation of the r-th SAR image. This represents the mean value of the r-th SAR image after correction.

[0073] Secondly, embodiments of this application provide a SAR image radiometric difference correction device, comprising:

[0074] The module is divided into sub-modules to acquire multiple SAR images of the target area and divide each SAR image into multiple grid images.

[0075] The first acquisition module acquires the root mean square height of each grid image.

[0076] The second acquisition module acquires the diffuse reflectance image in each SAR image based on the root mean square height of all grid images in the SAR image.

[0077] The module constructs a least-squares adjustment model for calculating the mean correction of SAR images and a least-squares adjustment model for calculating the standard deviation correction of SAR images, based on all diffuse reflectance images. The mean correction is used to correct the mean of SAR images, and the standard deviation correction is used to correct the standard deviation of SAR images.

[0078] The first solution module solves the least squares adjustment model used to solve the mean correction of SAR images, and obtains the mean correction of each SAR image.

[0079] The second solution module solves the least squares adjustment model used to solve the standard deviation correction of SAR images, and obtains the standard deviation correction of each SAR image.

[0080] The correction module corrects the radiometric differences of each SAR image based on the mean correction and standard deviation correction corresponding to the SAR image.

[0081] Thirdly, embodiments of this application provide a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the SAR image radiometric difference correction method described above.

[0082] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned SAR image radiometric difference correction method.

[0083] The above-mentioned solution in this application has the following beneficial effects:

[0084] In this embodiment, multiple SAR images of the target area are acquired, and each SAR image is divided into multiple grid images. Then, for each grid image, the root mean square height corresponding to the grid image is obtained. Then, for each SAR image, based on the root mean square heights corresponding to all grid images in the SAR image, the diffuse reflectance image in the SAR image is obtained. Then, based on all diffuse reflectance images, a least squares adjustment model for solving the mean correction of the SAR image and a least squares adjustment model for solving the standard deviation correction of the SAR image are constructed. Then, the least squares adjustment model for solving the mean correction of the SAR image is solved to obtain the mean correction of each SAR image. At the same time, the least squares adjustment model for solving the standard deviation correction of the SAR image is solved to obtain the standard deviation correction of each SAR image. Finally, for each SAR image, the radiometric difference of the SAR image is corrected based on the mean correction and standard deviation correction corresponding to the SAR image. This method divides each SAR image into multiple grid images, enabling analysis of large-scale SAR images. Based on the diffuse reflection region, a minimum adjustment model is constructed, which can analyze the radiometric differences in SAR images caused by different viewing angles. It also preserves the texture information of SAR images, ensuring that the information in the radiometrically corrected SAR images is completely preserved, thus improving the quality of radiometric difference correction for large-scale SAR images.

[0085] Furthermore, since dividing SAR images into gridded images is not limited by the size of SAR images, it is also applicable to small-scale SAR images.

[0086] Other beneficial effects of this application will be described in detail in the following detailed description section. Attached Figure Description

[0087] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0088] Figure 1 A flowchart of a SAR image radiometric difference correction method provided in an embodiment of this application;

[0089] Figure 2 This is a schematic diagram of surface roughness provided in an embodiment of this application;

[0090] Figure 3 This is a schematic diagram of a diffuse reflection image provided in an embodiment of this application;

[0091] Figure 4 A schematic diagram of the structure of a SAR image radiometric difference correction device provided in an embodiment of this application;

[0092] Figure 5 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. Detailed Implementation

[0093] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0094] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0095] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0096] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0097] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0098] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0099] To address the low quality issue in existing radiometric differential correction methods for large-scale SAR images, this application provides a SAR image radiometric differential correction method. This method acquires multiple SAR images of a target area, divides each SAR image into multiple grid images, and then obtains the root mean square (RMS) height corresponding to each grid image. Next, for each SAR image, based on the RMS heights of all grid images, it obtains the diffuse reflectance image within the SAR image. Finally, based on all diffuse reflectance images, it constructs a... This paper proposes a least-squares adjustment model for calculating the mean correction and a least-squares adjustment model for calculating the standard deviation correction of SAR images. The least-squares adjustment model for calculating the mean correction is then solved to obtain the mean correction for each SAR image. Simultaneously, the least-squares adjustment model for calculating the standard deviation correction is also solved to obtain the standard deviation correction for each SAR image. Finally, for each SAR image, radiometric differences are corrected based on the corresponding mean and standard deviation corrections. Specifically, each SAR image is divided into multiple grid images to enable analysis of large-scale SAR images. A least-squares adjustment model based on diffuse reflection regions can analyze radiometric differences in SAR images caused by different viewing angles, while preserving the texture information of the SAR images. This ensures that the information in the radiometrically corrected SAR images is completely preserved, improving the quality of radiometric difference correction for large-scale SAR images.

