Airborne LiDAR bathymetric wave correction method and system

Abstract

The present invention belongs to the technical field of marine surveying and mapping, and discloses an airborne LiDAR bathymetric wave correction method and system. The method calculates wave parameters according to the Elfouhaily spectral model; uses the K-nearest neighbor search method accelerated by KD-Tree to determine the search neighborhood with wave parameters to realize the search of sea surface points in the neighborhood; constructs a multi-scale irregular TIN triangulated network array sea surface model based on the efficient search results of sea surface points in the neighborhood, calculates the normal vector of each beam of sea surface laser point with seabed reflection echo, and performs ray tracing to correct the wave error to obtain accurate seabed point coordinates. The present invention can reduce the seabed bathymetric error caused by the path deflection of the laser when it enters the water from the air due to sea surface waves, improve the bathymetric accuracy of the ALB system, and provide strong support for marine resource development, marine environmental protection, etc.

Description

Technical Field

[0001] The present invention belongs to the technical field of ocean surveying and mapping, and in particular relates to an airborne LiDAR bathymetric wave correction method and system. Background Art

[0002] The Airborne LiDAR Bathymetry (ALB) system uses 532nm laser pulses to determine water depth information through the round-trip travel time difference between the sea surface and the seabed. It has the characteristics of high measurement accuracy, high efficiency, strong mobility, and amphibious use. It can obtain high-quality, full-coverage coastal integrated spatial geographic information data above and below the water. However, the process of the green bathymetric laser passing through the sea-air interface involves two media, the atmosphere and seawater. The different medium densities cause the transmission direction of the laser to change after it enters the water. If the sea surface is assumed to be flat, the laser ray tracing can be simply achieved using the law of refraction. However, in reality, the sea-air interface will form randomly undulating waves due to environmental factors such as sea surface winds, which is specifically manifested as the corresponding deflection of the instantaneous sea surface normal vector. Therefore, an effective method must be used to correct the wave error. The wave correction method for airborne LiDAR bathymetry can reduce the refraction effect error caused by sea surface waves. It can not only provide technical support for the refined processing of airborne LiDAR bathymetry data, but also provide high-precision data for underwater topography measurements in coastal areas and other areas, promoting the development and progress of my country's marine science, marine surveying and mapping and other disciplines.

[0003] The existing wave correction methods mainly include: using the B-spline surface to fit the sea surface points to construct a sea surface model for ray tracing. The sea surface model constructed by this method can meet the requirements of natural wave takeoff, but the calculation is complicated; by solving the fluid dynamics equation, the water surface is represented by the height field to solve the normal vector for ray tracing. This method has a realistic simulation effect, but the simulation efficiency is often low due to solving nonlinear equations; using the wave model constructed by the wave spectrum for ray tracing. This method has high accuracy but requires a lot of calculation time; the linear wave model method has high efficiency but low accuracy.

[0004] Through the above analysis, the problems and defects of the existing technology are as follows: the establishment of the existing ALB wave correction platform is mostly at the simulation level, and there are few studies on using the measured sea surface point cloud to construct the instantaneous sea surface model. The use of the wave spectrum to construct the wave model can accurately correct the wave error of each laser beam, but the calculation is cumbersome and time-consuming. Studying a sea surface modeling method that takes into account both algorithm accuracy and efficiency is an important direction that ALB wave error correction urgently needs to pay attention to. Summary of the invention

[0005] In order to overcome the problems existing in the related art, the disclosed embodiment of the present invention provides an airborne LiDAR bathymetric wave correction method and system, which specifically relates to a sea surface modeling and ray tracing method that takes into account both algorithm accuracy and efficiency to perform airborne LiDAR bathymetric wave correction.

[0006] The technical solution is as follows: an airborne LiDAR bathymetric wave correction method, comprising the following steps:

[0007] S1, calculate wave parameters according to the Elfouhaily spectrum model, wave parameters include wave width and wave height;

[0008] S2, using the K-nearest neighbor search method accelerated by KD-Tree, the search neighborhood is determined by wave parameters to realize the search of sea surface points within the neighborhood;

[0009] S3, based on the efficient search results of sea surface points in the neighborhood, construct a multi-scale irregular TIN triangulated network array sea surface model, calculate the normal vector of each sea surface laser point with seabed reflection echo, and perform ray tracing to correct the wave error to obtain the precise seabed point coordinates.

