High-precision complex sounding surface model construction method based on s-102 standard

Abstract

The application discloses a high-precision complex sounding surface model construction method based on an S-102 standard, relates to the field of intelligent channel sounding surface model construction, and comprises the following steps: segmenting original discrete sounding point data and extracting a target hard break line; constructing a constrained Delaunay triangulation model by using the original discrete sounding point data and the target hard break line; generating a channel regular grid DEM by using a natural neighborhood method based on the constrained Delaunay triangulation model, and calculating the uncertainty of sounding grid data by grid; and obtaining a high-precision complex sounding surface model construction result according to the channel regular grid DEM and the uncertainty of the sounding grid data. The application does not introduce additional forced conditions to the triangulation construction far from the mutation area while retaining the mutation topographic features, maximally reduces the artificial interference to the triangulation form of the non-mutation area, and realizes the high-fidelity reconstruction of the channel topography.

Description

Technical Field

[0001] This invention relates to the field of intelligent waterway sounding surface model construction, and particularly to a method for constructing high-precision complex sounding surface models based on the S-102 standard. Background Technology

[0002] With the deepening integration of information technologies such as big data and cloud computing with the water transport industry, the construction of smart waterways has become an inevitable requirement for promoting the digital and intelligent transformation of maritime transportation. A high-precision, high-fidelity digital elevation model (DEM) of the seabed is a key component of the digital foundation of smart waterways, directly impacting shipping operations such as depth analysis, navigation safety assessment, and maintenance workload calculation. The S-102 bathymetry surface product specification, as one of the S-100 series standards, defines the content model for high-resolution regular grid DEM products. Therefore, to ensure data standardization, a DEM that accurately reflects the waterway topography should be constructed based on the S-102 standard, providing a reliable data foundation for the smart waterway system.

[0003] Under artificial intervention, seabed topography often exhibits localized changes. For example, waterways and harbor basins are typically designed with a trapezoidal cross-section, characterized by a deeper central trough-like structure, with shallower sections retained on either side, transitioned by slope structures. However, in actual production, such as... Figure 1 As shown, the DEM constructed using the currently widely used interpolation method has the characteristics of continuity and smoothness, which makes it easy to over-smooth the terrain, resulting in distortion of terrain details. The smoothing of the waterway slope affects waterway maintenance decisions and increases unnecessary dredging workload.

[0004] In recent years, scholars have paid attention to the issue of high-fidelity DEMs and have begun to study abrupt terrain changes. Existing research includes Zhao Yuanyuan et al., who verified the advantages of elevation point densification and parallel break line methods in handling abrupt surface terrain changes; Wang Chun et al., who proposed a "vectorized simulation, rasterized organization" method and constructed a feature-embedded digital elevation model (F-DEM) that integrates terrain field features and object features; Xia Yuanhanzi et al., who proposed using virtual break line methods or topological break line methods to represent cases with low point cloud density; and Tian Fengmin et al., who focused on the construction method of seabed CD-TIN DEMs, achieving a reasonable representation of regional boundaries. However, most of the above studies focus on natural terrain, with very little research on seabed topography, and there is a lack of targeted research on high-fidelity DEMs of seabed shipping lanes that are closely related to human activities.

[0005] Triangulated Irregular Network (TIN) is a commonly used data structure in DEM (Digital Elevation Model) construction. It represents the terrain surface by connecting irregularly distributed discrete elevation points within a region to form a series of continuous triangles. Compared to regular grid DEMs, TINs can adaptively organize the triangular network structure according to the complexity of the terrain, showing better performance in representing important terrain features such as ridgelines and valley lines. Therefore, TINs are more suitable for generating DEMs for the concave slope characteristics of waterways. However, in actual terrain modeling, if the triangular network is constructed solely based on discrete points, the generated triangles may cross terrain structure boundaries, leading to inconsistencies between the triangular network structure and the actual terrain features.

