River underwater terrain reconstruction method based on single-beam data

Abstract

The invention discloses a river underwater terrain reconstruction method based on single-beam data. The method comprises the following steps: acquiring a river vector surface, a center line and section data of a target river; fitting a section and dividing section points to determine an elevation; discrete center points are divided on a river channel center line between adjacent sections, encrypted feature lines perpendicular to tangent lines are made through the center points, and feature points are divided. For each feature point, extracting a candidate section and determining a weight; selecting a corresponding section point and neighborhood points on the two sides of the section point from each candidate section to generate a neighborhood point set, determining the weight of the neighborhood points, calculating the weighted elevation of the candidate section according to the weight of the neighborhood points, and obtaining the comprehensive weighted elevation by combining the weight of the candidate section. Determining a dynamic slope adjustment coefficient according to the elevation change rate of the neighborhood point, determining a width adjustment coefficient based on the ratio of the river reach width to the average width of the whole river reach, comprehensively determining the elevation of the feature points, and constructing the river underwater topography in combination with the elevation of all the feature points and the original section data. According to the invention, the underwater terrain of the river can be accurately reconstructed.

Description

Technical Field

[0001] The present invention belongs to the technical field of underwater terrain modeling of rivers, and particularly relates to a method for reconstructing underwater terrain of rivers based on single-beam data. Background Art

[0002] Underwater terrain measurement of river channels is an important basis for hydrological engineering, waterway regulation, and ecological restoration. Currently, it mainly relies on technologies such as single-beam sounding, multi-beam sounding, lidar (LiDAR), and remote sensing inversion. Among them, single-beam sounding has become the main means for measuring small and medium-sized rivers due to its low cost and convenient operation. However, its data sparsity poses significant challenges to high-precision terrain reconstruction. The difficulties in reconstructing underwater terrain of rivers based on single-beam data are concentrated in three aspects: First, data sparsity and interpolation distortion. Single-beam data is only distributed along the cross-section, and there is no data between cross-sections. Traditional interpolation methods (such as inverse distance weighting or Kriging) are prone to smoothing distortion or false undulations in complex riverbeds (such as deep pools and sand waves), making it difficult to accurately reflect real terrain mutations. Second, insufficient expression of morphological features. Affected by water flow dynamics, rivers often form non-linear features such as steep slopes and deep pools. However, existing methods lack dynamic coupling of morphological parameters such as river width and slope, resulting in the loss of key geomorphic information. Third, single-beam data is easily interfered by aquatic plants, suspended matter, or bubbles, generating abnormal water depth points, and the noise interference is significant.

[0003] Existing solutions mostly rely on spatial interpolation with fixed parameters (such as IDW) or manual correction, and there are obvious limitations: Traditional interpolation methods only consider distance attenuation and ignore the dynamic influence of river channel morphology. At the same time, traditional interpolation methods usually lack self-adaptive ability and cannot take into account the different requirements of different river reaches (such as narrow river channels, wide and shallow river channels, and curved river channels). The above limitations lead to the reconstructed terrain being difficult to accurately reflect the actual geomorphic features, affecting the decision-making accuracy of hydrological engineering, waterway regulation, and ecological restoration.

[0004] Therefore, it is urgent to explore more efficient interpolation methods and data processing technologies to improve the application effect of single-beam data in underwater terrain reconstruction of river channels. Summary of the Invention

[0005] The technical objective of this application is to provide a method for reconstructing underwater terrain of rivers based on single-beam data to address the technical problem that the interpolation method adopted in the current underwater terrain modeling technology of rivers based on single-beam data lacks self-adaptive ability and cannot take into account the different requirements of different river reaches, thus affecting the reconstruction accuracy.

[0006] To achieve the above technical objective, the following technical solutions are adopted in this application.

