A method for correcting positioning errors of satellite laser footprints in mountainous forest areas

Abstract

The application discloses a mountainous forest satellite laser footprint positioning error correction method, which comprises the following steps: firstly, performing multi-condition constraint processing on satellite-borne laser footprints and GF7-DSM data, screening high-quality bare land laser footprints, and reconstructing RGF7-DSM data matched with the size of the laser footprints; secondly, introducing a small weight horizontal coordinate parameter, converting a traditional height residual minimum criterion into a spatial distance minimum criterion, calculating the distance between the weighted satellite-borne laser footprints and the RGF7-DSM different grids, and identifying the optimal position target grid of the laser footprints based on the multi-footprint distance and the minimum principle; and finally, constructing a distance residual function of the laser footprints to the target grid, minimizing the residual sum of squares by using a least square algorithm, directly solving the positioning error correction amount, and realizing the direct positioning of the optimal position of the laser footprints. Through the spatial distance conversion and multi-footprint joint solving strategy, the positioning precision of the satellite-borne laser footprints under the complex terrain of the mountainous forest is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of satellite-borne laser footprint positioning technology, and in particular to a method for correcting satellite laser footprint positioning errors in mountainous forest areas. Background Technology

[0002] Spaceborne lidar, with its ability to detect large-scale, high-precision three-dimensional structures of ground features, has been widely used in forest resource surveys such as global forest canopy height inversion and aboveground biomass estimation. High-precision positioning of laser footprints in forest areas is a crucial prerequisite for ensuring the accuracy and feasibility of these applications. However, in actual on-orbit operation, laser footprints generally suffer from significant positioning errors due to factors such as thermal effects and satellite jitter. For example, the positioning error of the GF7 laser footprint is approximately 5-10 meters, and the latest version of GEDI also exhibits a positioning error of around 10 meters. Especially in mountainous forest areas, the complex laser footprint echo signals, influenced by terrain undulations and vegetation structure, lead to even more significant positioning errors. Furthermore, laser positioning errors also cause more severe vertical elevation measurement errors in mountainous forest areas. This not only reduces the accuracy of ground elevation measurements using laser footprints in mountainous forest areas but also increases the error in canopy height inversion, limiting the effective application of spaceborne laser footprints in forest resource surveys. Therefore, correcting the positioning error of spaceborne laser footprints in mountainous forest areas is a primary research focus for improving the accuracy of its industry applications.

[0003] Currently, spaceborne laser positioning error correction methods can be divided into three categories. The first category is site-specific correction methods, including two strategies: ground-based calibration fields and ocean-based calibration fields. Ground-based calibration typically requires deploying reflective targets or ground detectors in open, flat areas to capture the true location of the footprints and use these as high-precision ground control points to correct footprint positioning errors. Ocean-based calibration typically selects a calm ocean surface as the experimental site. By designing pitch and roll attitude maneuvers, satellite conical scanning is achieved to collect periodic observation data, which is then combined with least-squares correction of footprint positioning errors. This type of method has high site requirements, needing to be implemented in specific locations such as flat, bare ground or calm oceans, and cannot be used for laser footprint positioning error correction in mountainous or forested areas.

[0004] The second type is the waveform matching-based correction method. This method simulates waveforms around the footprint recording location, calculates their similarity to the recorded waveform, and corrects the error based on the principle of highest similarity. For example, Wang et al. used this method to evaluate the positioning error of ICESat / GLAS laser footprints under different land cover types, concluding that this method is more applicable in urban scenarios with high spatial heterogeneity, and that the footprint positioning error was 8.19m. Lang et al. and Liu et al. used the GEDI simulator developed by Hancock's team to simulate footprint waveforms and corrected the GEDI footprint positioning error based on similarity indices. Ni et al. applied this method to a mountain forest biomass estimation scenario, and the GEDI footprint positioning accuracy improved after error correction. This method can effectively correct laser footprint positioning errors, but some problems still exist in mountainous forest areas. For example, because waveform simulation cannot completely reproduce laser parameters, ground reflection characteristics, and the complex terrain-vegetation features of mountainous forest areas, the accuracy of the simulated waveform is low, resulting in a certain difference between the simulated waveform and the recorded waveform. This difference reduces the matching accuracy between the analog waveform and the recorded waveform, which in turn affects the correction accuracy of footprint positioning errors in mountainous forest areas.