[0006] To address this issue, constrained Delaunay triangulation (CDT) is commonly used in engineering practice. This method introduces break lines or fracture lines as constraints, forcing certain line segments to participate as edges in the triangulation, thus preserving terrain structure features to some extent. Break lines are divided into soft lines and hard lines. Soft lines only represent the location of terrain structures and do not contain elevation information, thus having a relatively limited constraint effect on the terrain and cannot guarantee that all line segments will participate as edges in the triangulation. Hard lines, which contain elevation information, can meet this requirement. However, the addition of hard lines alters the original topology of the triangulation, and in some areas, may even affect the accuracy of the original terrain representation.

[0007] Therefore, there is an urgent need for a high-precision method for constructing complex sounding surface models based on the S-102 standard to address the shortcomings of existing technologies. Summary of the Invention

[0008] The purpose of this invention is to propose a method for constructing a high-precision complex bathymetric surface model based on the S-102 standard. While preserving the original topographic features as much as possible, this method improves the ability of bathymetric data to express seabed topographic details by making reasonable use of hard-line constraints, thereby realizing the construction of a high-fidelity DEM for the waterway area. Furthermore, the DEM is produced and packaged according to the S-102 standard, promoting the application of S-102 in the construction of smart waterways.

[0009] To achieve the above objectives, this invention provides a method for constructing a high-precision complex bathymetry surface model based on the S-102 standard, comprising:

[0010] S1. Segment the original discrete water depth point data and extract the target hard break line;

[0011] S2. Using the original discrete water depth point data and the target hard break line, construct a constrained Delaunay triangulation model;

[0012] S3. Based on the constrained Delaunay triangulation model, the natural neighborhood method is used to generate a channel regular grid DEM, and the uncertainty of the bathymetry grid data is calculated grid by grid.

[0013] S4. Based on the uncertainty of the channel rule grid DEM and the sounding grid data, obtain the construction results of the high-precision complex sounding surface model.

[0014] Optionally, S1, segment the original discrete water depth point data and extract the target hard break line, including:

[0015] The raw discrete water depth data is obtained and segmented according to spatial location to obtain water depth data inside the waterway and water depth data outside the waterway;

[0016] Based on the water depth point data in the waterway, Kriging is used to interpolate and encrypt the data to generate an encrypted water depth point set.

[0017] Based on the encrypted water depth point set, combined with the water depth point data outside the channel and the water depth point data inside the channel, the channel edge line is segmented and assigned values ​​to obtain the target hard break line. The target hard break line includes the channel edge line hard break line, the slope line and the slope toe line.

[0018] Optionally, based on the encrypted depth point set, combined with the external depth point data and the internal depth point data of the channel, segmentation and assignment of channel edge lines are performed to obtain the target hard break line, including:

[0019] The channel edge line is segmented at equal intervals to obtain several small edge line segments;

[0020] A spatial index is constructed based on the encrypted water depth point set using the nearest neighbor matching algorithm;

[0021] The target water depth point is found using Euclidean distance based on the spatial index and several edge segments, and the channel edge is assigned a value to obtain the initial assigned channel edge hard break line.

[0022] The continuity of the initially assigned hard break line of the waterway edge is detected to obtain the hard break line of the waterway edge.

[0023] The nodes of the channel edge are densified using a search radius threshold, and the nearest points are matched and assigned values ​​based on the water depth data outside the channel and the water depth data inside the channel, respectively generating the slope line and the slope toe line.

[0024] Optionally, the search radius threshold is:

[0025]

[0026] Where d1 is the survey line spacing, r is the search radius threshold, and d2 is the depth measurement point spacing on the same survey line.

[0027] Optionally, S2, using the original discrete water depth point data and the target hard break line, a constrained Delaunay triangulation model is constructed, including:

[0028] Using the original discrete water depth data and the target hard break line as constraints, triangulation is performed based on the constrained Delaunay triangulation algorithm to construct a constrained Delaunay triangulation model. The formula for calculating the constraints is as follows:

[0029]

[0030] Where e represents any hard-cut line segment, L0 represents the hard-cut line of the channel edge, and L outer For the slope line, L inner For the toe line, E final This is the set of all edges of the final generated triangular mesh.

[0031] Optionally, S3, based on the constrained Delaunay triangulation model, a channel regular grid DEM is generated using the natural neighborhood method, and the uncertainty of the bathymetry grid data is calculated grid by grid, including:

[0032] Based on the constrained Delaunay triangulation model, the water depth value of the grid center point is obtained by using the natural neighborhood method combined with interpolation weights.