[0010] In step S1, wave parameters are calculated according to the Elfouhaily spectral model, including:

[0011] S101, divide the entire sea surface point cloud into voxel grids, retain the highest point in each grid, collect wind speed synchronously during the experiment, and match the wind speed corresponding to each laser point according to time;

[0012] S102, Calculation of wave width using the Elfouhaily wave spectrum model and wave height , the expression is:

[0013] (1);

[0014] In the formula, is the Elfouhaily wave spectrum model, is the spectral amplitude coefficient, is the acceleration due to gravity, is the wave number, is an empirical parameter, is the spectrum peak frequency, is the direction distribution function, is the maximum wave number, is the minimum wave number, is the differential of the direction angle, is the wave number derivative, is the direction angle, is the wind speed.

[0015] In step S2, the K-nearest neighbor search method accelerated by KD-Tree is used to determine the search neighborhood with wave parameters to search for sea surface points in the neighborhood, including:

[0016] S201, construct KD-Tree: calculate the variance, select the dimension with the largest variance as the segmentation dimension of the current layer; in the selected segmentation dimension Above, sort the point cloud data, select the median as the split point, divide the data set into two left and right subsets with the median as the boundary, recursively build the subtree, and recursively perform the above operations on the left and right subsets until the preset tree depth is reached;

[0017] S202, K-nearest neighbor search: based on the wave width and wave height , adaptively determine the search radius and the number of nearest neighbors ; For query point , recursively traverse from the root node of the KD-Tree, and decide whether to enter the left subtree or the right subtree by comparing the coordinates of the query point on the current segmentation dimension with the segmentation value of the node. At each node, calculate the Euclidean distance between the query point and the current node, and update the nearest neighbor set. If the number of points in the current set is less than , then add directly, has reached points, replace the farthest point and update the maximum distance ; During recursive backtracking, determine whether the distance from the splitting hyperplane to the query point is less than the current maximum distance This judgment is used to decide whether to continue searching the other side subtree. If the distance from the hyperplane to the query point is greater than , then prune and no longer search the subtree on the other side, thus avoiding unnecessary calculations; if the distance is less than or equal to , then continue to recursively search the other side subtree to check if there is a closer point.

[0018] In step S201, the dimension is segmented The calculation formula is:

[0019] (2);

[0020] In the formula, For the dimensional variance, For all points The total number of are three spatial coordinate dimensions, For the dimension, For the The point is The value on the dimension, For the The mean of the dimension;

[0021] The data set is divided into two subsets, left and right, and the calculation formula for recursively constructing the subtree is:

[0022] (3);

[0023] In the formula, is the left subset, is the right subset, Summarize the dataset Points, is a set of data points, For data points In the split dimension The value on Split dimensions into a dataset The median value of all points on the .

[0024] In step S202, the search radius is adaptively determined and the number of nearest neighbors The calculation formula is:

[0025] (4);

[0026] In the formula, All are adjustment factors.

[0027] In step S3, before constructing the multi-scale irregular TIN triangulated network array sea surface model, multi-scale neighborhood point screening is performed, and the K-nearest neighbor search method accelerated by KD-Tree is used to select the nearest neighbor points according to each query point. Wave width and wave height , adaptively determine the search radius and the number of nearest neighbors ; By different search radius Value, divide the entire sea surface point cloud into multiple point sets of different scales .

[0028] In step S3, the normal vector of each sea surface laser point with seabed reflection echo is calculated, including: at each scale, Delaunay triangulation is performed on each neighborhood point set obtained by screening to construct a triangulated network. For each triangle , calculate the normal vector , the calculation formula is:

[0029] (5);

[0030] In the formula, are the three vertices of the triangle;

[0031] The normal vector information at different scales is integrated, and the weighted average method is used to fuse the normal vectors of each scale. The calculation formula for weighting is:

[0032] (6);

[0033] In the formula, is the normal vector The weight coefficient of For the The total number of triangles under scale;

[0034] Calculate the weighted average of the normal vectors of the triangles where the sea surface points at each scale are located , the calculation formula is:

[0035] (7);

[0036] In the formula, is the number of triangles.

[0037] In step S3, ray tracing is performed to correct the wave error to obtain the accurate coordinates of the seabed point, including: calculating the refraction vector according to Snell's law, and the calculation formula is:

[0038] (8);

[0039] In the formula, is the refraction vector, is the incident vector, is the refractive index ratio, is the laser incident angle, is the laser refraction angle;

[0040] The corresponding seabed point coordinates are calculated based on the sea surface point coordinates, the refraction vector and the corresponding underwater slope distance. The calculation formula is:

[0041] (9);

[0042] In the formula, is the coordinate of the laser seabed point in the WGS-84 coordinate system, for X , Y , Z The components of the refraction vector in three directions, is the coordinate of the laser sea surface point in the WGS-84 coordinate system, is the laser underwater slant range.