[0007] An embodiment of this application provides a method for reconstructing underwater terrain of rivers based on single-beam data, including:

[0008] Obtain the vector surface data of the target river channel, the center line data of the river channel, and the river channel cross-section data based on single-beam measurement;

[0009] According to the river channel cross-section data, fit the cross-section and divide the cross-section points to determine the elevation of each cross-section point;

[0010] According to the vector surface data of the river channel, discrete center points are drawn on the center line of the river channel between two adjacent cross-sections, and then straight lines are drawn through these center points along the direction perpendicular to the tangent of the center line of the river channel at the center point to obtain encrypted feature lines, and the feature points are divided for the encrypted feature lines;

[0011] For each of the feature points, the following steps are executed to determine the elevation of the feature point: extract the upstream and downstream set number of cross-sections as candidate cross-sections according to the spatial position of the feature point, and determine the weights of each candidate cross-section; for each candidate cross-section, select the cross-section point corresponding to the feature point in the candidate cross-section, and the cross-section points within the neighborhood radius on both sides of it as neighborhood points to generate the neighborhood point set of the candidate cross-section, and determine the weights of each neighborhood point in the neighborhood point set; according to the elevations of all neighborhood points in the neighborhood point set and the neighborhood point weights, determine the weighted elevation of the candidate cross-section; according to all candidate cross-section weights and the weighted elevations of all candidate cross-sections, obtain the comprehensive weighted elevation of the feature point; according to the elevation change rate of all neighborhood points in all neighborhood point sets, determine the dynamic slope adjustment coefficient; determine the width adjustment coefficient based on the ratio of the width of the river section where the feature point is located to the average width of the entire river section; according to the comprehensive weighted elevation, dynamic slope adjustment coefficient and width adjustment coefficient of the feature point, comprehensively determine the elevation of the feature point;

[0012] According to the elevations of all feature points, combined with the original river channel cross-section data, construct the underwater terrain of the river.

[0013] Furthermore, according to the river channel cross-section data, fitting the cross-section and dividing the cross-section points includes:

[0014] Use NURBS to fit the cross-section and set the curve order; determine the number of control points according to the complexity of the underwater terrain of the river, and after initializing the weights of the control points to 1, optimize the positions and weights of the control points by the least squares method to minimize the mean square error between the fitted cross-section and the original cross-section points; then parameterize the fitted curve and equally divide the parameter interval to obtain cross-section points.

[0015] Furthermore, the formula for determining the weight of the m-th candidate cross-section is as follows:

[0016] ;

[0017] Wherein, , f is the total number of cross-sections, is the weight of the m-th candidate section, λ represents the distance attenuation coefficient, and d m represents the length of the centerline of the river channel between the encrypted feature line where the feature point is located and the m-th candidate section.

[0018] Furthermore, the calculation expression of the neighborhood point weight is as follows:

[0019] ;

[0020] where is the weight of the p-th neighborhood point c m,p on the m-th candidate section, and k is the serial number of the feature point on the encrypted feature line where it is located.

[0021] Furthermore, the weighted elevation h m,k on the m-th candidate section is jointly determined by the elevations of all neighborhood points in the neighborhood point set and the neighborhood point weights, and the calculation formula is as follows:

[0022] ;

[0023] where h m,p represents the elevation of the p-th neighborhood point c m,p on the m-th candidate section, is the weight of the p-th neighborhood point c m,p on the m-th candidate section, and k is the serial number of the feature point on the encrypted feature line where it is located.

[0024] Furthermore, the expression for determining the comprehensive weighted elevation of the feature point is as follows:

[0025] ;

[0026] where f is the total number of sections, is the weight of the m-th candidate section, h m,k represents the weighted elevation of the m-th candidate section, and h k represents the comprehensive weighted elevation of the feature point with serial number k.

[0027] Furthermore, according to the elevation change rates of all neighborhood points in all neighborhood point sets, a dynamic slope adjustment coefficient is determined, including: determining the average elevation difference Δh of all neighborhood points in the neighborhood point sets of the feature point on all candidate sections, and the expression is as follows:

[0028] ;

[0029] where σ h is the standard deviation of the water depth values of the neighborhood points of the feature point in all neighborhood point sets; L represents the average water depth of the neighborhood points of the feature point in all neighborhood point sets;

[0030] Calculating the dynamic slope adjustment coefficient through the tanh function , and the expression is as follows:

[0031] ;

[0032] Among them, S represents the slope between the first candidate section and the encrypted feature line where the feature point is located; S0 represents the slope threshold; is a parameter to prevent division by zero; β is the maximum adjustment amplitude.