[0100] Furthermore, since dividing SAR images into gridded images is not limited by the size of SAR images, it is also applicable to small-scale SAR images.

[0101] The following is an exemplary description of the SAR image radiation difference correction method provided in this application.

[0102] like Figure 1 As shown, the SAR image radiometric difference correction method provided in this application includes the following steps:

[0103] Step 11: Acquire multiple SAR images of the target area and divide each SAR image into multiple grid images.

[0104] It should be noted that the target area mentioned above is the area that needs to be corrected for SAR image radiometric differences. The SAR image of the target area can be obtained by stitching together multiple small-sized SAR images. Since the satellite orbits corresponding to each small-sized SAR image are different, the stitched SAR image will have poor radiometric continuity and abrupt changes in radiometric brightness. In addition, interference with the system and equipment for acquiring SAR images will also cause the above-mentioned phenomena to occur in the SAR image of the target area.

[0105] In some embodiments of this application, before performing this step, each acquired SAR image is orthorectified using Rational Polynomial Coefficients (RPC) and Digital Elevation Model (DEM) data of the target area, and the DEM of the target area is interpolated to make the resolution of the SAR image the same as the resolution of the DEM, so as to eliminate the parallax caused by ground elevation in the SAR image and obtain elevation data, thus obtaining a SAR image that is convenient for subsequent processing.

[0106] For example, SAR images of the target area can be obtained by accessing the European Space Agency's website, and then the SAR images can be orthorectified and divided into gridded images using computer software such as Gamma for geographic image processing.

[0107] It is worth mentioning that dividing SAR images into multiple grid images enables subsequent processing of large-scale SAR images, and the grid image division is not limited by the size of the SAR image, and is also applicable to small-scale SAR images.

[0108] Step 12: For each grid image, obtain the root mean square height corresponding to the grid image.

[0109] Specifically, through the formula:

[0110]

[0111] Calculate the root mean square height s of the grid image;

[0112] Among them, Z i Let represent the elevation of the i-th pixel in the grid image, where i = 1, 2, ..., N, and N represents the total number of pixels in the grid image. This represents the average elevation of all pixels in the grid image.

[0113] For example, for a 5×5 grid image with 25 pixels, the root mean square height of the grid image is calculated based on the elevation of each of the 25 pixels.

[0114] In some embodiments of this application, the above formula can be run using data processing computer software such as Maltab to calculate the root mean square height corresponding to each grid image.

[0115] It is worth mentioning that obtaining the root mean square height corresponding to the grid image can describe the terrain height of the grid image, which is convenient for subsequent processing steps.

[0116] Step 13: For each SAR image, obtain the diffuse reflection image in the SAR image based on the root mean square height corresponding to all grid images in the SAR image.

[0117] In some embodiments of this application, the steps of obtaining the diffuse reflectance image in the SAR image based on the root mean square height of all grid images in the SAR image for each SAR image specifically include:

[0118] The first step is to add the root mean square height of each grid image to the region where the grid image is located in the geographic coordinate system to obtain the surface roughness of the region.

[0119] The second step is to determine whether the surface roughness is greater than or equal to the preset surface roughness value.

[0120] If so, the grid image corresponding to the surface roughness is used as the diffuse grid image.

[0121] The third step is to integrate all diffuse reflectance grid images onto a single map to obtain diffuse reflectance images from SAR imagery.

[0122] It should be noted that adding the root mean square height to the corresponding region results in a TIFF (Tag Image File Format) file. After obtaining the diffuse image, gross error removal is also required to reduce data errors.

[0123] Depending on the degree of ground roughness, SAR image echo reflection types can be categorized into specular reflection, diffuse reflection, and double echo reflection. Specular reflection appears as dark areas in the image, such as bodies of water; double echo reflection appears as two high points, such as towns. Diffuse reflection areas in SAR images refer to areas with relatively rough ground; the higher the root mean square height (RMS height) value, the rougher the ground. By filtering the RMS height data corresponding to the SAR image grid, when the RMS height value is below a certain threshold, the intensity value of the grid image is set to null, i.e., a specular reflection area. Further gross error removal, including double echo reflection areas, yields a SAR image containing only diffuse reflection areas.

