Geometric correction method and equipment for satellite remote sensing image and storage medium

Abstract

The invention provides a geometric correction method and device for a satellite remote sensing image and a storage medium, and relates to the technical field of image correction, and the method comprises the steps: obtaining remote sensing image data of at least two different spectral bands, synchronously recording satellite attitude parameters and orbit position parameters, and carrying out the verification processing; based on the satellite attitude parameters and the orbit position, geometric coordinate conversion is carried out on each remote sensing image data, and target remote sensing image data is generated; dividing each piece of target remote sensing image data into a plurality of continuous grid units, and constructing a local geometric transformation model at the vertex of each grid unit; and resampling the original pixels in each grid unit according to the local geometric transformation model, and outputting a plurality of pieces of target remote sensing image data subjected to geometric correction. According to the invention, the geometric correction precision of the high-resolution satellite remote sensing image can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image correction, and in particular, to a geometric correction method, device, and storage medium for satellite remote sensing images. Background Art

[0002] In the field of remote sensing images, geometric correction of remote sensing images mainly relies on ground control points. Specifically, the existing technology manually selects corresponding ground reference points, and establishes a geometric correction model based on the mapping relationship between ground coordinates and image pixel coordinates, so as to eliminate geometric distortions caused by factors such as sensor attitude and orbital offset.

[0003] However, the above method has significant limitations when applied to high-resolution remote sensing satellites. First, the hyperspectral characteristics of high-resolution remote sensing satellites have greatly increased their image data. The process of manually selecting ground control points is time-consuming and it is difficult to ensure the spatial consistency of all-band data. Second, the unique imaging parameters of high-resolution remote sensing satellites do not match the parameters of existing correction models, and direct application will reduce the accuracy. In addition, the accuracy of ground control points is easily limited by factors such as terrain occlusion and fuzzy ground object features, and the complex attitude changes of high-resolution remote sensing satellites further lead to a reduction in the correction accuracy of the existing method. Therefore, how to improve the geometric correction accuracy of high-resolution satellite remote sensing images has become an urgent problem for those skilled in the art in the current field. Summary of the Invention

[0004] The problem solved by the present invention is how to improve the geometric correction accuracy of high-resolution satellite remote sensing images.

[0005] To solve the above problems, the present invention provides a geometric correction method, device, and storage medium for satellite remote sensing images.

[0006] In a first aspect, the present invention provides a geometric correction method for satellite remote sensing images, including:

[0007] Obtain remote sensing image data of at least two different spectral bands, and synchronously record the satellite attitude parameters and orbital position parameters during the imaging period;

[0008] Perform verification processing on each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters respectively;

[0009] Based on the verified satellite attitude parameters and the orbital position, perform geometric coordinate transformation on each verified remote sensing image data respectively to generate target remote sensing image data that is initially corrected in the geographic coordinate system;

[0010] Divide each of the target remote sensing image data into a plurality of consecutive grid units according to a preset grid rule, and at the vertices of each grid unit, construct a local geometric transformation model based on the vertex coordinate differences before and after the initial correction of the geographic coordinate system;

[0011] Resample the original pixels of the remote sensing image data in each grid unit according to the local geometric transformation model, and output the geometrically corrected target remote sensing image data.

[0012] Optionally, the remote sensing image data of at least two different spectral bands includes panchromatic band remote sensing image data and at least one multispectral band remote sensing image data, and the multispectral band remote sensing image data includes at least one of infrared band remote sensing image data, green band remote sensing image data, and visible light band remote sensing image data.

[0013] Optionally, the satellite attitude parameters include quaternion attitude parameters, satellite attitude parameter timestamps, and satellite attitude change rate data, and the orbital position parameter data includes satellite azimuth data, satellite moving speed, satellite position timestamps, and satellite position change rate data.

[0014] Optionally, the respective verification processes for the remote sensing image data, the satellite attitude parameters, and the orbital position parameters include:

[0015] Perform resolution compliance verification on each of the remote sensing image data based on a preset scenario;

[0016] Perform spatial coverage integrity verification on each of the remote sensing image data that has passed the resolution compliance verification, where, in response to detecting that the missing area exceeds a preset integrity threshold, perform a data repair operation on the remote sensing image data;

[0017] Perform format normality verification and semantic label completeness verification on the satellite attitude parameters and the orbital position parameters respectively;

[0018] Perform rationality verification on the satellite attitude parameters and the orbital position parameters that have passed the format normality verification and the semantic label completeness verification.

[0019] Optionally, the rationality verification on the satellite attitude parameters and the orbital position parameters that have passed the format normality verification and the semantic label completeness verification includes:

[0020] Perform mathematical normalization verification on the quaternion attitude parameters that have passed the format normality verification and the semantic label completeness verification;

[0021] Perform time series continuity verification based on the adjacent interval distribution characteristics of the satellite attitude parameter timestamps obtained through the above-mentioned format normalization verification and semantic label completeness verification;

[0022] Perform dynamic permission domain verification on the satellite attitude change rate data obtained through the above-mentioned format normalization verification and semantic label completeness verification;

[0023] Perform spatio-temporal evolution compliance verification on the satellite azimuth data and the satellite moving speed obtained through the above-mentioned format normalization verification and semantic label completeness verification;

[0024] Perform time series continuity verification based on the adjacent interval distribution characteristics of the satellite position parameter timestamps obtained through the above-mentioned format normalization verification and semantic label completeness verification;

[0025] Perform dynamic permission domain verification on the satellite position change rate data obtained through the above-mentioned format normalization verification and semantic label completeness verification.