[0033] Furthermore, determine the elevation H of the feature point, and the formula is as follows:

[0034] ;

[0035] Among them, h k represents the comprehensive weighted elevation of the feature point with the serial number k, and α w (W) is the width adjustment coefficient, is the dynamic slope adjustment coefficient.

[0036] Furthermore, the value of the curve order is 3; for the navigable river with regular underwater terrain, the number of control points is 3 - 4, and for the natural river with complex underwater terrain, the number of control points is 5 - 8.

[0037] Furthermore, determine the width adjustment coefficient based on the ratio of the width of the river section where the feature point is located to the average width of the whole river section, and the expression is as follows:

[0038] ;

[0039] Among them, α w (W) is the width adjustment coefficient, W is the width of the river section where the encrypted feature line where the feature point is located, W0 is the average width of the whole river section, and γ represents the power index parameter.

[0040] Compared with the prior art, the beneficial technical effects achieved by the embodiments of the present application are as follows: Obtain the river channel vector surface data, center line data and cross-section data based on single-beam measurement of the target river. The multiple types of data complement each other and describe the river characteristics from different dimensions. It can flexibly adjust the curve shape according to the complexity of the underwater terrain of the river, making the fitted cross-section more conform to the actual terrain. The fused comprehensive weighted elevation can scientifically and reasonably reflect the influence degree of various factors on the target point, providing a basis for accurately calculating the elevation of the target point. Further, the dynamic slope adjustment coefficient can reflect the change of the terrain slope, and the width adjustment coefficient takes into account the difference in river morphology. The combination of the two makes the calculation of the target point elevation more in line with the actual terrain situation, and the present application can accurately reconstruct the underwater terrain of the river. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure of the present application in any way. Additionally, the shapes, proportional dimensions, etc. of the components in the drawings are only schematic and are used to assist in understanding the present application, rather than specifically defining the shapes and proportional dimensions of the components of the present application. Those skilled in the art can, under the teaching of the present application, select various possible shapes and proportional dimensions according to specific circumstances to implement the present application. In the accompanying drawings:

[0042] Figure 1 Schematic flow diagram of the river underwater terrain reconstruction method based on single-beam data provided by an embodiment of the present application;

[0043] Figure 2 Schematic diagram of the measurement section and encrypted feature lines in an embodiment of the present application;

[0044] Figure 3 Schematic diagram of the neighborhood point set of the second feature point (k = 2) when the number of candidate sections is 4 and the neighborhood radius is 1 in an embodiment of the present application;

[0045] Figure 4 Schematic diagram of the cross-section weight calculation of the second feature point (k = 2) when the number of cross-sections is 4 and the neighborhood radius is 1 in an embodiment of the present application. Detailed implementation manner

[0046] In order to enable those skilled in the art to better understand the technical solutions in the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present application.

[0047] An embodiment of the present application provides a river underwater terrain reconstruction method based on single-beam data, as Figure 1 shown, including the following steps:

[0048] Step 1: Obtain the target river channel vector surface data, river channel centerline data, and river channel cross-section data based on single-beam measurement;

[0049] Step 2 includes:

[0050] Step 2-1: According to the river channel cross-section data, fit the cross-section and divide the cross-section points to determine the elevation of each cross-section point;

[0051] Step 2-2: According to the river channel vector surface data, discrete center points are marked on the center line of the river channel between two adjacent cross-sections. Then, straight lines are drawn through these center points along the direction perpendicular to the tangent of the center line of the river channel at the center point to obtain encrypted feature lines, and the encrypted feature lines are divided into feature points;

[0052] Step 3 includes: for each feature point, perform the following steps 3-1 to step 3-6 to determine the elevation of the feature point:

[0053] Step 3-1: Extract a set number of upstream and downstream cross-sections as candidate cross-sections according to the spatial position of the feature point, and determine the weights of each candidate cross-section;