[0005] The third category is correction methods based on terrain matching. This method calculates the elevation residuals between the laser footprint record elevation and the reference terrain at different locations, and corrects the laser footprint positioning error based on the principle of minimizing the sum of multiple footprint elevation residuals. Schleich et al. used airborne LiDAR as the reference terrain data, set a fixed step size to traverse and calculate the footprint elevation residuals, and introduced a flow accumulation method to approximate the minimum residual position, thus correcting the GEDI forest footprint positioning error. Yang et al. used this method, introducing multiple methods such as minimum value, geometric center, and Gaussian fitting to approximate the optimal footprint position, revealing the positioning errors at three scales: GEDI footprint, beam, and orbit, and finding that the positioning error at the footprint scale is the largest, approximately 15m. Liu et al., focusing on GF-7, calculated laser ranging values ​​at different ground locations based on satellite DSM, and used surface fitting to approximate the minimum position of the residuals between the laser ranging and the actual ranging, achieving correction of the forest area laser footprint positioning error. Although these methods correct the forest area footprint positioning error to some extent, the correction accuracy is limited by the fitting model and the resolution of the reference terrain. More importantly, these methods cannot directly determine the optimal location of the footprint, but rather approximate it gradually through indirect fitting. In mountainous forest areas with significant topographic relief and complex vegetation, there is still a certain error between the approximate optimal location and the actual location, thus limiting the accuracy of laser footprint positioning error correction in mountainous forest areas.

[0006] Furthermore, both waveform matching-based and terrain matching-based methods rely on high-precision airborne LiDAR data, which has significant limitations in terms of data coverage and acquisition cost. Summary of the Invention

[0007] The purpose of this invention is to provide a method for correcting satellite laser footprint positioning errors in mountainous forest areas, thereby solving the aforementioned problems existing in the prior art.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] A method for correcting satellite laser footprint positioning errors in mountainous forest areas includes the following steps:

[0010] Spaceborne laser footprint and GF7-DSM data processing: using multiple constraints to screen high-quality bare ground laser footprints and reconstructing RGF7-DSM that can match the scale of the laser footprints;

[0011] The "optimal location target grid identification" of the spaceborne laser footprint is achieved by introducing a low-weight horizontal coordinate parameter, which transforms the traditional minimum elevation residual into the minimum spatial distance, calculates the distance between the spaceborne laser footprint and different grids of RGF7-DSM after horizontal coordinate weighting, and identifies the optimal location target grid of the laser footprint based on the principle of minimum distance sum.

[0012] The "optimal position direct positioning" of the spaceborne laser footprint is achieved by constructing the distance residual function from the laser footprint to the target grid, and by jointly minimizing the sum of squared residuals using least squares to calculate the positioning error correction.

[0013] Furthermore, the processing of spaceborne laser footprints and GF7-DSM data specifically includes:

[0014] A multi-condition constraint quality control strategy was constructed to systematically screen high-quality GF7 bare ground laser footprints: basic quality identification screening to retain valid laser observations; terrain consistency verification to select GF7 laser footprints with an absolute elevation difference of less than 2m from the reference terrain data.

[0015] Elevation consistency reconstruction of GF7-DSM data considering the scale of laser footprints: First, taking advantage of the simultaneous and same-platform nature of GF7 satellite full-waveform laser / optical imagery, a 3m resolution DSM is produced using the laser footprint as elevation control and in conjunction with optical imagery; then, to ensure the consistency between the reference data elevation and the footprint ground elevation, a neighborhood window of the same size as the GF7 laser footprint is constructed, all DSM pixels within the window are statistically analyzed, and their average elevation is used as the reconstructed elevation of that pixel, generating the elevation-reconstructed DSM.