[0026] Optionally, based on the verified satellite attitude parameters and orbital positions, perform geometric coordinate transformation on each remote sensing image data respectively to generate target remote sensing image data with preliminary correction in the geographic coordinate system, including:

[0027] According to the camera factory angle parameters, convert each verified remote sensing image data from the focal plane coordinate system to the camera coordinate system to obtain the corresponding remote sensing image data in the camera coordinate system;

[0028] Construct a rotation matrix based on the quaternion attitude parameters of the satellite attitude parameters to convert each remote sensing image data in the camera coordinate system to the satellite body coordinate system to obtain the remote sensing image data in the satellite body coordinate system;

[0029] Based on the satellite geocentric coordinates and the earth's angular velocity of rotation in the orbital position parameters, convert the remote sensing image data in the satellite body coordinate system to the geocentric inertial coordinate system to obtain the remote sensing image data in the geocentric inertial coordinate system;

[0030] Based on the preset precession model matrix, nutation matrix and polar motion matrix, convert the remote sensing image data in the geocentric inertial coordinate system to the geographic coordinate system to generate the target remote sensing image data with preliminary correction in the geographic coordinate system.

[0031] Optionally, divide each target remote sensing image data into multiple continuous grid cells according to the preset grid rules, and at the vertices of each grid cell, construct a local geometric transformation model according to the vertex coordinate differences before and after the preliminary correction in the geographic coordinate system, including:

[0032] Generate a dynamic grid division rule according to the geographical coverage area characteristic parameters of each piece of the target remote sensing image data, and divide the target remote sensing image data into a plurality of polygon grid units that are seamlessly spliced with each other based on the dynamic grid division rule;

[0033] Extract the geographical coordinates of the vertices of each grid unit before and after the initial correction of the geographical coordinate system, and construct a two-dimensional displacement vector group representing the differences in the coordinate positions of the vertices of each grid unit before and after;

[0034] Establish a geometric transformation model for each grid unit based on the distribution characteristics of the two-dimensional displacement vector group.

[0035] Optionally, the resampling the original pixels of the remote sensing image data within each grid unit according to the local geometric transformation model and outputting the geometrically corrected target remote sensing image data includes:

[0036] Establish a continuous mapping relationship between the original pixels and the target geographical coordinates based on the geometric transformation model of each grid unit;

[0037] According to the dynamic change parameters of the mapping relationship, resample the original pixels by using an interpolation algorithm, and eliminate the geometric deformation residuals between adjacent grid units to obtain secondary pixel data;

[0038] Bind the geographical coordinates to the secondary pixel data, and output the geometrically corrected target remote sensing image data.

[0039] In a second aspect, the present invention provides an electronic device, including a memory and a processor;

[0040] The memory is used to store a computer program;

[0041] The processor is used to implement the geometric correction method of the satellite remote sensing image as described in any one of the above when executing the computer program.

[0042] In a third aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the geometric correction method of the satellite remote sensing image as described in any one of the above is implemented.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows: By synchronously collecting different remote sensing image data (such as visible light band, short-wave infrared band, and panchromatic band) and performing geometric correction separately, it is possible to avoid the decrease in correction accuracy caused by multi-band imaging differences during unified correction. Secondly, through the separate verification of remote sensing image data, satellite attitude parameters, and orbital position parameters, a reliable input data set can be established for geometric correction. Furthermore, using the satellite's own satellite attitude parameters and orbital position parameters to perform geometric transformation on remote sensing image data avoids the dependence on ground control points in geometric correction in related technologies, thereby simplifying the correction process of remote sensing image data and avoiding the problem of reduced correction accuracy caused by limited accuracy of ground control points, improving the correction accuracy. Moreover, using the grid cell division of the target remote sensing image data can further quantify the matching errors caused by terrain undulation or sensor dynamic jitter. Finally, resampling the original pixels of the remote sensing image data in each grid cell according to the local geometric transformation model ensures the sub-pixel alignment of remote sensing image data in different spectral bands, thereby completing the geometric correction of the target remote sensing image data. The present invention can effectively improve the geometric correction accuracy of high-resolution satellite remote sensing images. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic flowchart of a geometric correction method for a satellite remote sensing image provided by an embodiment of the present invention;

[0045] Figure 2 It is a schematic flowchart of step S13 in a geometric correction method for a satellite remote sensing image provided by an embodiment of the present invention;

[0046] Figure 3 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention is provided in conjunction with the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments described herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not used to limit the protection scope of the present invention.

[0048] It should be understood that the various steps recorded in the method embodiments of the present invention can be executed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this regard.