[0054] Step 3-2: For each candidate cross-section, select the cross-section point corresponding to the feature point on the candidate cross-section, and the cross-section points within the neighborhood radius on both sides of it as neighborhood points to generate a neighborhood point set, and determine the weights of each neighborhood point in the neighborhood point set;

[0055] Step 3-3: Determine the weighted elevation of the candidate cross-section according to the elevations and weights of all neighborhood points in the neighborhood point set; obtain the comprehensive weighted elevation of the feature point according to the weights of all candidate cross-sections and the weighted elevations of all candidate cross-sections;

[0056] Step 3-4: Determine the dynamic slope adjustment coefficient according to the elevation change rates of all neighborhood points in all neighborhood point sets;

[0057] Step 3-5: Determine the width adjustment coefficient based on the ratio of the width of the river section where the feature point is located to the average width of the entire river section;

[0058] Step 3-6: Comprehensively determine the elevation of the feature point according to the comprehensive weighted elevation, dynamic slope adjustment coefficient, and width adjustment coefficient of the feature point;

[0059] Step 4: Construct the underwater terrain of the river according to the elevations of all feature points and in combination with the original river channel cross-section data.

[0060] In some embodiments, the cross-section points, center points, and feature points are equally spaced. Discrete center points are marked on the river centerline according to the equal-spacing principle, and then a densified feature line is obtained and feature points are marked, increasing the density and details of the terrain data. The equal-spacing division makes the data acquisition process have clear rules and standards, facilitating operation and implementation. In actual measurement and data processing, dividing at a fixed interval can reduce human errors, improve the efficiency and accuracy of data acquisition. At the same time, the data distribution with equal intervals is also conducive to subsequent calculations and analyses, making the algorithm easier to implement and reducing the complexity of data processing. The equal-spacing division can also ensure relatively uniform sampling of terrain features throughout the river area. Whether in the straight section or the curved section of the river, terrain information can be obtained at the same density, avoiding the situation where terrain features in some areas are over-sampled or missed due to uneven sampling intervals.

[0061] In some embodiments, before step 1, the vector surface data of the river channel, the centerline data of the river channel, and the cross-section data of the river channel measured by a single-beam (a total of s cross-sections) in the basic investigation of water resources can be collected to complete the preprocessing of the cross-section data.

[0062] The single-beam measurement can be shipborne single-beam measurement, portable single-beam measurement, airborne single-beam measurement, or single-beam measurement carried by an underwater robot. For example, shipborne single-beam measurement is a measurement technology using sound waves as the propagation medium. By emitting sound waves towards the bottom of the water and receiving the reflected sound waves on the ship, calculating the propagation time and speed of the sound waves, the elevation, water depth, and material distribution of the underwater terrain, etc., can be determined. This technology is mainly applied to fields such as underwater terrain measurement and marine resource exploration.

[0063] In some embodiments, the preprocessing of the cross-section data includes operations such as coordinate transformation, noise filtering, and outlier removal to ensure the accuracy and consistency of the data. As an example, the noise filtering of the cross-section data can adopt the sliding window standard deviation method, with a window size of 7 for filtering, and data points deviating from the mean by ±3 times the standard deviation are removed.

[0064] In some embodiments, step 2-1 may include: First, each cross-section is fitted using NURBS to determine key parameters such as the curve order, the number of control points, and the weights of the control points. The positions and weights of the control points are optimized by the least squares method to minimize the mean square error between the fitted cross-section curve and the original cross-section points.

[0065] As an example, the key parameters are set as follows: The curve order determines the smoothness and flexibility of the curve and is generally set to 3. The number of control points is set according to the complexity of the underwater terrain of the river. For a regular waterway river with an underwater terrain, it is recommended to take 3 - 4 control points. For a natural river with a complex underwater terrain, it is recommended to take 5 - 8 control points. The more control points there are, the greater the degree of freedom of the curve, and the more precisely it can express the underwater terrain. The weight of each control point is initially set to 1 and is adjusted later according to the fitting accuracy.

[0066] As an example, based on the NURBS definition, a curve equation is constructed:

[0067] ;

[0068] where C(u) represents the coordinates of the cross-sectional curve at parameter u; N i,k (u) is the i-th k-th order B-spline basis function; w i represents the weight of the i-th control point; P i represents the coordinate vector of the i-th control point, and z is the total number of control points.