[0124] For example, for the second grid image, the region in the geographic coordinate system is from 20°N to 22°N and 108°E to 109°E. The root mean square height (RMS) of this grid image is 50. Therefore, for the region from 20°N to 22°N and 108°E to 109°E, its RMS height is 50. Adding the RMS height to the corresponding region in the geographic coordinate system yields the surface roughness data for that region. If the preset surface roughness value is 70, and the surface roughness of the region corresponding to the second grid image is 50, it indicates that the region corresponding to the second grid image is a non-diffuse grid image, and this point is assigned a null value. The surface roughness of the region corresponding to the third grid image is 90, indicating that the region corresponding to the third grid image is a diffuse grid image.

[0125] It is worth mentioning that extracting the diffuse reflectance grid image of SAR imagery can remove the non-diffuse reflectance grid imagery from SAR imagery, retaining only the data needed for subsequent steps.

[0126] The following example illustrates this step.

[0127] The TIFF file of surface roughness obtained by adding the root mean square height corresponding to each grid image in the SAR image to the geographic coordinate system is as follows: Figure 2 As shown, based on the preset surface roughness value, the non-diffuse grid images are removed, and the diffuse reflection images obtained from the diffuse grid images are retained. Figure 3 As shown, Figure 3 The black area represents the diffuse reflection area, and the white area represents the non-diffuse reflection area.

[0128] Step 14: Based on all diffuse reflectance images, construct a least squares adjustment model for solving the mean correction of SAR images and a least squares adjustment model for solving the standard deviation correction of SAR images.

[0129] The mean corrections mentioned above are used to correct the mean of SAR images, and the standard deviation corrections mentioned above are used to correct the standard deviation of SAR images.

[0130] In some embodiments of this application, the steps of constructing a least squares adjustment model for solving the mean correction of SAR images and a least squares adjustment model for solving the standard deviation correction of SAR images based on all diffuse reflectance images specifically include:

[0131] The first step is to iterate through each other diffuse image for each diffuse image. If the regions corresponding to the diffuse images overlap, then the diffuse images and the other diffuse images are considered as an overlapping pair corresponding to the diffuse images.

[0132] For example, if the area corresponding to the first diffuse reflection image partially overlaps with the area corresponding to the second diffuse reflection image, then the first diffuse reflection image and the second diffuse reflection image are considered as an overlapping pair. At the same time, if the area corresponding to the second diffuse reflection image partially overlaps with the area corresponding to the third diffuse reflection image, then the second diffuse reflection image and the third diffuse reflection image are also considered as an overlapping pair.

[0133] The second step involves constructing a least-squares adjustment model for calculating the mean correction of SAR images and a least-squares adjustment model for calculating the standard deviation correction of SAR images, based on the overlapping regions corresponding to all overlapping pairs.

[0134] The least-two adjustment model used to solve for the mean correction of SAR images is as follows:

[0135] C·X u -L u =0

[0136] Where, C = [A T B T ] T , C represents the parameter matrix, X u L represents the mean correction matrix. u Let A be the mean constant matrix, B be the correction coefficient matrix, and L be the reference coefficient matrix. A,u L represents the correction mean constant matrix. B,u Represents the reference mean constant matrix:

[0137]

[0138]

[0139]

[0140]

[0141] L B,u =|0 … 0| T

[0142] Where u'0 represents the mean correction of the 0th SAR image, u' M-1 This represents the mean correction for the (M-1)th SAR image. Elements in matrix A are correction coefficients, taking values ​​of 1, -1, or 0. 0,0 a 0,M-1 a N-1,0 and a N-1,M-1 All elements are from matrix A, and the elements in matrix B are reference coefficients, taking values ​​of 0 or 1. 0,0 b 0,M-1 bM-1,0 and b M-1,M-1 All are elements in matrix B. This represents the mean value of the overlapping region in the first diffuse image of the 0th overlapping pair. This represents the mean value of the overlapping region in the second diffuse image of the 0th overlapping pair. This represents the mean value of the overlapping region in the first diffuse reflection image of the (N-1)th overlapping pair. This represents the mean value of the overlapping region in the second diffuse reflection image of the N-1th overlapping pair.