[0049] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to"; the term "based on" means "based at least in part on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optionally" means "optional embodiments". The relevant definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc. mentioned in the present invention are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0050] It should be noted that the modifications of "one" and "plurality" mentioned in the present invention are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0051] In the field of remote sensing image geometric correction technology, the ground control point (GCP)-dependent method is still the core solution of traditional correction. This solution relies on the spatial correspondence between manually annotated ground reference points and image pixels, and compensates for geometric deformations induced by factors such as satellite platform posture offset and orbital perturbation by constructing a coordinate mapping model. However, with the iterative upgrade of high-spectral resolution remote sensing payloads, this technical system encounters three adaptability bottlenecks: first, with the surge in the dimension of hyperspectral satellite imaging data, manual point-by-point calibration not only leads to an increase in working hours, but also causes multi-spectral channel spatial benchmark misalignment due to visual interpretation bias; secondly, the high dynamic imaging characteristics of satellite-borne multi-spectral sensors (such as inter-band visual axis discreteness and integration time asynchrony) are systematically offset from the preset parameters, making it difficult for traditional single correction models to adapt to the correction needs of wide spectral bands; finally, complex terrain occlusion causes uneven ground control space density, which further reduces the credibility of control points, thereby leading to a decrease in correction accuracy.

[0052] Based on this, the present invention provides a method, device and storage medium for geometric correction of satellite remote sensing images.

[0053] Reference Figure 1 The present invention provides a method for geometric correction of satellite remote sensing images, comprising:

[0054] S11. Acquire remote sensing image data of at least two different spectral bands, and synchronously record satellite attitude parameters and orbital position parameters during the imaging period.

[0055] Specifically, visible light, short-wave infrared, and panchromatic band remote sensing image data can be synchronously acquired through a spaceborne multispectral sensor, and a sensitive gyro component is used to record the satellite attitude data and orbital position parameters at the moment of imaging in real time. Through the linkage of multi-band remote sensing image data, satellite attitude parameters, and orbital position parameters, it is possible to separately process the differences in the size, focal length, imaging method, etc. of the band images in each band of remote sensing image data, thereby improving the accuracy of geometric correction.

[0056] S12. Check and process each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters respectively.

[0057] Furthermore, in this embodiment, a method of joint calibration of multi-source data can be adopted to check and process each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters respectively, so as to improve the accuracy of multi-source data.

[0058] S13. Based on the verified satellite attitude parameters and the orbital position, perform geometric coordinate transformation on each of the verified remote sensing image data respectively to generate target remote sensing image data with preliminary correction in the geographic coordinate system.

[0059] In this embodiment, geometric coordinate transformation can be performed on remote sensing image data through the satellite's own satellite attitude parameters and orbital position, which can avoid the dependence on ground control points in geometric correction, thereby simplifying the geometric correction of satellite remote sensing images. Moreover, by performing independent geometric coordinate transformation on each remote sensing image data, geometric coordinate transformation can be performed according to the characteristics of remote sensing image data in different bands, thereby avoiding the cross-band mapping distortion caused by the overall geometric coordinate transformation of feature data in different bands in the traditional method, and further improving the accuracy of geometric correction.

[0060] S14. Divide each of the target remote sensing image data into multiple continuous grid units according to a preset grid rule, and at the vertices of each grid unit, construct a local geometric transformation model based on the vertex coordinate differences before and after the preliminary correction in the geographic coordinate system.

[0061] The preset grid rule can be to divide the target remote sensing image data into grid units of a preset fixed size, or a dynamic grid division rule constructed according to the geographical characteristics of the target area, so as to improve the transformation accuracy of the local geometric transformation model of the target remote sensing image data.

[0062] S15. Resample the original pixels of the remote sensing image data in each grid unit according to the local geometric transformation model, and output the target remote sensing image data after geometric correction.

[0063] Compared with the prior art, the beneficial effects of the present invention are as follows: By synchronously collecting different remote sensing image data (such as visible light band, short-wave infrared band, and panchromatic band), and performing geometric correction respectively, it is possible to avoid the decrease in correction accuracy caused by multi-band imaging differences when performing unified correction; Secondly, by separately verifying the remote sensing image data, satellite attitude parameters, and orbital position parameters, a reliable input data set can be established for geometric correction; Furthermore, using the satellite's own satellite attitude parameters and orbital position parameters to perform geometric transformation on the remote sensing image data avoids the dependence on ground control points in geometric correction in related technologies, thus simplifying the correction process of remote sensing image data and avoiding the problem of reduced correction accuracy caused by limited accuracy of ground control points, and improving the correction accuracy; Moreover, using the grid cell division of the target remote sensing image data can further quantify the matching errors caused by terrain undulation or sensor dynamic jitter; Finally, resampling the original pixels of the remote sensing image data in each of the grid cells according to the local geometric transformation model ensures the sub-pixel alignment of the remote sensing image data in different spectral bands, thereby completing the geometric correction of the target remote sensing image data. The present invention can effectively improve the geometric correction accuracy of high-resolution satellite remote sensing images.

[0064] In one embodiment, the remote sensing image data of at least two different spectral bands includes panchromatic band remote sensing image data and at least one multi-spectral band remote sensing image data, and the multi-spectral band remote sensing image data includes at least one of infrared band remote sensing image data, green light band remote sensing image data, and visible light band remote sensing image data.