[0043] Another object of the present invention is to provide an airborne LiDAR bathymetric wave correction system, which implements the airborne LiDAR bathymetric wave correction method, and the system comprises:

[0044] The wave parameter calculation module is used to calculate the wave parameters according to the Elfouhaily spectrum model. The wave parameters include wave width and wave height;

[0045] The sea surface point search module in the neighborhood is used to use the K-nearest neighbor search method accelerated by KD-Tree to determine the search neighborhood with wave parameters to realize the search of sea surface points in the neighborhood;

[0046] The wave error correction module is used to construct a multi-scale irregular TIN triangulated network array sea surface model based on the efficient search results of sea surface points in the neighborhood, calculate the normal vector of each sea surface laser point with seabed reflection echo, and perform ray tracing to correct the wave error to obtain the precise coordinates of the seabed point.

[0047] Furthermore, the airborne LiDAR bathymetric wave correction system is carried on a computer-readable storage medium, and the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the functions of the above-mentioned airborne LiDAR bathymetric wave correction system can be realized.

[0048] Combining all the above technical solutions, the beneficial effects of the present invention are as follows: In view of the problem that the cumbersome construction of the sea surface model leads to low efficiency in calculating the sea surface normal vector, a wave error correction algorithm based on KNN-TIN space optimization is proposed. First, the wave parameters are calculated according to the Elfouhaily spectral model; then, the K-NN algorithm accelerated by KD-Tree is used to realize efficient search of sea surface measurement points in the neighborhood, and the search neighborhood is determined by the wave parameters; finally, a multi-scale irregular TIN triangulated network array sea surface model is constructed to calculate the normal vector of each laser point, and ray tracing is performed to correct the wave error to obtain accurate seabed point coordinates. The airborne LiDAR wave correction of the present invention solves the problem of insufficient accuracy of airborne LiDAR bathymetry and seabed bathymetry caused by sea surface waves. It can not only provide technical support for the refined processing of airborne LiDAR bathymetry data, but also provide high-precision data for underwater topography measurements in coastal areas and other areas, and promote research and application development in the fields of marine science and marine surveying and mapping.

[0049] The present invention can reduce the seabed sounding error caused by the deflection of the laser path when it enters the water from the air due to sea surface waves, improve the sounding accuracy of the ALB system, and surpass the depth measurement accuracy standard of the "Marine Engineering Topography Specification", thereby obtaining an integrated high-precision airborne LiDAR sounding three-dimensional point cloud above and below the water, and then a high-precision seabed terrain model can be established, providing strong support for marine resource development, marine environmental protection, etc. At present, the ALB system has been commercialized abroad, while my country's ALB technology started late, and there is no widely used commercial data processing software. Foreign equipment manufacturers usually regard their manufacturing technology and data processing methods as trade secrets and keep them confidential. Therefore, as the key basis for laser sounding data processing and analysis, this technical solution is of great significance for promoting the refined processing of domestic ALB sounding data. At present, the establishment of ALB wave correction platforms in China is mostly at the simulation level, and there are few studies on using measured sea surface point clouds to construct instantaneous sea surface models. Compared with most wave correction methods that need to use wave spectra to construct wave models to fit the sea surface, the present invention can quickly perform wave correction without fitting the sea surface, and can ensure accuracy while improving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure;

[0051] Figure 1 is a flow chart of an airborne LiDAR bathymetric wave correction method provided by an embodiment of the present invention;

[0052] Figure 2 Schematic diagram of the airborne LiDAR bathymetric wave correction method provided by an embodiment of the present invention;

[0053] Figure 3 It is the ALB bathymetric laser sea-air transmission path diagram provided by the present invention;

[0054] Figure 4 It is a cross-sectional view of the sea surface model of the multi-scale irregular TIN triangulated network array provided by the present invention;

[0055] Figure 5 It is the ALB and SBES point cloud profile diagram before strip wave correction provided by the present invention;

[0056] Figure 6 It is the ALB and SBES point cloud profile after strip-wave correction provided by the present invention;

[0057] Figure 7 It is the ALB and SBES point cloud profile of the strip 2 before wave correction provided by the present invention;

[0058] Figure 8 It is the ALB and SBES point cloud profile diagram after the strip two wave correction provided by the present invention;

[0059] Fig. 9 It is the relative distribution diagram of Mapper 20KU and SBES points provided by the present invention;

[0060] Fig.10 It is a scatter plot of the error before wave correction of the elevation difference between the same-name points of Mapper 20KU and SBES provided by the present invention;

[0061] Fig.11 It is a scatter diagram of the error after wave correction of the elevation difference between the same-name points of Mapper 20KU and SBES provided by the present invention. DETAILED DESCRIPTION

[0062] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.