[0069] Furthermore, the fitted cross-sectional curve is parameterized to generate a parameter interval [u min , u max . The parameter interval is equally divided into n cross-sectional points, and each parameter value is calculated: , . Substitute the parameter values into the NURBS curve equation to obtain the corresponding cross-sectional point coordinates (x i , y i ).

[0070] In some embodiments, step 2-2 may specifically include: Refining the underwater terrain features between cross-sections by drawing encrypted feature lines. According to the accuracy requirements, the center line of the river between two adjacent cross-sections can be divided into evenly distributed center points according to the equal-spacing principle to capture the morphological changes of the river. Subsequently, for each discrete center point, a perpendicular line is drawn along the direction perpendicular to the tangent of the river center line at that point (i.e., the vertical direction of the river flow direction) to reflect the width change and local morphological features of the river in the lateral dimension. When the perpendicular line intersects the water surface boundary of the entire river (i.e., the river surface), the formed intersection segment is the so-called encrypted feature line. In the same way as the cross-sectional point division method, the encrypted feature line is equally divided into n feature points. The number of points on each cross-section and the encrypted feature line is the same (n), and the serial numbers of the cross-sectional points and the feature points correspond one by one to facilitate the subsequent generation of the neighborhood point set of the feature points.

[0071] In some embodiments, step 3-1 may specifically include: sequentially determining the candidate cross-sections and candidate cross-section weights of each feature point. According to the spatial position of the feature point, f cross-sections with a specific number upstream and downstream of it are extracted as candidate cross-sections to ensure the spatial correlation and data integrity between the cross-section and the feature point, and to reflect the influence of the underwater terrain of the river along the water flow direction on the elevation to be interpolated.

[0072] The exponential function has the characteristics of a sharp decline in weight within a short distance and a gentle change after a long distance. According to the spatial relationship between the candidate cross-section and the feature point, the cross-section weight is constructed using the fast initial decay and long-tail effect of the exponential function to reasonably distribute the weight, so as to reflect the contribution degree of the candidate cross-section to the parameter estimation of the feature point. The candidate cross-section weight is determined by distance and has globality.

[0073] In this embodiment, a total of f cross-sections upstream and downstream of the encrypted feature line where the feature point is located are selected as candidate cross-sections, which can be referred to Figure 2 as shown.

[0074] In some embodiments, for the mth ( ) candidate cross-section weight the calculation formula is as follows:

[0075] ;

[0076] where λ represents the distance attenuation coefficient, which controls the rate of weight decline with distance. The larger λ is, the more obvious the advantage of the proximal cross-section is. λ is adjusted according to the cross-section density. For dense cross-sections (cross-section spacing < 50m), λ can take 0.02 - 0.03, and for sparse cross-sections (spacing > 100m), λ can take 0.01 - 0.02; d m represents the length of the river centerline between the encrypted feature curve and the mth candidate cross-section, which can be seen Figure 4 as shown.

[0077] In some embodiments, for the encrypted feature line located between the 1st and f / 2th cross-sections, or the encrypted feature line between the s - f / 2th and sth cross-sections, that is, the encrypted feature lines at both ends of the river, the candidate cross-sections are the cross-sections that can be searched within the range of f / 2 cross-sections before and after the encrypted feature line.

[0078] Example: If f = 4, for the encrypted feature line located between the 1st and 2nd cross-sections, within the range of 4 / 2, the 1st, 2nd, and 3rd cross-sections can be searched as candidate cross-sections.

[0079] In some embodiments, step 3-2 may specifically include: determining the neighborhood point set of the feature point on each candidate cross-section and the neighborhood point weights of each neighborhood point, and quantifying the influence of cross-section points at different positions within the same cross-section on the feature point through the neighborhood point weights. The riverbed elevation usually has spatial continuity, and the influence of single-point measurement error is smoothed through weighted averaging of multiple neighborhood points.