[0143] The least squares adjustment model used to solve for the standard deviation correction of SAR images is as follows:

[0144] C·X δ -L δ =0

[0145] Where, C = [A T B T ] T , C represents the parameter matrix, X δ L represents the standard deviation correction matrix. δ L represents the standard deviation constant matrix. A,δ L represents the correction standard deviation constant matrix. B,δ Represents the reference standard deviation constant matrix:

[0146]

[0147]

[0148] L B,δ =[0 … 0] T

[0149] Where δ'0 represents the standard deviation correction of the 0th SAR image, δ' M-1 This represents the standard deviation correction for the (M-1)th SAR image. This represents the standard deviation of the overlapping region in the first diffuse image of the 0th overlapping pair. This represents the standard deviation of the overlapping region in the second diffuse image of the 0th overlapping pair. This represents the standard deviation of the overlapping region in the first diffuse image of the (N-1)th overlapping pair. It represents the standard deviation of the overlapping area in the second diffuse reflection image of the N-1th overlapping pair.

[0150] It should be noted that for one overlap pair (3, 7), this overlap pair consists of the 3rd diffuse image and the 7th diffuse image. In this overlap pair, the 1st diffuse image is the 3rd diffuse image, and the 2nd diffuse image is the 7th diffuse image. For another overlap pair (7, 3), this overlap pair also consists of the 3rd diffuse image and the 7th diffuse image, but in this overlap pair, the 1st diffuse image is the 7th diffuse image, and the 2nd diffuse image is the 3rd diffuse image. Regarding element a in the above matrix A... n,r n = 0, 1, ..., N-1, r = 0, 1, ..., M-1, when the r-th diffuse image and the first diffuse image in the n-th overlapping pair are the same diffuse image, a n,r =1, when the r-th diffuse image and the second diffuse image in the n-th overlapping pair are the same diffuse image, a n,r =-1, otherwise, a n,r =0. For element b in the matrix B above. s,r s,r=0,1,...,M-1, when s=r, b s,r =1, otherwise, b s,r =0. The reference mean constant matrix and reference standard deviation constant matrix are obtained by using all diffuse reflectance images as reference images, that is, assuming that the SAR image corresponding to each diffuse reflectance image does not need to be corrected, and the mean correction and standard deviation correction are both 0.

[0151] It is worth mentioning that the minimum adjustment model based on the diffuse reflectance region is based on the information of the regions in the diffuse reflectance region that overlap with other diffuse reflectance regions. It can analyze the radiation differences of SAR images caused by different viewing angles, and does not require processing of texture information, thus preserving the texture information of SAR images.

[0152] Step 15: Solve the least squares adjustment model used to calculate the mean correction of the SAR images to obtain the mean correction for each SAR image.

[0153] Specifically, through the formula:

[0154] X u =(C T PC) -1 C T PL u

[0155] Calculate the mean correction matrix;

[0156] Where P represents the weight matrix:

[0157]

[0158] Where E represents the identity matrix, O represents the zero matrix, and W represents the weight matrix of the overlapping pairs:

[0159]

[0160] Where w0 represents the weight of the 0th overlapping pair, w n w represents the weight of the nth overlapping pair. N-1 The weight of the (N-1)th overlapping pair is:

[0161]

[0162] Among them, P n P represents the number of pixels in the overlapping region of the nth overlapping pair. j This represents the number of pixels in the overlapping region of the j-th overlapping pair, where n,j = 0, 1, ..., N-1, and N represents the total number of overlapping pairs.

[0163] For example, the above formula can be calculated using mathematical calculation computer software such as Mathematica to obtain the mean correction for each SAR image.

[0164] It is worth mentioning that by solving the mean least squares adjustment model, the mean correction of each SAR image can be obtained based on the number of pixels in the overlapping area of ​​the overlapping pair.

[0165] Step 16: Solve the least squares adjustment model used to solve the standard deviation correction of SAR images to obtain the standard deviation correction of each SAR image.

[0166] Specifically, through the formula:

[0167] X δ =(C T PC) -1 C T PL δ

[0168] Calculate the standard deviation correction matrix.

[0169] Where P represents the weight matrix:

[0170]

[0171]

[0172] Where E represents the identity matrix, O represents the zero matrix, and W represents the weight matrix of the overlapping pairs.