[0065] In this embodiment, by synchronously acquiring panchromatic band remote sensing image data with high spatial resolution and multi-spectral band remote sensing image data including infrared, green light, or visible light, a geometric correction input source for the composite band is constructed. Based on the high-frequency texture features of the panchromatic band remote sensing image data, the fine structure of ground objects can be captured, while the enhanced separability of the wide-spectrum response features of the multi-spectral band remote sensing image data enables dual driving in the spatial and spectral dimensions during geometric correction; Furthermore, the panchromatic band can be used as the main correction benchmark to improve the edge alignment accuracy (such as eliminating the influence of the blurring effect caused by atmospheric scattering on multi-spectral registration), the thermal radiation characteristics of the infrared band remote sensing image data can assist in correcting the projection difference caused by terrain undulation (such as identifying elevation anomalies in shadow areas through the difference in thermal inertia), and the green light band remote sensing image data can strengthen the geometric constraints of landmarks such as water bodies and vegetation through high albedo features, ultimately forming a correction framework with multi-band collaborative enhancement, reducing the cross-band registration error and the geometric restoration accuracy of the precision shadow area in images with high vegetation coverage.

[0066] In one embodiment, the satellite attitude parameters include quaternion attitude parameters, satellite attitude parameter timestamps, and satellite attitude change rate data, and the orbital position parameter data includes satellite azimuth data, satellite moving speed, satellite position timestamps, and satellite position change rate data.

[0067] In this embodiment, by integrating quaternion attitude parameters, satellite attitude parameter timestamps, and satellite attitude change rate data, a high-precision satellite three-axis pointing dynamic model is constructed. At the same time, by integrating satellite azimuth data, satellite moving speed, satellite position timestamps, and satellite position change rate data, six-degree-of-freedom pose four-dimensional reconstruction of each frame of image in satellite remote sensing image data can be realized, thereby improving the geometric correction accuracy of satellite remote sensing images.

[0068] In one embodiment, the verification processing of each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters includes:

[0069] Performing resolution compliance verification on each of the remote sensing image data based on a preset scenario;

[0070] Performing spatial coverage integrity verification on each of the remote sensing image data that passes the resolution compliance verification, wherein, when it is detected that the missing area exceeds a preset integrity threshold, a data repair operation is performed on the remote sensing image data;

[0071] Performing format normality verification and semantic label completeness verification on the satellite attitude parameters and the orbital position parameters respectively;

[0072] Performing rationality verification on the satellite attitude parameters and the orbital position parameters that pass the format normality verification and the semantic label completeness verification.

[0073] Specifically, in this embodiment, resolution compliance verification of each of the remote sensing image data (such as only retaining remote sensing image data that meets the preset accuracy requirements) is performed through typical ground object scenarios (such as airport runways, standard targets), and combined with spatial integrity verification (such as detecting the spatial coverage integrity of images through grid scanning). When the missing area exceeds a preset ratio, a sub-pixel-level data repair algorithm is started, which can significantly reduce the time-consuming for complementing remote sensing image data; at the same time, the satellite attitude parameters and orbital position parameters are verified respectively, which not only verifies whether the data format conforms to the international standard of spatial metadata, but also checks the integrity of semantic elements such as time tags and physical dimensions. This embodiment constructs a multi-level quality inspection barrier for multi-source data, which can effectively improve the geometric distortion correction efficiency of high-resolution satellite images.

[0074] In one embodiment, the rationality verification of the satellite attitude parameters and the orbital position parameters after passing the format normality verification and the semantic label completeness verification includes:

[0075] Perform a mathematical normalization check on the quaternion attitude parameters that have passed the format standardization check and the semantic tag completeness check;

[0076] Perform a time series continuity check based on the adjacent interval distribution characteristics of the satellite attitude parameter timestamps that have passed the format standardization check and the semantic tag completeness check;

[0077] Perform a dynamic permission domain check on the satellite attitude change rate data that has passed the format standardization check and the semantic tag completeness check;

[0078] Perform a spatio-temporal evolution compliance check on the satellite azimuth data and the satellite moving speed that have passed the format standardization check and the semantic tag completeness check;

[0079] Perform a time series continuity check based on the adjacent interval distribution characteristics of the satellite position parameter timestamps that have passed the format standardization check and the semantic tag completeness check;

[0080] Perform a dynamic permission domain check on the satellite position change rate data that has passed the format standardization check and the semantic tag completeness check.

[0081] Specifically, in this embodiment, first, a normalization calibration is performed on the quaternion attitude parameters, which can significantly improve the description accuracy of the satellite attitude parameters; secondly, through the analysis of the timestamp interval distribution of the satellite attitude and position parameters, the coherence of the time series data is ensured; and the dynamic permission domain check of the satellite attitude change rate data and the satellite position change rate data can screen out the abnormal values of the satellite attitude change rate data and the satellite position change rate data by combining with the dynamic characteristics of the spacecraft; the spatio-temporal evolution compliance check of the satellite azimuth data and the satellite moving speed can screen out the obviously abnormal data. This embodiment can systematically perform a rationality check on the satellite attitude parameters and orbital position parameters, and significantly improve the reliability of high-precision remote sensing satellite image data.