[0063] The innovation of the present invention lies in that the dynamic changes and irregularities of the sea surface lead to the problem of low efficiency and accuracy of bathymetric ray tracing. An ALB wave error correction algorithm based on KNN-TIN space optimization is proposed. First, the wave parameters are calculated based on the wave spectrum model, and then the efficient search of sea surface points is realized based on the KD-Tree and K-nearest neighbor fusion model. Finally, a multi-scale TIN triangulated sea surface model is constructed, and the sea surface normal vector of the local sea surface fluctuation at the moment each laser pulse enters the water is calculated. Ray tracing is performed to solve the problem of insufficient ALB bathymetric accuracy caused by wave errors.

[0064] Aiming at the problem of low efficiency of sea surface normal vector calculation caused by cumbersome construction of sea surface model, a wave error correction algorithm based on KNN-TIN spatial optimization is proposed.

[0065] Embodiment 1, as Figure 1 As shown, the airborne LiDAR bathymetric wave correction method provided by the embodiment of the present invention includes:

[0066] S1, calculate wave parameters according to the Elfouhaily spectrum model, wave parameters include wave width and wave height;

[0067] It lays the foundation for the subsequent multi-scale irregular TIN (Triangulated Irregular Network, TIN) triangulated network array sea surface model modeling;

[0068] S2, using the K-nearest neighbor search method accelerated by KD-Tree, the search neighborhood is determined by wave parameters to realize the search of sea surface points within the neighborhood;

[0069] S3, based on the efficient search results of sea surface points in the neighborhood, construct a multi-scale irregular TIN triangulated network array sea surface model, calculate the normal vector of each sea surface laser point with seabed reflection echo, and perform ray tracing to correct the wave error to obtain the precise seabed point coordinates.

[0070] For example, in step S1 of the embodiment of the present invention, when the green laser of the airborne LiDAR bathymetry (ALB) system passes through the sea-air interface, the optical path will be deflected due to the influence of sea surface waves, such as Figure 3 As shown. Currently, commonly used wave spectrum models include PM spectrum, JONSWAP spectrum and Elfouhaily spectrum. Among them, PM spectrum (Pierson-Moskowitz spectrum) is mainly used to describe the fully developed wave spectrum under the steady sea surface; JONSWAP spectrum is based on PM spectrum, and the gravity wave part is corrected by introducing peak enhancement factor to be closer to the actual observed wave characteristics. In contrast, Elfouhaily spectrum comprehensively considers the low-frequency and high-frequency components in the wave spectrum to form a complete wave number spectrum. Since only the theoretical values ​​of the sea surface slope variance and height variance of Elfouhaily spectrum are consistent with the Cox-Munk formula, the Elfouhaily spectrum is selected to calculate the wave parameters.

[0071] In step S1 of the embodiment of the present invention, the wave parameter calculation based on the Elfouhaily wave spectrum model includes:

[0072] S101, divide the entire sea surface point cloud into voxel grids, retain the highest point in each grid, collect wind speed synchronously during the experiment, and match the wind speed corresponding to each laser point according to time;

[0073] S102, Calculation of wave width using the Elfouhaily wave spectrum model and wave height , the expression is:

[0074] (1);

[0075] In the formula, is the Elfouhaily wave spectrum model, is the spectral amplitude coefficient, is the acceleration due to gravity, is the wave number, is an empirical parameter, is the spectrum peak frequency, is the direction distribution function, is the maximum wave number, is the minimum wave number, is the differential of the direction angle, To differentiate the beam, is the direction angle, is the wind speed.

[0076] In step S2 of the embodiment of the present invention, the sea surface point cloud data obtained by the ALB system is scattered, and the distribution of the point cloud is irregular and cannot be described by a mathematical model. Therefore, it is impossible to query the point cloud coordinate information. Therefore, it is necessary to establish topological information to make the scanned point cloud data more regular, so as to realize the queryable functions and requirements, and KDTree is the most widely used method for establishing topological relationships.