[0016] Furthermore, the "optimal location target grid recognition" of spaceborne laser footprints includes the following process:

[0017] Construction of a square ground search domain based on laser footprints;

[0018] Laser footprint and square ground search domain horizontal coordinate weighting;

[0019] Calculation of the distance from the laser footprint to the square ground search grid;

[0020] Grid distances and set generation;

[0021] Laser footprint "optimal location target grid" recognition.

[0022] Furthermore, the construction of the square ground search domain for laser footprints involves creating a square search domain centered on the laser footprint recording location, with the search domain range set to... The grid spacing is set to 3m, and the elevation is obtained by interpolation using RGF7-DSM, forming a three-dimensional grid set:

[0023] In the formula, Let the side length of the square search domain be . This refers to the officially released laser footprint positioning error.

[0024] ;

[0025] In the formula, Represents the three-dimensional coordinates of the grid in the i-th row and j-th column. The value is obtained by interpolation at this grid point using RGF7-DSM; The number of rows and columns of the square ground search domain; It is the set of three-dimensional points formed by all grids within the square search domain.

[0026] Furthermore, the horizontal coordinates of the laser footprint and the square ground search area are weighted, with a horizontal coordinate weighting factor β=10 for both the laser footprint and the square ground search area. -3 After weighting, the footprint coordinates were obtained respectively. And search domain grid coordinates:

[0027] ;

[0028] In the formula, The horizontal coordinate weighting coefficient is taken in this paper. ; Indicates the first The weighted 3D coordinates of the laser footprints; This indicates all grid lines within the square ground search domain. The weighted 3D point set, where , , The number of rows and columns of the square ground search domain. Obtained by interpolation from RGF7-DSM.

[0029] Furthermore, it also includes: calculating the weighted distances between different grids of the square ground search domain from the spaceborne laser footprints, including:

[0030] First, fit the grid. Obtaining a finite plane ;

[0031] Secondly, calculate laser footprints. to plane Spatial distance: if The projection falls on If it is inside, then the perpendicular distance is calculated directly. If the projection falls on the extended surface Above, then for To P The shortest distance to the boundary.

[0032] Furthermore, the generation of grid distances and sets is used to eliminate random errors caused by individual laser footprints.

[0033] Introduction A continuous high-quality laser footprint joint solution, calculating The distances from each laser footprint to the grid are summed to obtain... ;

[0034] Traverse the entire square ground search area and perform statistics Distances and sets of footprints to all grid cells The formula is:

[0035] In the formula, The number of rows and columns of the square ground search domain. for A footprint OK Leggnet The distance and, for A footprint OK Leggnet The distance and.

[0036] Furthermore, the laser footprint "optimal location target grid" identification specifically involves: jointly weighting the grid coordinates and distance sums to construct a three-dimensional discrete point set; identifying the grid corresponding to the minimum distance sum, which is the optimal location target grid, thus obtaining a three-dimensional target grid set.

[0037] Furthermore, the "direct optimal location positioning" of spaceborne laser footprints is based on n laser footprints and the identified target grid. It constructs a distance residual function from the laser footprint to the target grid and uses a least squares algorithm to minimize the sum of squared residuals to directly calculate the laser footprint positioning error correction. Then The coordinates are superimposed onto the laser footprint record to directly locate the optimal position of the footprint and complete the laser footprint positioning error correction;

[0038] The formula for minimizing the sum of squared residuals using the joint least squares algorithm is as follows:

[0039]

[0040] In the formula, This is the correction amount for laser footprint positioning error. For the first An index of footprints, This represents the number of laser footprints involved in the calibration. For laser footprints To the target grid The minimum distance, βx+βΔ, βy+βΔ, are the corrected coordinates obtained by applying corrections Δ to the weighted horizontal coordinates (x, y) of the laser footprint, and z is the elevation coordinate of the laser footprint recording position.