[0082] In one embodiment, based on the tested satellite attitude parameters and the orbital position, perform a geometric coordinate transformation on each remote sensing image data respectively to generate target remote sensing image data with an initial correction in the geographic coordinate system, including:

[0083] S131. According to the camera factory angle parameters, convert each tested remote sensing image data from the focal plane coordinate system to the camera coordinate system to obtain the corresponding remote sensing image data in the camera coordinate system.

[0084] Specifically, the camera factory angle parameters include the principal point offset coordinates, focal length, and focal plane mounting rotation matrix; among them, the principal point offset coordinates refer to the row and column coordinates of the center point of the camera imaging sensor on the focal plane; the conversion formula from the focal plane coordinate system to the camera coordinate system is:

[0085]

[0086] Among them, (i, j) represents the pixel coordinates of the remote sensing image data in the focal plane coordinate system, (c x , c y ) represents the principal point offset coordinates, (X cam , Y cam , Z cam ) represents the three-dimensional coordinates of the pixel coordinates in the camera coordinate system, f represents the focal length, R cam represents the focal plane mounting rotation matrix, and the focal plane mounting rotation matrix is used to describe the relative attitude between the focal plane coordinate system and the camera coordinate system, and a represents the pixel physical size.

[0087] S132. Construct a rotation matrix based on the quaternion attitude parameters of the satellite attitude parameters to convert each piece of remote sensing image data in the camera coordinate system to the satellite body coordinate system, and obtain the remote sensing image data in the satellite body coordinate system.

[0088] Specifically, the formula for constructing a rotation matrix based on the quaternion attitude parameters of the satellite attitude parameters is:

[0089]

[0090] Among them, (q1, q2, q3, q4) represents the quaternion attitude parameters of the satellite attitude parameters, and the quaternion attitude parameters are used to represent the attitude of the satellite body relative to the inertial system, and R att represents the rotation matrix;

[0091] The formula for converting each piece of remote sensing image data in the camera coordinate system to the satellite body coordinate system to obtain the remote sensing image data in the satellite body coordinate system is:

[0092]

[0093] represents the inverse matrix of the rotation matrix, P cam represents the coordinates in the camera coordinate system, and P sat represents the coordinates in the satellite body coordinate system.

[0094] S133. Based on the satellite geocentric coordinates and the earth's angular velocity in the orbital position parameters, convert the remote sensing image data in the satellite body coordinate system to the geocentric inertial coordinate system to obtain the remote sensing image data in the geocentric inertial coordinate system.

[0095] Specifically, convert the remote sensing image data in the satellite body coordinate system to the geocentric inertial coordinate system to obtain the remote sensing image data in the geocentric inertial coordinate system, where the conversion formula is:

[0096]

[0097] where P ECI represents the coordinates of the remote sensing image data in the geocentric inertial coordinate system, T sat represents the coordinates of the satellite body in the geocentric inertial coordinate system, P sat represents the coordinates in the satellite body coordinate system, and w e represents the angular velocity of the Earth's rotation.

[0098] S134. Based on the preset precession model matrix, nutation matrix, and polar motion matrix, convert the remote sensing image data in the geocentric inertial coordinate system to the geographic coordinate system to generate the target remote sensing image data with initial correction in the geographic coordinate system.

[0099] Specifically, the precession model matrix, nutation matrix, and polar motion matrix can be constructed according to the 2010 specifications of the International Earth Rotation and Reference Systems Service, and then the remote sensing image data in the geocentric inertial coordinate system is converted to the geographic coordinate system based on the constructed precession model matrix, nutation matrix, and polar motion matrix. The conversion process includes:

[0100] Convert the remote sensing image data in the geocentric inertial coordinate system to the geocentric rectangular coordinate system. The conversion formula is:

[0101] P ITRF =R pm ·R nut ·R pre ·R GAST ·P ECI ;

[0102] where P ITRF represents the coordinates of the remote sensing image data in the geocentric rectangular coordinate system, R pm represents the polar motion matrix, R nut represents the nutation matrix, R pre represents the precession model matrix, R GAST represents the rotation matrix around the Z-axis corresponding to the Greenwich Apparent Sidereal Time (GAST), and P ECI represents the coordinates of the remote sensing image data in the satellite body coordinate system;

[0103] Then convert the coordinates of the remote sensing image data in the geocentric rectangular coordinate system to the geographic coordinate system. The conversion formula is:

[0104]

[0105] Wherein, λ represents the longitude of the target remote sensing image data in the geographic coordinate system, φ represents the latitude of the remote sensing image data in the geographic coordinate system, (X ITRF , Y ITRF , Z ITRF ) represents the three-dimensional coordinates of the remote sensing image data in the geocentric rectangular coordinate system, a represents the semi-major axis of the earth, e 2 represents the square of the first eccentricity, and h represents the elevation.

[0106] In this embodiment, first, according to the camera factory parameters, the remote sensing image data is converted from the focal plane to the camera coordinate system to eliminate the internal distortion of the sensor; then, based on the quaternion attitude parameters, a rotation matrix is constructed to map the remote sensing image data from the camera coordinate system to the satellite body coordinate system to ensure the mathematical rigor of the attitude representation; further, in combination with the orbital position parameters and the earth rotation model, the remote sensing image data is converted to the geocentric inertial coordinate system to complete the coupling correction of the satellite motion trajectory and the earth rotation; finally, through the correction of geophysical models such as precession, nutation and polar motion, the initially calibrated image in the geographic coordinate system is generated. Through the step-by-step progressive coordinate system conversion, this embodiment significantly improves the geometric accuracy and spatio-temporal consistency of the geographic coordinate system positioning of the remote sensing image data on the basis of strictly relying on the satellite attitude, orbital parameters and the earth motion model.