[0173] For example, the above formula can be calculated using mathematical calculation computer software such as Mathematica to obtain the standard deviation correction for each SAR image.

[0174] It is worth mentioning that by solving the standard deviation least squares adjustment model, the standard deviation correction of each SAR image can be obtained based on the number of pixels in the overlapping area of ​​the overlapping pair.

[0175] Step 17: For each SAR image, correct the radiometric differences of the SAR image based on the mean correction and standard deviation correction corresponding to the SAR image.

[0176] In some embodiments of this application, the steps of correcting the radiometric differences of SAR images based on the mean correction and standard deviation correction corresponding to each SAR image specifically include:

[0177] The first step is through the formula:

[0178]

[0179] Correct the mean of all SAR images;

[0180] in, This represents the mean value of the 0th SAR image after correction. This represents the mean value of the (M-1)th SAR image after correction. This represents the mean of the 0th SAR image. This represents the mean of the (M-1)th SAR image.

[0181] The second step is to use the formula:

[0182]

[0183] Correct the standard deviation of all SAR images;

[0184] in, This represents the standard deviation of the 0th SAR image after correction. This represents the standard deviation of the (M-1)th SAR image after correction. This represents the standard deviation of the 0th SAR image. This represents the standard deviation of the (M-1)th SAR image.

[0185] The third step is to use the formula:

[0186]

[0187] Correct the radiance value of the r-th SAR image at (x,y);

[0188] Where, f(x,y) r Let g(x,y) be the corrected radiance value of the r-th SAR image at (x,y), where x represents the horizontal axis, y represents the vertical axis, r = 0, 1, ..., M-1, and M represents the total number of SAR images. r Let represent the radiance value of the i-th SAR image at (x, y). Let represent the mean value of the r-th SAR image. This represents the standard deviation of the r-th SAR image after correction. This represents the standard deviation of the r-th SAR image. This represents the mean value of the r-th SAR image after correction.

[0189] For example, computer software such as Mathematica and Gamma can be used to calculate the above formula and correct the radiation intensity value of each SAR image.

[0190] It is worth mentioning that each SAR image is divided into multiple grid images, enabling analysis of large-scale SAR images. Based on the diffuse reflection region, a minimum adjustment model is constructed, which can analyze the radiometric differences of SAR images caused by different viewing angles, and preserve the texture information of SAR images. This ensures that the information in the radiometrically corrected SAR images is preserved intact, thereby improving the quality of radiometric difference correction for large-scale SAR images.

[0191] Furthermore, since dividing SAR images into gridded images is not limited by the size of SAR images, it is also applicable to small-scale SAR images.

[0192] The SAR image radiation difference correction device provided in this application is described below as an example.

[0193] like Figure 4 As shown, this application embodiment provides a SAR image radiometric difference correction device, the SAR image radiometric difference correction device 400 including:

[0194] The segmentation module 401 acquires multiple SAR images of the target area and divides each SAR image into multiple grid images;

[0195] The first acquisition module 402 acquires the root mean square height of each grid image.

[0196] The second acquisition module 403 acquires the diffuse reflection image in the SAR image based on the root mean square height of all grid images in the SAR image for each SAR image.

[0197] Module 404 is constructed based on all diffuse reflectance images to build a least squares adjustment model for solving the mean correction of SAR images and a least squares adjustment model for solving the standard deviation correction of SAR images; the mean correction is used to correct the mean of SAR images, and the standard deviation correction is used to correct the standard deviation of SAR images.

[0198] The first solution module 405 solves the least squares adjustment model used to solve the mean correction of SAR images, and obtains the mean correction of each SAR image.

[0199] The second solution module 406 solves the least squares adjustment model used to solve the standard deviation correction of SAR images, and obtains the standard deviation correction of each SAR image.

[0200] The correction module 407 corrects the radiometric differences of each SAR image based on the mean correction and standard deviation correction corresponding to the SAR image.

[0201] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0202] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0203] like Figure 5 As shown, an embodiment of this application provides a terminal device, wherein the terminal device D10 of this embodiment includes: at least one processor D100 ( Figure 5The diagram shows only one processor, a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, which, when executing the computer program D102, implements the steps in any of the above method embodiments.