[0041] The beneficial effects of this invention are:

[0042] This invention overcomes the challenge of traditional terrain matching methods that rely on high-precision airborne LiDAR data and fitting models, and cannot directly locate the optimal position of laser footprints, by employing a satellite laser footprint positioning error correction method for mountainous forest areas without requiring high-precision airborne LiDAR point cloud data. The method innovatively introduces a small-weighted horizontal coordinate parameter to weight the horizontal coordinates of the satellite laser footprint and the RGF7-DSM, iteratively calculating the weighted distances between the laser footprint and different terrain grids, and identifying the target grid for the optimal footprint position based on the principle of minimum distance. Subsequently, it uses least squares to directly calculate the minimum distance within the target grid, accurately locating the optimal footprint position, thus effectively achieving high-precision correction of laser footprint positioning errors in mountainous forest areas. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the positioning error of the spaceborne laser footprint described in this invention;

[0044] Figure 2 A flowchart illustrating an embodiment of the present invention;

[0045] Figure 3 Schematic diagram of GF7DSM elevation consistency reconstruction according to an embodiment of the present invention;

[0046] Figure 4 A schematic diagram of the minimum distance calculation model between spaceborne laser footprints and different grids of GF7-DSM according to the embodiments of the invention;

[0047] Figure 5 A schematic diagram illustrating the distances between each grid cell in the laser footprint search domain on a square ground surface according to an embodiment of the present invention;

[0048] Figure 6 This is a schematic diagram illustrating the minimum value of fitting a three-dimensional discrete point set according to an embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] Reference Figures 1 to 6 The method for correcting satellite laser footprint positioning errors in mountainous forest areas, as shown, includes the following steps:

[0051] S100, spaceborne laser footprint and GF7-DSM data processing, namely, using multi-condition constraints to screen high-quality bare ground laser footprints and reconstruct RGF7-DSM that can match the scale of laser footprints;

[0052] S200, the "Optimal Position Target Grid Recognition" of the spaceborne laser footprint, that is, by introducing a small weighted horizontal coordinate parameter, the traditional minimum elevation residual is transformed into the minimum spatial distance, and the distance between the spaceborne laser footprint and different grids of RGF7-DSM after horizontal coordinate weighting is calculated, and the optimal position target grid of the laser footprint is identified based on the principle of minimum distance sum.

[0053] The S300 satellite-borne laser footprint "optimal position direct positioning" is achieved by constructing the distance residual function from the laser footprint to the target grid, and by jointly minimizing the sum of squared residuals using least squares to calculate the positioning error correction, thus realizing the optimal position direct positioning of the laser footprint.

[0054] Furthermore, the processing of spaceborne laser footprints and GF7-DSM data specifically includes:

[0055] A multi-condition constraint quality control strategy was constructed to systematically screen high-quality GF7 bare ground laser footprints: basic quality identification screening to retain valid laser observations; terrain consistency verification to select GF7 laser footprints with an absolute elevation difference of less than 2m from the reference terrain data.

[0056] In the above content, it should be noted that the data processing process of spaceborne laser footprints and GF7-DSM data is explained first. This part is the preprocessing of the input data of the method, which involves two parts: high-quality, bare ground spaceborne laser footprint data screening based on multiple constraints and GF7-DSM data elevation consistency reconstruction taking into account the scale of laser footprints.

[0057] In the screening of high-quality spaceborne laser footprint data based on multiple constraints, a quality control strategy with multiple constraints was constructed to systematically screen high-quality GF7 laser footprints. This mainly includes basic quality identification screening: retaining valid laser observations; terrain consistency verification: selecting GF7 laser footprints with an absolute elevation difference of less than 2m from the reference terrain data and a terrain slope greater than 2m.

[0058] To ensure consistent elevation of GF7-DSM data at the laser footprint scale: First, leveraging the advantages of simultaneous and simultaneous spatial and platform-based full-waveform laser / optical imagery from the GF7 satellite, a 3m resolution DSM was produced using the laser footprint as elevation control, in conjunction with optical imagery; this serves as the reference data for this study. Subsequently, to ensure consistency between the reference data elevation and the footprint ground elevation, a neighborhood window of the same size as the GF7 laser footprint was constructed. All DSM pixels within the window were statistically analyzed, and their average elevation was used as the reconstructed elevation of that pixel. This process was repeated to reconstruct the elevation of all pixels, generating the reconstructed DSM (RGF7-DSM).