[0077] Among them, the K-nearest neighbor search method (K-NN algorithm) accelerated by KD-Tree is used to determine the search neighborhood with wave parameters to realize the search of sea surface points in the neighborhood, including:

[0078] S201, construct KD-Tree: calculate the variance, select the dimension with the largest variance as the segmentation dimension of the current layer; in the selected segmentation dimension Above, sort the point cloud data, select the median as the split point, divide the data set into two left and right subsets with the median as the boundary, recursively build the subtree, and recursively perform the above operations on the left and right subsets until the preset tree depth is reached;

[0079] Split Dimension The calculation formula is:

[0080] (2);

[0081] In the formula, For the dimensional variance, For all points The total number of are three spatial coordinate dimensions, For the dimension, For the The point is The value on the dimension, For the The mean of the dimension;

[0082] The data set is divided into two subsets, left and right, and the calculation formula for recursively constructing the subtree is:

[0083] (3);

[0084] In the formula, is the left subset, is the right subset, Summarize the dataset Points, is a set of data points, For data points In the split dimension The value on Split dimensions into a dataset The median value of all points on the .

[0085] S202, K-nearest neighbor search: based on the wave width and wave height , adaptively determine the search radius and the number of nearest neighbors ; For query point , recursively traverse from the root node of the KD-Tree, and decide whether to enter the left subtree or the right subtree by comparing the coordinates of the query point on the current segmentation dimension with the segmentation value of the node. At each node, calculate the Euclidean distance between the query point and the current node, and update the nearest neighbor set. If the number of points in the current set is less than , then add directly, has reached points, replace the farthest point and update the maximum distance ; During recursive backtracking, determine whether the distance from the splitting hyperplane to the query point is less than the current maximum distance This judgment is used to decide whether to continue searching the other side subtree. If the distance from the hyperplane to the query point is greater than , then prune and no longer search the subtree on the other side, thus avoiding unnecessary calculations; if the distance is less than or equal to , then continue to recursively search the other side subtree to check if there is a closer point.

[0086] Adaptive determination of search radius and the number of nearest neighbors The calculation formula is:

[0087] (4);

[0088] In the formula, All are adjustment factors.

[0089] In step S3 of the embodiment of the present invention, the construction of the multi-scale irregular TIN triangulated network array sea surface model can save a lot of time compared to the sea surface fitting of the wave spectrum. The K-nearest neighbor search method accelerated by KD-Tree can adaptively divide the grid size. The construction of a multi-scale triangulated network according to different wave widths can more accurately reflect the detailed characteristics of waves at different scales and adapt to the wave performance of different wavelengths.

[0090] Wherein, step S3 specifically includes:

[0091] S301, multi-scale neighborhood point screening. In order to capture the multi-scale characteristics of sea surface waves, the KD-Tree accelerated K nearest neighbor search method is used to select the nearest neighbor points for each query point. Wave width and wave height , adaptively determine the search radius and the number of nearest neighbors ; By different search radius Value, divide the entire sea surface point cloud into multiple point sets of different scales , through neighborhood point sets of different scales , can capture large-scale and small-scale wave features respectively, and improve the comprehensiveness and accuracy of normal vector calculation. Specifically, a larger search radius can capture large-scale wave features with longer wavelengths, larger amplitudes, and smoother wave surfaces, while a smaller search radius can more accurately capture small-scale wave features with shorter wavelengths, smaller amplitudes, and more complex wave surfaces.

[0092] S302: Multi-scale irregular TIN triangulated sea surface model construction and normal vector calculation fusion. At each scale, each neighborhood point set obtained by screening is Delaunay triangulated to construct a triangulated network. For each triangle , calculate the normal vector , such as the innovative formula proposed by the present invention:

[0093] (5);

[0094] In the formula, are the three vertices of the triangle;

[0095] In order to integrate the normal vector information at different scales, the weighted average method is used to fuse the normal vectors of each scale. The weighting is performed, such as the formula innovatively proposed by the present invention:

[0096] (6);

[0097] In the formula, is the normal vector The weight coefficient of For the The total number of triangles under scale;

[0098] Calculate the weighted average of the normal vectors of the triangles where the sea surface points at each scale are located , as the calculation formula innovatively proposed by the present invention is:

[0099] (7);

[0100] In the formula, is the number of triangles.