[0033] Based on the water depth value at the center point of the grid, a channel rule grid DEM is generated, and the corresponding neighborhood measurement points and weights of the grid are obtained.

[0034] The depth measurement error of natural neighboring points is retrieved using the neighborhood measurement points and weights corresponding to the grid, and the uncertainty of the depth measurement grid data is obtained by performing grid-by-grid calculations using the weights corresponding to the grid and the depth measurement errors of the natural neighboring points.

[0035] Optionally, the formula for calculating the water depth at the center point of the grid is as follows:

[0036]

[0037] Where D(x) is the water depth value at each grid center point x, n is the total number of natural neighbor points around the grid center point x, and w i Let D(p) be the interpolation weight for the i-th neighboring point. i ) represents the water depth value of the i-th natural neighbor point.

[0038] Optionally, the uncertainty of the sounding grid data is calculated using the following formula:

[0039]

[0040] in, To measure the uncertainty of the grid data, n is the total number of natural neighborhood points around the grid center point x, and w i The interpolation weights for the i-th neighboring point are... Let be the depth sounding error of the i-th natural neighbor point.

[0041] Optionally, S4, based on the uncertainty of the channel rule grid DEM and the sounding grid data, obtain the construction results of a high-precision complex sounding surface model, including:

[0042] The uncertainty of the waterway rule grid DEM is matched with that of the bathymetry grid data to obtain a matching dataset;

[0043] Based on the S-102 data specification, the matching dataset is encapsulated into the feature instance group of the S-102 product to obtain the rule grid of the two-dimensional array;

[0044] The regular grid of the two-dimensional array is verified, and the S-102 standard sounding surface product is obtained as the result of constructing a high-precision complex sounding surface model.

[0045] Optionally, the attribute storage of the feature instance group of the S-102 product includes geographic location information, such as the bounding box range, grid origin latitude and longitude, grid latitude and longitude interval, number of points in latitude and longitude direction, starting sequence, and number of instance groups.

[0046] Compared with the closest existing technology, the present invention has the following advantages:

[0047] This invention preserves the characteristics of abrupt terrain changes while avoiding the introduction of additional mandatory conditions for constructing triangulation networks far from abrupt areas. This minimizes human interference with the triangulation morphology in non-abrupt areas, achieving high-fidelity reconstruction of waterway topography. Furthermore, the entire process is automated through Python programming, improving data processing efficiency. Compared to conventional interpolation models, this invention extracts hard-line data points with water depth values ​​as constraints and employs a side-by-side assignment strategy to avoid water depth interference. Simultaneously, the "three hard-line constraints" mechanism effectively constrains the water depth differences at the upper and lower boundaries of abrupt terrain changes, enhancing the reconstruction effect. This invention encodes and encapsulates DEM products according to the S-102 standard, generating standardized bathymetric surface products for intelligent waterway systems, promoting the alignment of intelligent waterways with international standards. Attached Figure Description

[0048] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0049] Figure 1 The image shows the effect of DEM over-smoothing problem constructed by the existing interpolation method proposed in this invention.

[0050] Figure 2 This is a flowchart illustrating the high-precision complex depth sounding surface model construction method based on the S-102 standard according to an embodiment of the present invention.

[0051] Figure 3 This is a flowchart of the S-102 product generation process based on hard-break line constraints proposed in an embodiment of the present invention;

[0052] Figure 4 This is a schematic diagram illustrating the principle of a single beam proposed in an embodiment of the present invention;

[0053] Figure 5 This is a flowchart illustrating the generation process of a hard break line along the waterway edge, as proposed in an embodiment of the present invention.