[0059] Furthermore, the "optimal location target grid recognition" of spaceborne laser footprints includes the following process:

[0060] Construction of a square ground search domain based on laser footprints;

[0061] Laser footprint and square ground search domain horizontal coordinate weighting;

[0062] Calculation of the distance from the laser footprint to the square ground search grid;

[0063] Grid distances and set generation;

[0064] Laser footprint "optimal location target grid" recognition.

[0065] Furthermore, the construction of the square ground search domain for laser footprints involves creating a square search domain centered on the laser footprint recording location, with the search domain range set to... The grid spacing is set to 3m, and the elevation is obtained by interpolation using RGF7-DSM, forming a three-dimensional grid set:

[0066] In the formula, Let the side length of the square search domain be . This refers to the officially released laser footprint positioning error.

[0067] ;

[0068] In the formula, Represents the three-dimensional coordinates of the grid in the i-th row and j-th column. The value is obtained by interpolation at this grid point using RGF7-DSM; The number of rows and columns of the square ground search domain; It is the set of three-dimensional points formed by all grids within the square search domain.

[0069] Furthermore, the horizontal coordinates of the laser footprint and the square ground search area are weighted, with a horizontal coordinate weighting factor β=10 for both the laser footprint and the square ground search area. -3 After weighting, the footprint coordinates were obtained respectively. And search domain grid coordinates:

[0070] ;

[0071] In the formula, The horizontal coordinate weighting coefficient is taken in this paper. ; Indicates the first The weighted 3D coordinates of the laser footprints; This indicates all grid lines within the square ground search domain. The weighted 3D point set, where , , The number of rows and columns of the square ground search domain. Obtained by interpolation from RGF7-DSM.

[0072] Furthermore, it also includes: calculating the weighted distances between different grids of the square ground search domain from the spaceborne laser footprints, including: using the footprints... Grid For example:

[0073] First, fit the grid. Obtaining a finite plane ;

[0074] Secondly, calculate laser footprints. to plane Spatial distance: if The projection falls on If it is inside, then the perpendicular distance is calculated directly. If the projection falls on the extended surface Above, then for To P The shortest distance to the boundary.

[0075] Furthermore, the generation of distances and sets for each grid cell in the spaceborne laser footprint-square ground search domain is used to eliminate random errors caused by individual laser footprints.

[0076] Introduction A continuous high-quality laser footprint joint solution, using a grid in the i-th row and j-th column. For example, calculate The distances from each laser footprint to the grid are summed to obtain... ;

[0077] Traverse the entire square ground search area and perform statistics Distances and sets of footprints to all grid cells The formula is: ;

[0078] In the formula, The number of rows and columns of the square ground search domain. for A footprint OK Leggnet The distance and, for A footprint OK Leggnet The distance and.

[0079] Furthermore, the laser footprint "optimal location target grid" identification specifically involves: jointly weighting the grid coordinates and distance sums to construct a three-dimensional discrete point set; identifying the grid corresponding to the minimum distance sum, which is the optimal location target grid, thus obtaining a three-dimensional target grid set.

[0080] Furthermore, the "direct optimal location positioning" of spaceborne laser footprints is based on n laser footprints and the identified target grid. It constructs a distance residual function from the laser footprint to the target grid and uses a least squares algorithm to minimize the sum of squared residuals to directly calculate the laser footprint positioning error correction. Then The coordinates are superimposed onto the laser footprint record to directly locate the optimal position of the footprint and complete the laser footprint positioning error correction;

[0081] The formula for minimizing the sum of squared residuals using the joint least squares algorithm is as follows:

[0082]

[0083] In the formula, This is the correction amount for laser footprint positioning error. For the first An index of footprints, This represents the number of laser footprints involved in the calibration. For laser footprints To the target grid The minimum distance, βx+βΔ, βy+βΔ, are the corrected coordinates obtained by applying corrections Δ to the weighted horizontal coordinates (x, y) of the laser footprint, and z is the elevation coordinate of the laser footprint recording position.