[0101] S303, the refraction vector is calculated according to Snell's law, as shown in formula (8):

[0102] (8);

[0103] In the formula, is the refraction vector, is the incident vector, is the refractive index ratio, is the laser incident angle, is the laser refraction angle;

[0104] The corresponding seabed point coordinates are calculated based on the sea surface point coordinates, the refraction vector and the corresponding underwater slope distance, as shown in formula (9):

[0105] (9);

[0106] In the formula, is the coordinate of the laser seabed point in the WGS-84 coordinate system, are the refraction vector components in the X, Y, and Z directions, is the coordinate of the laser sea surface point in the WGS-84 coordinate system, is the laser underwater slant range.

[0107] Another example, Figure 2 The embodiment of the present invention provides an airborne LiDAR bathymetric wave correction method principle.

[0108] It can be seen from the above embodiments that the method of the present invention can efficiently complete wave correction without constructing a wave model based on the wave spectrum, and can ensure accuracy while being highly efficient. It can reduce the wave error caused by the deflection of the laser path caused by sea surface waves when the laser is at the sea-air interface. It can not only provide technical support for the refined processing of airborne LiDAR bathymetric data, but also provide high-precision data for underwater topographic measurements in coastal areas and other regions, and promote research and application development in the fields of marine science and marine surveying and mapping.

[0109] Embodiment 2: An airborne LiDAR bathymetric wave correction system provided by an embodiment of the present invention comprises:

[0110] The wave parameter calculation module is used to calculate the wave parameters according to the Elfouhaily spectrum model; the wave parameters include wave width and wave height, which lays the foundation for the subsequent multi-scale irregular TIN (Triangulated Irregular Network, TIN) triangulated network array sea surface model modeling;

[0111] The sea surface point search module in the neighborhood is used to use the K-nearest neighbor search method (K-NN algorithm) accelerated by KD-Tree to determine the search neighborhood with wave parameters to realize the search of sea surface points in the neighborhood;

[0112] The wave error correction module is used to construct a multi-scale irregular TIN triangulated network array sea surface model based on the efficient search results of sea surface points in the neighborhood, calculate the normal vector of each sea surface laser point with seabed reflection echo, and perform ray tracing to correct the wave error to obtain the precise coordinates of the seabed point.

[0113] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0114] Simulation experiment: In order to verify the performance of the proposed airborne LiDAR bathymetric wave correction method, the airborne LiDAR bathymetric data along the coast of a certain area was used for verification. The ALB equipment used in the experiment is Mapper 20KU developed by a certain optical precision instrument research institute. The laser center wavelength is 532nm, the laser repetition rate is 20kHz, the scanning angle is ±20°, the laser divergence angle is 1mrad, and the weight is 6kg. The equipment integrates a company's EG370N high-precision inertial navigation, and the flight carrier is a vertical take-off and landing fixed-wing UAV RY-V50.

[0115] In order to verify the bathymetric accuracy of the ALB system after wave correction, verification data were collected along the coast of a certain area in October 2023. The test altitude was 120m, the flight speed was 22m / s, and the point cloud density was about 10 / m 2 The overlap of the navigation zone is 30%, the transparency of the saddle disc is 3.7m, and the maximum detection depth is about 7.55m. At the same time, the data collected by the ship-borne Single Beam EchoSounder (SBES) system is used as the seabed sounding accuracy verification data.

[0116] Qualitative analysis of wave correction results: Using coastal data of a certain area, the wave error of Mapper 20KU was corrected according to the proposed method. The profile of the multi-scale irregular TIN triangulated network array sea surface model is shown in Figure 4 It can be seen that the application of this method to construct TIN triangulation network for sea surface point cloud can better express the ups and downs of sea surface waves.

[0117] Depend on Figure 5 , Figure 7 It can be seen that before wave correction, there is a large elevation deviation between the ALB data and the SBES data. The distribution of green points is relatively discrete, and there are large fluctuations in some areas. The overall trend is inconsistent, indicating that the wave error has significantly interfered with the ALB data, making the bathymetric results inaccurate. Figure 6 , Figure 8 It can be seen that after wave correction, the ALB data are highly consistent with the SBES data, the elevation deviation is significantly reduced, and the distribution of blue points becomes smoother and more regular, especially in areas with large deviations before correction. The consistency of the data has been significantly improved, indicating that the wave correction method effectively reduces the interference of sea surface waves on ALB data and significantly improves the reliability and accuracy of bathymetric data, providing scientific and reliable support for the application of ALB in actual ocean measurements, and reflecting the practicality and necessity of the wave correction method.