[0080] Each cross-section and the encrypted feature line have n points. Through step 3-1, f candidate cross-sections upstream and downstream of each feature point are determined, and the neighborhood point set of each feature point on each candidate cross-section is determined in sequence.

[0081] Assume that the point number of the currently interpolated feature point is k (k ∈ (1, 2,..., n)), and r is the neighborhood radius. The neighborhood radius is determined according to the complexity of the river. For natural rivers with many bends and bifurcations, the neighborhood radius is taken as 3-5, and for straight rivers, it is taken as 1-2. Select the k-th cross-section point and the r points on both sides of it in the m-th candidate cross-section (abbreviated as "candidate cross-section m") (m ∈ (1, 2,..., f)) to generate the neighborhood point set N(m, k). The neighborhood point set includes the cross-section points with serial numbers from k-r to k+r, that is .

[0082] In some embodiments, if the feature point number k ∈ (1, 2,..., r) or k ∈ (n-r, n-r+1,..., n), that is, the feature points within the neighborhood radius at both ends of the encrypted feature line, the relationship between the feature point and the cross-section point is one-to-one, and the neighborhood point set in the candidate cross-section m is { }.

[0083] Example of neighborhood point set selection 1: For a feature point with serial number k = 6 on a certain encrypted feature line, the neighborhood radius r = 3, and 4 cross-sections upstream and downstream have been determined as candidate cross-sections. The neighborhood point set of this feature point on the candidate cross-section 1 (m = 1) is N(1, 6), including the cross-section points with serial numbers from 3 to 9 on the candidate cross-section 1, that is:

[0084] .

[0085] Example of neighborhood point set selection 2: For a feature point with serial number k = 2 on a certain encrypted feature line, the neighborhood radius r = 3, and k = 2 ∈ (1, 2, 3). This feature point is within the neighborhood radius, and 4 cross-sections upstream and downstream have been determined as candidate cross-sections. The neighborhood point set of this feature point on the candidate cross-section 1 (m = 1) is N(1, 2), including the cross-section point with serial number 2 on the candidate cross-section 1, that is .

[0086] Example of neighborhood point set selection 3: As Figure 3As shown in the figure, for the feature point with serial number k = 2 on a certain encryption feature line, the neighborhood radius r = 1, k = 2, and this feature point is not within the neighborhood radius. Four cross-sections upstream and downstream have been determined as candidate cross-sections. The neighborhood point set of this feature point on candidate cross-section 1 (m = 1) is N(1, 2), including the cross-section points with serial numbers 1 and 2 on candidate cross-section 1, that is, {c1,1, c1,2, c1,3}.

[0087] The weights of neighborhood points are determined by the positions within the neighborhood, having locality, and are used to reflect the local continuity of cross-section data. For the neighborhood point c m,p on candidate cross-section m at the p-th point, its elevation value is h m,p and its neighborhood point weight is calculated as follows:

[0088] ;

[0089] where p represents the serial number of the neighborhood point within the neighborhood point set on candidate cross-section m, and the range is [k - r, k + r].

[0090] Then the elevation value h m,k of candidate cross-section m is jointly determined by the elevations of all neighborhood points within the neighborhood point set and the neighborhood point weights, and the calculation formula is as follows:

[0091] ;

[0092] where h m,p represents the elevation of neighborhood point c m,p , and h m,k represents the elevation of candidate cross-section m after weighting. For each encryption feature line, there are n feature points. Each feature point corresponds to f candidate cross-sections and f cross-section weights, generating f neighborhood point sets. The neighborhood points within each neighborhood point set have neighborhood point weights, and the comprehensive weighted elevation of this feature point is calculated. The comprehensive weighted elevations of n feature points on each encryption feature line are calculated in sequence, more detailedly reflecting the undulation of the underwater terrain.

[0093] In some embodiments, step 3 - 3 may specifically include: calculating the comprehensive weighted elevation of the feature point by fusing the above-mentioned cross-section weights and neighborhood point weights, and this process takes into account the influence of the water flow movement along the water flow direction and perpendicular to the riverbank direction on the underwater terrain simultaneously.

[0094] .