[0084] The following is a specific example. Taking the laser footprint data from my country's GF7 satellite as an example, the experimental data is selected from the GF7 laser footprint data passing through Pu'er, Yunnan Province from 2021 to 2024. The invention is implemented based on the above data.

[0085] First, based on constraints such as basic quality, terrain, and elevation consistency, high-quality bare ground laser footprints were selected. Simultaneously, to match the observation scale of the laser footprints, the advantages of simultaneous spatial-temporal and platform-based analysis of GF7 satellite full-waveform laser / optical imagery were utilized. Using the laser footprints as elevation controls, a 3m resolution DSM (Digital Smart Scale) was produced in conjunction with the optical imagery, serving as the reference data for this study. Then, a circular region with a diameter of 20m (taking the size of a GF7 footprint as an example) was constructed, centered on each grid vertex of the DSM. The average grid elevation within the circular region was used as the reconstructed elevation value for that vertex. The elevations of all grid vertices in the DSM were reconstructed sequentially to obtain the RGF7-DSM. Subsequently, a square ground search domain with a side length of 100m was constructed, centered on each laser footprint recording location, with a grid spacing of 3m. The corresponding elevation information was obtained through interpolation using the RGF7-DSM. A planar coordinate weighting factor β=10 was introduced. -3 The planar coordinates of the laser footprints and each grid cell in the search domain are uniformly weighted. The spatial distances from each footprint to the grid cells are calculated and summed to identify the grid cell with the minimum distance, which is then determined as the optimal target grid cell. Finally, a distance residual function from the laser footprints to the target grid cells is constructed. The sum of squared residuals is minimized using the least squares method to solve for the error correction, thereby achieving direct localization of the optimal laser footprint position.

[0086] The experimental results are shown in Table 1.

[0087] Table 1

[0088]

[0089] By adopting the above-disclosed technical solution of this invention, the following beneficial effects are obtained:

[0090] This invention overcomes the challenge of traditional terrain matching methods that rely on high-precision airborne LiDAR data and fitting models, and cannot directly locate the optimal position of laser footprints, by employing a satellite laser footprint positioning error correction method for mountainous forest areas without requiring high-precision airborne LiDAR point cloud data. This method innovatively introduces a small-weighted horizontal coordinate parameter to weight the horizontal coordinates of the satellite laser footprint and the RGF7-DSM, iteratively calculating the weighted distances between the laser footprint and different terrain grids, and identifying the target grid for the optimal footprint position based on the principle of minimum distance. Subsequently, it uses least squares to directly calculate the minimum distance within the target grid, accurately locating the optimal footprint position, effectively achieving high-precision correction of laser footprint positioning errors in mountainous forest areas. The above description is merely a preferred embodiment of this invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention.

Claims

1. A method for correcting satellite laser footprint positioning errors in mountainous forest areas, characterized in that, Includes the following steps: Spaceborne laser footprints and GF7-DSM data processing: High-quality bare ground laser footprints are screened using multi-condition constraints, and RGF7-DSM that can match the scale of the laser footprints are reconstructed. The optimal location target grid for spaceborne laser footprints is identified by introducing a low-weight horizontal coordinate parameter, which transforms the traditional minimum elevation residual into the minimum spatial distance. The distance between the spaceborne laser footprints and different grids of RGF7-DSM after horizontal coordinate weighting is calculated, and the optimal location target grid for the laser footprints is identified based on the principle of minimum distance sum. The optimal location of the spaceborne laser footprint can be directly located by constructing the distance residual function from the laser footprint to the target grid, and by combining least squares minimization of the sum of squared residuals to calculate the positioning error correction, thus achieving the direct location of the optimal laser footprint. The optimal location target grid identification of the spaceborne laser footprint includes the following process: Construction of a square ground search domain based on laser footprints; Laser footprint and square ground search domain horizontal coordinate weighting; Calculation of the distance from the laser footprint to the square ground search grid; Grid distances and set generation; Target grid identification of optimal location for laser footprints; The horizontal coordinates of the laser footprint and the square ground search area are weighted, with a horizontal coordinate weighting factor applied to both. β=10 -3 After weighting, the footprint coordinates were obtained respectively. And search domain grid coordinates: ； In the formula, The horizontal coordinate weighting coefficient is taken in this paper. ; Indicates the first The weighted 3D coordinates of the laser footprints; This indicates all grid lines within the square ground search domain. The weighted 3D point set, where , , The number of rows and columns of the square ground search domain. Obtained by interpolation from RGF7-DSM.