[0118] Quantitative analysis of wave correction results:

[0119] In order to verify the accuracy of Mapper 20KU seabed point sounding after wave correction, the data collected by the more mature underwater acoustic single beam sounding system (SBES) was used as the benchmark data to conduct a comparative analysis of the seabed point sounding accuracy before and after wave correction. The distribution of points in the measurement area is as follows: Fig. 9 shown.

[0120] In order to quantitatively evaluate the accuracy of Mapper 20KU bathymetry before and after wave correction, SBES data were used as verification data. Since the point density of the ALB point cloud is different from that of the verification data, the inverse distance weighted method was used to determine the same-name points, and then the mean absolute error (MAE) and root mean square error (RMSE) were calculated. Fig.10 This is the elevation difference distribution diagram of the same-name sounding points of ALB and SBES before wave correction. Fig.11 This is the distribution diagram of the elevation difference between the ALB and SBES sounding points with the same name after wave correction. The statistical results are shown in Table 1.

[0121] Table 1 Comparison of depth measurement accuracy of Mapper 20KU before and after wave correction

[0122]

[0123] Depend on Fig.10 , Fig.11It can be seen that the error after wave correction is closer to 0 and is more concentrated in the range of ±0.15m. As shown in Table 1, the maximum height difference between the point cloud collected by Mapper 20KU and the SBES same-name point after wave correction is 28.0cm, the minimum is 0.1cm, the MAE is 10.5cm, and the RMSE is 12.8cm; the RMSE after wave correction is 4.3cm less than that before correction. The depth measurement accuracy after wave correction is better than the depth measurement accuracy standard of the "Marine Engineering Survey Code": in depth measurement, when the water depth is ≤20m, the depth measurement error is ≤0.2m. It can be seen that the wave error of the Mapper 20KU system has been basically eliminated, and it has good measurement accuracy, which further shows that the proposed method is highly applicable to reducing wave errors.

[0124] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered within the protection scope of the present invention.

Claims

1. A wave correction method for airborne LiDAR bathymetry, characterized in that: The method comprises the following steps: S1, calculate wave parameters according to the Elfouhaily spectrum model, wave parameters include wave width and wave height; S2, using the K-nearest neighbor search method accelerated by KD-Tree, the search neighborhood is determined by wave parameters to realize the search of sea surface points within the neighborhood; S3, based on the efficient search results of sea surface points in the neighborhood, construct a multi-scale irregular TIN triangulated network array sea surface model, calculate the normal vector of each sea surface laser point with seabed reflection echo, perform ray tracing to correct the wave error and obtain the precise seabed point coordinates; In step S1, wave parameters are calculated according to the Elfouhaily spectral model, including: S101, divide the entire sea surface point cloud into voxel grids, retain the highest point in each grid, collect wind speed synchronously during the experiment, and match the wind speed corresponding to each laser point according to time; S102, Calculation of wave width using the Elfouhaily wave spectrum model and wave height , the expression is: （1）； In the formula, is the Elfouhaily wave spectrum model, is the spectral amplitude coefficient, is the acceleration due to gravity, is the wave number, is an empirical parameter, is the spectrum peak frequency, is the direction distribution function, is the maximum wave number, is the minimum wave number, is the differential of the direction angle, To differentiate the beam, is the direction angle, is the wind speed.

2. The airborne LiDAR bathymetric wave correction method according to claim 1, characterized in that: In step S2, the K-nearest neighbor search method accelerated by KD-Tree is used to determine the search neighborhood with wave parameters to search for sea surface points in the neighborhood, including: S201, construct KD-Tree: calculate the variance, select the dimension with the largest variance as the segmentation dimension of the current layer; in the selected segmentation dimension Above, sort the point cloud data, select the median as the split point, divide the data set into two left and right subsets with the median as the boundary, recursively build the subtree, and recursively perform the above operations on the left and right subsets until the preset tree depth is reached; S202, K-nearest neighbor search: based on the wave width and wave height , adaptively determine the search radius and the number of nearest neighbors ; For query point , recursively traverse from the root node of the KD-Tree, and decide whether to enter the left subtree or the right subtree by comparing the coordinates of the query point on the current segmentation dimension with the segmentation value of the node. At each node, calculate the Euclidean distance between the query point and the current node, and update the nearest neighbor set. If the number of points in the current set is less than , then add directly, has reached points, replace the farthest point and update the maximum distance ; During recursive backtracking, determine whether the distance from the splitting hyperplane to the query point is less than the current maximum distance This judgment is used to decide whether to continue searching the other side subtree. If the distance from the hyperplane to the query point is greater than , then prune and no longer search the subtree on the other side, thus avoiding unnecessary calculations; if the distance is less than or equal to , then continue to recursively search the other side subtree to check if there is a closer point.