[0107] In one embodiment, the step of dividing each of the target remote sensing image data into a plurality of continuous grid units according to a preset grid rule and constructing a local geometric transformation model at the vertices of each of the grid units according to the vertex coordinate differences before and after the initial correction in the geographic coordinate system includes:

[0108] Generating a dynamic grid division rule according to the geographic coverage area characteristic parameters of each of the target remote sensing image data, and dividing the target remote sensing image data into a plurality of polygon grid units that are seamlessly spliced with each other.

[0109] Specifically, for the target remote sensing image data (including the pixel matrix and the geographic coordinate range) after the initial correction, the geographic coverage area characteristic parameters include the terrain complexity index, the spatial resolution, and the preset range of the preset grid side length; the dynamic grid division rule can be a rule for grid division according to the terrain complexity. Specifically, based on the quadtree segmentation algorithm on the basis of the dynamic grid division rule, the target remote sensing image data is divided into a plurality of polygon grid units that are seamlessly spliced with each other, and at the same time, the vertex coordinate indexes of each polygon grid unit are generated. It should be noted that the quadtree segmentation algorithm is a prior art and will not be described in detail here.

[0110] Extracting the geographic coordinates of the vertices of each grid unit before and after the initial correction in the geographic coordinate system, and constructing a two-dimensional displacement vector group representing the differences in the vertex coordinate positions of each grid unit before and after.

[0111] It should be noted that the two-dimensional displacement vector group is constructed based on the differences in the coordinate positions of the vertices of the aforementioned grid cells before and after. For example, for a certain grid vertex, the coordinate before initial correction is (x1, y1, z1), and the coordinate after initial correction is (x2, y2, z2), then the corresponding two-dimensional displacement vector is: (x2 - x1, y2 - y1, z2 - z1). Furthermore, the two-dimensional displacement vectors of the vertices of each grid cell are aggregated to form the two-dimensional displacement vector group D.

[0112] Based on the distribution characteristics of the two-dimensional displacement vector group, a geometric transformation model is established for each grid cell.

[0113] Specifically, this step includes: performing covariance analysis on the two-dimensional displacement vector group D. If the eigenvalue of the covariance matrix in the two-dimensional displacement vector group D is lower than the preset threshold, then a geometric transformation model is constructed for each grid cell based on the reflection transformation model. If the eigenvalue of the covariance matrix in the two-dimensional displacement vector group D is not lower than the preset threshold, then a rigid body transformation model is selected to construct a geometric transformation model for each grid cell.

[0114] In this embodiment, first, seamless mosaicked polygon grid cells are adaptively generated according to the geographical coverage characteristics of the target remote sensing image data, ensuring that the division result fits the surface space structure. Then, by comparing the geographical coordinate differences of the grid cell vertices before and after initial correction, the two-dimensional displacement vector group is extracted to quantify the local deformation characteristics. Furthermore, based on the displacement distribution characteristics, an independent geometric transformation model is established for each grid to specifically correct the deformation errors in each region.

[0115] Through refined local deformation characterization and dynamic block correction, this embodiment can effectively eliminate local geometric distortions while maintaining the overall consistency of the target remote sensing image data, significantly improving the spatial geometric accuracy and surface feature matching ability of high-resolution remote sensing images.

[0116] In one embodiment, the resampling of the original pixels of the remote sensing image data in each grid cell according to the local geometric transformation model and outputting the geometrically corrected target remote sensing image data includes:

[0117] Establishing a continuous mapping relationship between the original pixels and the target geographical coordinates based on the geometric transformation model of each grid cell.

[0118] According to the dynamic change parameters of the mapping relationship, using an interpolation algorithm to resample the original pixels and eliminating the geometric deformation residuals between adjacent grid cells to obtain secondary pixel data.

[0119] Specifically, the dynamic change parameter here can be either the degree of geometric change or the value of the Jacobian determinant to characterize the deformation strength of the grid cell. Subsequently, the dynamic change parameter is judged (the judgment rule can be flexibly set according to the actual situation and will not be specifically limited here). If it belongs to the strong deformation area, the large window interpolation algorithm can be used to interpolate the original data to generate the resampling result of the original pixels. If it belongs to the weak deformation area, the bilinear interpolation method can be used to interpolate the original pixels to generate the resampling result of the original pixels. Then, based on the distance from the resampled original pixels to the grid cell boundary, the mixing weight of the original pixels is calculated (the design rule of the mixing weight can be: the weight at the boundary of the original pixels is the standard mixing weight, and the mixing weight of the original pixels gradually decreases towards the center of the grid cell), and the pixel values of adjacent grid cells are weighted and averaged to complete the geometric deformation elimination operation between adjacent grid cells, thereby obtaining the secondary pixel data.

[0120] Bind the geographical coordinates to the secondary pixel data and output the geometrically corrected target remote sensing image data.