[0204] Specifically, when the processor D100 executes the computer program D102, it acquires multiple SAR images of the target area and divides each SAR image into multiple grid images. Then, for each grid image, it acquires the root mean square height corresponding to the grid image. For each SAR image, based on the root mean square heights corresponding to all grid images in the SAR image, it acquires the diffuse reflection image in the SAR image. Then, based on all diffuse reflection images, it constructs a least squares adjustment model for solving the mean correction of the SAR image and a least squares adjustment model for solving the standard deviation correction of the SAR image. Then, it solves the least squares adjustment model for solving the mean correction of the SAR image to obtain the mean correction of each SAR image. At the same time, it solves the least squares adjustment model for solving the standard deviation correction of the SAR image to obtain the standard deviation correction of each SAR image. Finally, for each SAR image, it corrects the radiometric differences of the SAR image based on the mean correction and standard deviation correction corresponding to the SAR image. This method divides each SAR image into multiple grid images, enabling analysis of large-scale SAR images. Based on the diffuse reflection region, a minimum adjustment model is constructed, which can analyze the radiometric differences in SAR images caused by different viewing angles. It also preserves the texture information of SAR images, ensuring that the information in the radiometrically corrected SAR images is completely preserved, thus improving the quality of radiometric difference correction for large-scale SAR images.

[0205] Furthermore, since dividing SAR images into gridded images is not limited by the size of SAR images, it is also applicable to small-scale SAR images.

[0206] The processor D100 can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0207] In some embodiments, the memory D101 may be an internal storage unit of the terminal device D10, such as a hard disk or memory of the terminal device D10. In other embodiments, the memory D101 may be an external storage device of the terminal device D10, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the terminal device D10. Furthermore, the memory D101 may include both internal and external storage units of the terminal device D10. The memory D101 is used to store the operating system, applications, bootloader, data, and other programs, such as the program code of the computer program. The memory D101 can also be used to temporarily store data that has been output or will be output.

[0208] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0209] This application provides a computer program product that, when run on a terminal device, enables the terminal device to implement the steps described in the various method embodiments above.

[0210] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to a SAR image radiometric difference correction method apparatus / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0211] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0212] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0213] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principles described in this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for correcting radiometric differences in SAR images, characterized in that, include: Acquire multiple SAR images of the target area, and divide each SAR image into multiple grid images; For each of the grid images, obtain the root mean square height corresponding to the grid image; For each of the SAR images, the diffuse reflectance image in the SAR image is obtained based on the root mean square height corresponding to all grid images in the SAR image; Based on all diffuse reflectance images, a least squares adjustment model is constructed to solve for the mean correction of SAR images and a least squares adjustment model is constructed to solve for the standard deviation correction of SAR images; the mean correction is used to correct the mean of the SAR images and the standard deviation correction is used to correct the standard deviation of the SAR images. The least squares adjustment model used to solve for the mean correction of SAR images is solved to obtain the mean correction of each SAR image. The least squares adjustment model used to solve the standard deviation correction of SAR images is solved to obtain the standard deviation correction of each SAR image. For each SAR image, the radiometric differences of the SAR image are corrected based on the mean correction and standard deviation correction corresponding to the SAR image. The least squares adjustment model used to solve for the standard deviation correction of SAR images is as follows: in, , , Represents the parameter matrix, Represents the correction coefficient matrix. Represents the reference coefficient matrix. This represents the standard deviation correction matrix. Represents the standard deviation constant matrix. Represents the correction standard deviation constant matrix. Represents the reference standard deviation constant matrix: in, This represents the standard deviation correction for the 0th SAR image. Indicates the first Standard deviation correction for SAR images This represents the standard deviation of the overlapping region in the first diffuse image of the 0th overlapping pair. This represents the standard deviation of the overlapping region in the second diffuse image of the 0th overlapping pair. Indicates the first The standard deviation of the overlapping region in the first diffuse reflection image of the overlapping pair Indicates the first The standard deviation of the overlapping area in the second diffuse reflection image of the overlapping pair.

2. The SAR image radiometric difference correction method according to claim 1, characterized in that, Obtaining the root mean square height corresponding to the grid image includes: Through the formula: Calculate the root mean square height corresponding to the grid image. ; in, Indicates the first in the grid image Elevation of each pixel , This represents the total number of pixels in the grid image. This represents the average elevation of all pixels in the grid image.