3. The airborne LiDAR bathymetric wave correction method according to claim 2, characterized in that: In step S201, the dimension is segmented The calculation formula is: （2）； In the formula, For the dimensional variance, For all points The total number of are three spatial coordinate dimensions, For the dimension, For the The point is The value on the dimension, For the The mean of the dimension; The data set is divided into two subsets, left and right, and the calculation formula for recursively constructing the subtree is: （3）； In the formula, is the left subset, is the right subset, Summarize the dataset Points, is a set of data points, For data points In the split dimension The value on Split dimensions into a dataset The median value of all points on the .

4. The airborne LiDAR bathymetric wave correction method according to claim 2, characterized in that: In step S202, the search radius is adaptively determined and the number of nearest neighbors The calculation formula is: （4）； In the formula, All are adjustment factors.

5. The airborne LiDAR bathymetric wave correction method according to claim 1, characterized in that: In step S3, before constructing the multi-scale irregular TIN triangulated network array sea surface model, multi-scale neighborhood point screening is performed, and the K-nearest neighbor search method accelerated by KD-Tree is used to select the nearest neighbor points according to each query point. Wave width and wave height , adaptively determine the search radius and the number of nearest neighbors ; By different search radius Value, divide the entire sea surface point cloud into multiple point sets of different scales .

6. The airborne LiDAR bathymetric wave correction method according to claim 5, characterized in that: In step S3, the normal vector of each sea surface laser point with seabed reflection echo is calculated, including: at each scale, Delaunay triangulation is performed on each neighborhood point set obtained by screening to construct a triangulated network. For each triangle , calculate the normal vector , the calculation formula is: （5）； In the formula, are the three vertices of the triangle; The normal vector information at different scales is integrated, and the weighted average method is used to fuse the normal vectors of each scale. The calculation formula for weighting is: （6）； In the formula, is the normal vector The weight coefficient of For the The total number of triangles at each scale; Calculate the weighted average of the normal vectors of the triangles where the sea surface points are located at each scale , the calculation formula is: （7）； In the formula, is the number of triangles.

7. The airborne LiDAR bathymetric wave correction method according to claim 6, characterized in that: In step S3, ray tracing is performed to correct the wave error to obtain the accurate coordinates of the seabed point, including: calculating the refraction vector according to Snell's law, and the calculation formula is: （8）； In the formula, is the refraction vector, is the incident vector, is the refractive index ratio, is the laser incident angle, is the laser refraction angle; The corresponding seabed point coordinates are calculated based on the sea surface point coordinates, the refraction vector and the corresponding underwater slope distance. The calculation formula is: （9）； In the formula, is the coordinate of the laser seabed point in the WGS-84 coordinate system, for X , Y , Z The components of the refraction vector in three directions, is the coordinate of the laser sea surface point in the WGS-84 coordinate system, is the laser underwater slant range.

8. An airborne LiDAR bathymetric wave correction system, characterized in that: The system implements the airborne LiDAR bathymetric wave correction method as claimed in any one of claims 1 to 7, and the system comprises: The wave parameter calculation module is used to calculate the wave parameters according to the Elfouhaily spectrum model. The wave parameters include wave width and wave height. The sea surface point search module in the neighborhood is used to use the K-nearest neighbor search method accelerated by KD-Tree to determine the search neighborhood with wave parameters to realize the search of sea surface points in the neighborhood; The wave error correction module is used to construct a multi-scale irregular TIN triangulated network array sea surface model based on the efficient search results of sea surface points in the neighborhood, calculate the normal vector of each sea surface laser point with seabed reflection echo, and perform ray tracing to correct the wave error to obtain the precise coordinates of the seabed point.

9. The airborne LiDAR bathymetric wave correction system according to claim 8, characterized in that: The airborne LiDAR bathymetric wave correction system is mounted on a computer-readable storage medium, and the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the functions of the airborne LiDAR bathymetric wave correction system can be realized.

Images