[0121] Binding the geographical coordinates to the secondary pixel data means determining the coordinates of the secondary pixel data in the geographical coordinate system. The specific steps may include calculating the transformation relationship matrix between the secondary pixel coordinates and the geographical coordinate system according to the satellite geometric parameters (such as the sensor attitude) in the mapping relationship, so as to transform the matrix position of each secondary pixel (such as the row and column positions in the grid cell) into longitude and latitude coordinates through the transformation relationship matrix, thereby obtaining the target remote sensing image data.

[0122] In this embodiment, first, a continuous mapping relationship from pixels to geographical coordinates is established based on the grid cell geometric transformation model to ensure the mathematical rigor of local deformation correction. Then, the interpolation algorithm is used to resample the original pixels in combination with the dynamic distribution characteristics of the mapping parameters, which can effectively eliminate the geometric residuals between adjacent grids and generate the transition pixel data with secondary accuracy. Finally, by binding the secondary pixel data with the geographical coordinates, the geometrically aligned multi-band target remote sensing image data is output.

[0123] Furthermore, in this embodiment, the deformation gradient is kept smooth through continuous mapping, and the problems of seams and edge mutations in the target remote sensing image data are reduced by means of residual elimination and pixel fusion technologies, significantly improving the spatial consistency, detail fidelity, and cross-region stitching quality of multi-source remote sensing data, and effectively improving the geometric correction accuracy of satellite remote sensing images.

[0124] Referring to Figure 3 , the present invention provides an electronic device 30, including a memory 31 and a processor 32;

[0125] The memory 31 is used to store computer programs;

[0126] The processor 32 is configured to implement the above geometric correction method for satellite remote sensing images when executing the computer program.

[0127] Alternatively, an electronic device 30 includes a memory 31 and a processor 32 coupled to the memory 31; the memory 31 is configured to store a computer program; the processor 32 is configured to perform the following operations when executing the computer program:

[0128] Obtain remote sensing image data of at least two different spectral bands, and synchronously record the satellite attitude parameters and orbital position parameters during the imaging period;

[0129] Perform verification processing on each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters respectively;

[0130] Based on the verified satellite attitude parameters and the orbital position, perform geometric coordinate transformation on each verified remote sensing image data respectively to generate target remote sensing image data with initial correction in the geographic coordinate system;

[0131] Divide each of the target remote sensing image data into a plurality of consecutive grid units according to a preset grid rule, and at the vertices of each grid unit, construct a local geometric transformation model according to the vertex coordinate differences before and after the initial correction in the geographic coordinate system;

[0132] Resample the original pixels of the remote sensing image data in each grid unit according to the local geometric transformation model, and output the target remote sensing image data after geometric correction.

[0133] The present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above geometric correction method for satellite remote sensing images is implemented.

[0134] Alternatively, a non-volatile computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor is made to perform the following operations:

[0135] Obtain remote sensing image data of at least two different spectral bands, and synchronously record the satellite attitude parameters and orbital position parameters during the imaging period;

[0136] Perform verification processing on each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters respectively;

[0137] Based on the verified satellite attitude parameters and the orbital positions, perform geometric coordinate transformation on each verified remote sensing image data respectively to generate target remote sensing image data with initial correction in the geographic coordinate system;

[0138] Divide each of the target remote sensing image data into a plurality of consecutive grid units according to a preset grid rule, and at the vertices of each grid unit, construct a local geometric transformation model based on the vertex coordinate differences before and after the initial correction in the geographic coordinate system;

[0139] Resample the original pixels of the remote sensing image data in each grid unit according to the local geometric transformation model, and output the target remote sensing image data after geometric correction.

[0140] Now, an electronic device 30 that can be a server or a client of the present invention will be described. It is an example of a hardware device that can be applied to various aspects of the present invention. The electronic device 30 is intended to represent various forms of digital electronic computer devices, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 30 can also represent various forms of mobile devices, such as, personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0141] The electronic device 30 includes a computing unit, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) or a computer program loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The computing unit, the ROM, and the RAM are connected to each other through a bus. An input / output (I / O) interface is also connected to the bus.

[0142] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc. In this application, the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present invention. In addition, the functional units in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0143] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will all fall within the protection scope of the present invention.

Claims

1. A geometric correction method for satellite remote sensing images, characterized in that, including: obtaining remote sensing image data of at least two different spectral bands, and synchronously recording satellite attitude parameters and orbital position parameters at the imaging time; performing verification processing on each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters respectively; based on the verified satellite attitude parameters and the orbital position, respectively performing geometric coordinate transformation on each verified remote sensing image data to generate target remote sensing image data with preliminary correction in the geographic coordinate system; dividing each of the target remote sensing image data into a plurality of consecutive grid units according to a preset grid rule, and at the vertex of each grid unit, constructing a local geometric transformation model according to the vertex coordinate difference before and after the preliminary correction in the geographic coordinate system; performing resampling on the original pixels of the remote sensing image data in each grid unit according to the local geometric transformation model, and outputting the target remote sensing image data after geometric correction.