[0054] Figure 6 This is a schematic diagram illustrating the principle of channel edge densification and nearest neighbor assignment proposed in an embodiment of the present invention;

[0055] Figure 7 This is a flowchart illustrating the construction process of the hard-break line on the inner and outer sides of the waterway as proposed in an embodiment of the present invention;

[0056] Figure 8 This is a schematic diagram of the construction of hard-break lines on the inner and outer sides of the waterway according to an embodiment of the present invention;

[0057] Figure 9 This is a schematic diagram of the TIN construction of the waterway under hard break line constraints proposed in an embodiment of the present invention;

[0058] Figure 10 This is a diagram of the regular grid DEM encoding structure based on the S-102 standard proposed in an embodiment of the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions in the embodiments of this invention will be clearly and completely described below with reference to specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0060] The terminology used in the embodiments section of this invention is for the purpose of explaining specific embodiments of the invention only, and is not intended to limit the invention.

[0061] Generally, multibeam bathymetry data can effectively improve the accuracy of seabed topographic mapping, while single-beam bathymetry data remains the most commonly used data source in routine waterway maintenance due to its low cost, ease of operation, and fast data processing. Therefore, this invention addresses the needs of smart waterway construction by proposing a method for constructing high-precision complex bathymetry surface models based on the S-102 standard. This method supports scientific decision-making in smart waterways. Based on the actual data usage, experimental verification is conducted using single-beam bathymetry data as an example to address the problem of excessive smoothing in representing waterway slopes using water depth data.

[0062] like Figure 2-3 As shown, this embodiment of the invention provides a method for constructing a high-precision complex sounding surface model based on the S-102 standard, including:

[0063] S1. Segment the original discrete water depth point data and extract the target hard break line;

[0064] The raw discrete water depth data of a single beam is acquired. Based on the channel boundary vector, the measuring points are divided into two categories of data: inside and outside the channel, by spatial location determination. Then, Kriging interpolation is used to densify the water depth points inside the channel and complete the assignment of the main hard line of the channel edge. Subsequently, based on the segmented channel edge, the measured water depth points on the outside and inside of the channel are called and matched to form lines and assigned measured water depths, generating two hard lines with water depth: the slope line and the slope toe line. This completes the three types of hard lines used to constrain the terrain boundary, providing a forced terrain constraint element for the subsequent construction of the triangulation network, and avoiding the problem of mutual interference between water depth data inside and outside the channel from the source.

[0065] S2. Using the original discrete water depth point data and the target hard break line, construct a constrained Delaunay triangulation model;

[0066] Discrete water depth points, channel edge hard break lines, slope lines, and slope toe lines are used as constraints input into the triangulation algorithm. Constrained Delaunay triangulation operations are performed to ensure that all hard break line segments are fixed as permanent boundary edges of the triangulation network and are not flipped or deleted by the algorithm. This isolates the two types of different terrains, namely deep channels and shallow shoals on the outer side, and prevents the generated triangles from crossing the abrupt boundary of the channel slope. This solves the defects of traditional unconstrained triangulation networks that overly smooth channel slopes and lose details of terrain abrupt changes, forming a CDT terrain triangulation base that can realistically reproduce the trapezoidal cross-section morphology of artificial channels.

[0067] S3. Based on the constrained Delaunay triangulation model, the natural neighborhood method is used to generate a channel regular grid DEM, and the uncertainty of the bathymetry grid data is calculated grid by grid.

[0068] Using the constructed constrained Delaunay triangulation model as the computational basis, rasterization processing was carried out. The natural neighborhood interpolation method was used to generate a uniform resolution channel regular grid DEM. During the interpolation operation, three hard break lines were used as the boundary of the neighborhood search. The interpolation weights were calculated based on the area ratio of Voronoi cells to ensure that the interpolation results at abrupt slope changes closely match the actual seabed topography. At the same time, based on the measurement error inherent in the original measuring points, combined with the natural neighborhood interpolation weights, the corresponding water depth uncertainty was calculated for each grid point according to the error propagation law. A standardized water depth grid and supporting accuracy quantification indicators were generated simultaneously to meet the S-102 standard's mandatory requirement for uncertainty information in bathymetry data.

[0069] S4. Based on the uncertainty of the channel rule grid DEM and the sounding grid data, obtain the construction results of the high-precision complex sounding surface model;

[0070] The regular grid DEM water depth data generated in this step, along with the water depth uncertainty calculated grid by grid, are encoded and encapsulated according to the S-100 HDF5 data model and the S-102 bathymetry coverage product layering specification. A standardized HDF5 file containing geographic metadata, HDF5 dataset, and extreme value attributes is constructed. The final output is a high-precision bathymetry surface model that conforms to the international S-102 standard, fully preserves the details of the waterway slope, and includes quantitative accuracy information. It can be directly connected to the smart waterway platform to carry out business applications such as navigation safety assessment and waterway maintenance calculation.