[0095] In some embodiments, step 3-4 includes: The dynamic slope adjustment coefficient reflects the elevation change rate of the river along the water flow direction, is directly related to the scouring-silting dynamics process of the water flow, and conforms to the mechanism better than simple geometric indicators. First, calculate the average elevation difference Δh of all neighboring points within the neighboring point set of the feature point on all candidate cross-sections, which reflects the undulation degree of the riverbed in the direction perpendicular to the riverbank. Δh identifies complex terrain areas and adjusts the model sensitivity, reducing manual parameter debugging. Then, calculate the dynamic slope adjustment coefficient through the tanh function .

[0096] When Δh is large, indicating a sudden change in terrain, the denominator is increased to suppress the change and weaken the slope weight; when Δh is small, indicating a gentle terrain, the influence of the slope on interpolation is enhanced. The calculation formulas for Δh and are as follows:

[0097] ;

[0098] ;

[0099] where σ h represents the standard deviation of the water depth values within the neighborhood of the feature point; represents the average water depth of the neighborhood; S represents the slope between the first candidate cross-section and the encrypted feature line where the feature point is located; S0 represents the slope threshold, which is determined by the average slope of the entire river section; is a minimum value (such as 1e-6) to prevent division by zero; β is the maximum adjustment amplitude.

[0100] In some embodiments, step 3-5 specifically includes using the width ratio (W / W0) to reflect the characteristics of the lateral constraint of the river channel, and calculating the width adjustment coefficient α w (W), where γ represents the power exponent parameter, W represents the width of the river section where the encrypted feature line where the feature point is located, and W0 represents the average width of the entire river section.

[0101] .

[0102] In some embodiments, step 3-6 specifically includes finally determining the elevation of the feature point according to the comprehensive weighted elevation, the dynamic slope adjustment coefficient, and the width adjustment coefficient , and the formula is as follows:

[0103] ;

[0104] That is: .

[0105] Step 4: Combine the cross-section point data and the encrypted feature point data to obtain gridded and refined elevation points, which serve as the input data for underwater terrain modeling. In ARCGIS, based on the Triangulated Irregular Network (TIN) method, use the "Create TIN" tool to generate the TIN of the river underwater terrain, and use the "TIN to Raster" tool to set the sampling interval parameter (such as 10 meters) for refined modeling of the river underwater terrain.

[0106] The above has introduced in detail the method for reconstructing the river underwater terrain based on single-beam data. In this article, specific examples are used to elaborate on the principle and implementation manner of this application. The description of the above embodiments is only for helping to understand the concept of this application and should not be construed as a limitation on the protection scope of this application.

Claims

1. A method for reconstructing underwater riverbed topography based on single-beam data, characterized in that, Including: Obtaining the vector surface data of the target river channel, the center line data of the channel, and the channel cross-section data based on single-beam measurement; Fitting the cross-section according to the channel cross-section data, dividing cross-section points, and determining the elevation of each cross-section point; According to the vector surface data of the channel, discrete center points are drawn on the center line of the channel between two adjacent cross-sections, and then straight lines are drawn through these center points along the direction perpendicular to the tangent of the center line of the channel at the center point to obtain encrypted feature lines, and the feature points are divided for the encrypted feature lines; For each of the feature points, the following steps are executed to determine the elevation of the feature point: extracting a set number of upstream and downstream cross-sections as candidate cross-sections according to the spatial position of the feature point, and determining the weights of each candidate cross-section; for each candidate cross-section, selecting the cross-section point corresponding to the feature point on the candidate cross-section and the cross-section points within the neighborhood radius on both sides thereof as neighborhood points to generate a neighborhood point set of the candidate cross-section, and determining the weights of each neighborhood point in the neighborhood point set; Determining the weighted elevation of the candidate cross-section according to the elevations of all neighborhood points in the neighborhood point set and the neighborhood point weights; obtaining the comprehensive weighted elevation of the feature point according to all candidate cross-section weights and the weighted elevations of all candidate cross-sections; determining the dynamic slope adjustment coefficient according to the elevation change rate of all neighborhood points in all neighborhood point sets; determining the width adjustment coefficient based on the ratio of the width of the river section where the feature point is located to the average width of the entire river section; comprehensively determining the elevation of the feature point according to the comprehensive weighted elevation, dynamic slope adjustment coefficient, and width adjustment coefficient of the feature point; Constructing the underwater topography of the river according to the elevations of all feature points and combining with the original channel cross-section data.