2. The geometric correction method of satellite remote sensing images according to claim 1, characterized in that, The remote sensing image data of the at least two different spectral bands includes panchromatic band remote sensing image data and at least one multispectral band remote sensing image data, and the multispectral band remote sensing image data includes at least one of infrared band remote sensing image data, green light band remote sensing image data, and visible light band remote sensing image data.

3. The geometric correction method of satellite remote sensing images according to claim 1, characterized in that The satellite attitude parameters include quaternion attitude parameters, satellite attitude parameter timestamps, and satellite attitude change rate data, and the orbital position parameter data includes satellite azimuth data, satellite moving speed, satellite position timestamps, and satellite position change rate data.

4. The geometric correction method of satellite remote sensing images according to claim 3, wherein, The performing verification processing on each of the remote sensing image data, the satellite attitude parameters, and the orbital position parameters respectively includes: performing resolution compliance verification on each of the remote sensing image data based on a preset scenario; performing spatial coverage integrity verification on each of the remote sensing image data that passes the resolution compliance verification, wherein in response to detecting that the missing area exceeds a preset integrity threshold, performing a data repair operation on the remote sensing image data; respectively performing format standardization verification and semantic label completeness verification on the satellite attitude parameters and the orbital position parameters; performing rationality verification on the satellite attitude parameters and the orbital position parameters that pass the format standardization verification and the semantic label completeness verification.

5. The geometric correction method for satellite remote sensing images according to claim 4, characterized in that, The performing rationality verification on the satellite attitude parameters and the orbital position parameters that pass the format standardization verification and the semantic label completeness verification includes: performing mathematical normalization verification on the quaternion attitude parameters that pass the format standardization verification and the semantic label completeness verification; performing time series continuity verification based on the adjacent interval distribution characteristics of the satellite attitude parameter timestamps that pass the format standardization verification and the semantic label completeness verification; performing dynamic permission domain verification on the satellite attitude change rate data that pass the format standardization verification and the semantic label completeness verification; performing spatio-temporal evolution compliance verification on the satellite azimuth data and the satellite moving speed that pass the format standardization verification and the semantic label completeness verification; Perform time series continuity verification based on the adjacent interval distribution characteristics of the satellite position parameter timestamps through the format normality verification and the semantic tag completeness verification; Perform dynamic permission domain verification on the satellite position change rate data that has passed the format normality verification and the semantic tag completeness verification.

6. The geometric correction method for satellite remote sensing images according to claim 1, characterized in that Based on the verified satellite attitude parameters and the orbital positions, perform geometric coordinate transformation on each remote sensing image data respectively to generate target remote sensing image data with initial correction in the geographic coordinate system, including: According to the camera factory angle parameters, convert each verified remote sensing image data from the focal plane coordinate system to the camera coordinate system to obtain corresponding remote sensing image data in the camera coordinate system; Construct a rotation matrix based on the quaternion attitude parameters of the satellite attitude parameters to convert each remote sensing image data in the camera coordinate system to the satellite body coordinate system to obtain remote sensing image data in the satellite body coordinate system; Based on the satellite geocentric coordinates and the earth's angular velocity of rotation in the orbital position parameters, convert the remote sensing image data in the satellite body coordinate system to the geocentric inertial coordinate system to obtain remote sensing image data in the geocentric inertial coordinate system; Based on the preset precession model matrix, nutation matrix and polar motion matrix, convert the remote sensing image data in the geocentric inertial coordinate system to the geographic coordinate system to generate target remote sensing image data with initial correction in the geographic coordinate system.

7. The geometric correction method of satellite remote sensing images according to claim 1, characterized in that, The process of dividing each target remote sensing image data into multiple consecutive grid cells according to a preset grid rule, and at the vertices of each grid cell, constructing a local geometric transformation model based on the vertex coordinate differences before and after the initial correction in the geographic coordinate system, includes: Generate a dynamic grid division rule according to the geographic coverage area characteristic parameters of each target remote sensing image data, and divide the target remote sensing image data into multiple polygon grid cells that are seamlessly spliced with each other based on the dynamic grid division rule; Extract the geographic coordinates of the vertices of each grid cell before and after the initial correction in the geographic coordinate system, and construct a two-dimensional displacement vector group representing the differences in the vertex coordinate positions of each grid cell before and after; Based on the distribution characteristics of the two-dimensional displacement vector group, establish a geometric transformation model for each grid cell.

8. The geometric correction method of satellite remote sensing images according to claim 7, characterized in that, The process of resampling the original pixels of the remote sensing image data in each grid cell according to the local geometric transformation model and outputting the geometrically corrected target remote sensing image data includes: Establish a mapping relationship between the original pixels and the target geographic coordinates based on the geometric transformation model of each grid cell; According to the dynamic change parameters of the mapping relationship, use an interpolation algorithm to resample the original pixels and eliminate the geometric deformation residuals between adjacent grid cells to obtain secondary pixel data; Bind the geographic coordinates to the secondary pixel data and output the geometrically corrected target remote sensing image data.

9. An electronic device, characterized in that, Include a memory and a processor; The memory is used to store computer programs; The processor is used to, when executing the computer program, implement the geometric correction method of the satellite remote sensing image as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the storage medium, and when the computer program is executed by a processor, the geometric correction method of satellite remote sensing images according to any one of claims 1 to 8 is implemented.

Images