2. The method according to claim 1, characterized in that, The specific processing of the spaceborne laser footprint and GF7-DSM data includes: A multi-condition constraint quality control strategy was constructed to systematically screen high-quality GF7 bare ground laser footprints: basic quality identification screening to retain valid laser observations; terrain consistency verification to select GF7 laser footprints with an absolute elevation difference of less than 2m from the reference terrain data. Elevation consistency reconstruction of GF7-DSM data considering the scale of laser footprints: First, taking advantage of the simultaneous and same-platform nature of GF7 satellite full-waveform laser / optical imagery, a 3m resolution DSM is produced using the laser footprint as elevation control and in conjunction with optical imagery; then, to ensure the consistency between the reference data elevation and the footprint ground elevation, a neighborhood window of the same size as the GF7 laser footprint is constructed, all DSM pixels within the window are statistically analyzed, and their average elevation is used as the reconstructed elevation of that pixel, generating the elevation-reconstructed DSM.

3. The method according to claim 2, characterized in that, The laser footprint square ground search domain is constructed by using the laser footprint recording location as the center to construct a square search domain, and the search domain range is set to... The grid spacing is set to 3m, and the elevation is obtained by interpolation using RGF7-DSM, forming a three-dimensional grid set: In the formula, Let the side length of the square search domain be . This refers to the officially released laser footprint positioning error. ； In the formula, Represents the three-dimensional coordinates of the grid in the i-th row and j-th column. The value is obtained by interpolation at this grid point using RGF7-DSM; The number of rows and columns of the square ground search domain; It is the set of three-dimensional points formed by all grids within the square search domain.

4. The method according to claim 3, characterized in that, Also includes: Calculate the weighted distances between different grids of the square ground search domain from the spaceborne laser footprints, including: First, fit the grid. Obtaining a finite plane ; Secondly, calculate laser footprints. to plane Spatial distance: if The projection falls on If it is inside, then the perpendicular distance is calculated directly. If the projection falls on the extended surface Above, then for To P The shortest distance to the boundary.

5. The method according to claim 3, characterized in that, The generation of the grid distances and sets is intended to eliminate random errors caused by individual laser footprints. Introduction A continuous high-quality laser footprint joint solution, calculating The distances from each laser footprint to the grid are summed to obtain... ; Traverse the entire square ground search area and perform statistics Distances and sets of footprints to all grid cells The formula is: ； In the formula, The number of rows and columns of the square ground search domain. for A footprint OK Leggnet The distance and, for A footprint OK Leggnet The distance and.

6. The method according to claim 3, characterized in that, The laser footprint optimal position target grid identification specifically involves: jointly weighting the grid coordinates and distance sum to construct a three-dimensional discrete point set, identifying the grid corresponding to the minimum distance sum as the optimal position target grid, and obtaining a three-dimensional target grid set.

7. The method according to claim 1, characterized in that, The direct localization of the optimal location of the spaceborne laser footprints is based on n laser footprints and the identified target grid. A distance residual function from the laser footprint to the target grid is constructed, and the sum of squared residuals is minimized using a least squares algorithm to directly calculate the laser footprint localization error correction. Then The coordinates are superimposed onto the laser footprint record to directly locate the optimal position of the footprint and complete the laser footprint positioning error correction; The formula for minimizing the sum of squared residuals using the joint least squares algorithm is as follows: ； In the formula, This is the correction amount for laser footprint positioning error. For the first An index of footprints, This represents the number of laser footprints involved in the calibration. For laser footprints To the target grid The minimum distance, βx+βΔ, βy+βΔ, are the corrected coordinates obtained by applying corrections Δ to the weighted horizontal coordinates (x, y) of the laser footprint, and z is the elevation coordinate of the laser footprint recording position.

Images