2. The method for reconstructing the underwater topography of a river based on single-beam data according to claim 1, wherein Fitting the cross-section according to the channel cross-section data and dividing cross-section points, including: Using NURBS to fit the cross-section and setting the curve order; determining the number of control points according to the complexity of the underwater topography of the river, and after initializing the weights of the control points to 1, optimizing the positions and weights of the control points by the least squares method to minimize the mean square error between the fitted cross-section and the original cross-section points; then parameterizing the fitted curve and equally dividing the parameter interval to obtain cross-section points.

3. The method for reconstructing underwater riverbed topography based on single-beam data according to claim 1, wherein The formula for determining the weight of the m-th candidate cross-section is as follows: ； Among them, , where f is the total number of cross-sections, is the weight of the m-th candidate cross-section, λ represents the distance attenuation coefficient, and d m represents the length of the center line of the river channel between the encrypted feature line where the feature point is located and the m-th candidate cross-section.

4. The method for reconstructing the underwater topography of a river based on single-beam data according to claim 1, wherein The calculation expression for the neighborhood point weight is as follows: ； wherein is the weight of the p-th neighborhood point c on the m-th candidate section m,p , and k is the serial number of the feature point on the encrypted feature line where it is located.

5. The method for reconstructing underwater river topography based on single-beam data according to claim 1, wherein The elevation h of the m-th candidate section after weighting m,k The calculation formula is as follows: ； Among them, h m,p represents the elevation of the p-th neighborhood point c m,p on the m-th candidate cross-section, and is the weight of the p-th neighborhood point c m,p on the m-th candidate cross-section. k is the serial number of the feature point on its encrypted feature line.

6. The method for reconstructing underwater riverbed topography based on single-beam data according to claim 1, wherein The expression for determining the comprehensive weighted elevation of the feature point is as follows: ； where f is the total number of cross-sections, is the weight of the m-th candidate cross-section, h m,k represents the elevation after weighting of the m-th candidate cross-section, h k represents the comprehensive weighted elevation of the feature point with the serial number k.

7. The method for reconstructing underwater river topography based on single-beam data according to claim 1, characterized in that Determining the dynamic slope adjustment coefficient according to the elevation change rate of all neighborhood points in all neighborhood point sets, including: Determining the average elevation difference Δh of all neighborhood points in the neighborhood point set of the feature point on all candidate cross-sections, and the expression is as follows: ； where, σ h is the standard deviation of the water depths of the neighborhood points within all neighborhood point sets of the feature point; L represents the average water depth of the neighborhood points within all neighborhood point sets of the feature point; Calculate the dynamic slope adjustment coefficient through the tanh function , and the expression is as follows: ； Where S represents the slope of the first candidate section and the encrypted feature line where the feature point is located; S0 represents the slope threshold; is a parameter to prevent division by zero; β is the maximum adjustment range.

8. The method for reconstructing underwater topography of a river based on single-beam data according to claim 1, wherein Determining the elevation H of the feature point, and the formula is as follows: ； Among them, h k represents the comprehensive weighted elevation of the feature point with serial number k, and α w (W) is the width adjustment coefficient, is the dynamic slope adjustment coefficient.

9. The method for reconstructing underwater river topography based on single-beam data according to claim 2, wherein The value of the curve order is 3; for the navigable river with regular underwater topography, the number of control points is 3 - 4, and for the natural river with complex underwater topography, the number of control points is 5 - 8.

10. The method for reconstructing underwater riverbed topography based on single-beam data according to claim 1, wherein Determining the width adjustment coefficient based on the ratio of the width of the river section where the feature point is located to the average width of the entire river section, and the expression is as follows: ； where α w (W) is the width adjustment coefficient, W is the width of the river reach where the encrypted feature line where the feature point is located, W0 is the average width of the entire river reach, and γ represents the power exponent parameter.

Images