3. The SAR image radiometric difference correction method according to claim 1, characterized in that, The step of obtaining the diffuse reflectance image in the SAR image based on the root mean square height corresponding to all grid images in the SAR image includes: For each of the grid images, the root mean square height corresponding to the grid image is added to the region where the grid image is located in the geographic coordinate system to obtain the surface roughness of the region; Determine whether the surface roughness is greater than or equal to a preset surface roughness value; If so, the grid image corresponding to the surface roughness is taken as the diffuse grid image; All diffuse reflectance grid images are integrated onto a single image to obtain the diffuse reflectance image from the SAR imagery.

4. The SAR image radiometric difference correction method according to claim 1, characterized in that, The construction of a least-squares adjustment model for calculating the mean correction of SAR images and a least-squares adjustment model for calculating the standard deviation correction of SAR images, based on all diffuse reflectance images, includes: For each of the diffuse reflection images, each other diffuse reflection image is traversed. If the regions corresponding to the diffuse reflection image and the other diffuse reflection image overlap, then the diffuse reflection image and the other diffuse reflection image are regarded as an overlapping pair corresponding to the diffuse reflection image. Based on the overlapping regions corresponding to all overlapping pairs, a least squares adjustment model is constructed to solve for the mean correction of SAR images and a least squares adjustment model is constructed to solve for the standard deviation correction of SAR images.

5. The SAR image radiometric difference correction method according to claim 4, characterized in that, The least squares adjustment model used to solve for the mean correction of SAR images is as follows: in, , , Represents the parameter matrix, Represents the mean correction matrix. Represents a matrix with constant mean. Represents the correction coefficient matrix. Represents the reference coefficient matrix. Represents the correction mean constant matrix. Represents the reference mean constant matrix: in, This represents the mean correction for the 0th SAR image. Indicates the first Mean correction for each SAR image, matrix The elements in the table are correction coefficients, with values ​​of 1, -1, or 0. , , and All are matrices Elements in the matrix The elements in the table are reference coefficients, taking values ​​of 0 or 1. , , and All are matrices The elements in This represents the mean value of the overlapping region in the first diffuse image of the 0th overlapping pair. This represents the mean value of the overlapping region in the second diffuse image of the 0th overlapping pair. Indicates the first The mean value of the overlapping region in the first diffuse reflection image of the overlapping pair. Indicates the first The mean value of the overlapping region in the second diffuse reflection image of the overlapping pair.

6. The SAR image radiometric difference correction method according to claim 1, characterized in that, Solving the least squares adjustment model used to calculate the mean correction for SAR images, to obtain the mean correction for each SAR image, includes: Through the formula: Calculate the mean correction matrix; in, Weight matrix: in, Represents the identity matrix. Represents a 0 matrix. The weight matrix representing the overlapping pairs: in, This represents the weight of the 0th overlapping pair. Indicates the first The weights of overlapping pairs, Indicates the first Weights of overlapping pairs: in, Indicates the first The number of pixels in the overlapping region of an overlapping pair Indicates the first The number of pixels in the overlapping region of an overlapping pair , This represents the total number of overlapping pairs; The process of solving the least squares adjustment model used to calculate the standard deviation correction of SAR images to obtain the standard deviation correction for each SAR image includes: Through the formula: Calculate the standard deviation correction matrix.

7. The SAR image radiometric difference correction method according to claim 6, characterized in that, The step of correcting the radiometric differences of the SAR image based on the mean correction and standard deviation correction corresponding to the SAR image includes: Through the formula: Correct the mean of all SAR images; in, This represents the mean value of the 0th SAR image after correction. Indicates the first Mean value of SAR image after correction This represents the mean value of the 0th SAR image. Indicates the first The mean of a SAR image; Through the formula: Correct the standard deviation of all SAR images; in, This represents the standard deviation of the 0th SAR image after correction. Indicates the first The standard deviation of the corrected SAR image This represents the standard deviation of the 0th SAR image. Indicates the first Standard deviation of a SAR image; Through the formula: For the first SAR images in The radiation intensity value at that location is corrected; in, The first one is shown SAR images in The corrected radiation intensity value, Represents the x-axis, Represents the ordinate, , This represents the total number of SAR images. Indicates the first SAR images in The radiation intensity value at that location, Indicates the first The mean of a SAR image, Indicates the first The standard deviation of the corrected SAR image Indicates the first The standard deviation of a SAR image Indicates the first The mean value of a SAR image after correction.

8. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the SAR image radiometric difference correction method as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the SAR image radiometric difference correction method as described in any one of claims 1 to 7.

Images