[0071] In summary, steps S1 to S4 first partition the original discrete water depth points and extract multiple sets of hard discontinuities with water depth as topographic constraints. Then, combining the original measuring points and hard discontinuities, a constrained Delaunay triangulation network is constructed to isolate the topography inside and outside the waterway and avoid excessive smoothing distortion of the slope topography. Next, based on the triangulation network, a regular grid DEM that fits the real abrupt changes in waterway topography is generated using the natural neighborhood method. The water depth uncertainty is calculated grid by grid according to the error propagation law. Finally, the water depth values ​​and uncertainties are integrated and packaged according to the S-102 standard to obtain the S-102 standard bathymetric surface product. This invention relies on the three hard discontinuities constraint mechanism to completely preserve key topographic details such as waterway slopes and toes, significantly improving the fidelity of the seabed topographic model. The quantified water depth uncertainty meets the S-102 standard data specifications, and the entire process is automated to output standardized HDF5 products. This can effectively reduce the measurement deviation of dredging engineering quantities and provide a high-precision, standardized, and shareable seabed digital elevation data base for intelligent waterway navigation safety assessment and waterway maintenance decision-making.

[0072] As one possible implementation, in the above embodiments, step S1 may specifically include the following steps:

[0073] S1-1. Obtain the original discrete water depth point data by dividing it according to spatial location, and obtain the water depth point data inside the waterway and the water depth point data outside the waterway.

[0074] After acquiring single-beam bathymetry data, it is divided into two parts according to spatial location: water depth data within the channel and water depth data outside the channel. Specifically:

[0075] The system collects raw discrete water depth data using a single beam and performs spatial location discrimination by combining it with the channel boundary vector. Through the point-area inclusion relationship, all measuring points are divided into two independent sets of measuring point data: one inside the channel and one outside the channel. This distinguishes the measured points in the deep channel area from the external shallow shoal area, providing a regional data source for subsequent regional densification of water depth points and the construction of slope lines and toe lines, thus avoiding mutual interference between internal and external topographic measuring points.

[0076] S1-2. Based on the water depth point data in the channel, interpolation and encryption are performed using the Kriging method to generate an encrypted water depth point set;

[0077] From a physical perspective, the water depth at the hard break line of a waterway boundary should be determined by the water depth point within the waterway to accurately reflect the topographic features of the waterway area. However, as... Figure 4 As shown in the schematic diagram of the single-beam principle, single-beam bathymetry data has the limitation of being dense along the survey line and sparse perpendicular to the survey line. The original survey points are often difficult to directly provide a reliable water depth reference for the entire channel edge.

[0078] To address this issue, the depth sounding data within the waterway was first interpolated and densified. Given the relatively flat terrain of the deep channel and considering the "bull's-eye" effect that the inverse distance weighting method is prone to in sparse data areas, ordinary kriging was chosen to interpolate the depth points within the waterway, generating a continuous depth raster surface. The raster size was set according to the target DEM resolution. Then, the raster was converted into a dense set of depth points, serving as a candidate data source for subsequent hard-line assignment.

[0079] S1-3. Based on the encrypted depth point set, combined with the external depth point data and the internal depth point data of the waterway, segment and assign values ​​to the waterway edge line to obtain the target hard break line.

[0080] Based on the generated set of encrypted water depth points, the water depth of the channel edge is assigned to segments, generating hard cut lines of the channel edge as the benchmark for terrain abrupt changes. Then, using the channel edge as the benchmark, the positioning nodes are encrypted and matched with the original measured points outside and inside the channel segmented in the first step. The matched points are connected sequentially and assigned measured water depth values, generating two types of hard cut lines: slope lines and toe lines. Finally, all target hard cut lines of the channel edge, slope lines, and toe lines are collected to form a three-layer terrain constraint control line, providing forced terrain boundary conditions for the subsequent construction of the constraint triangulation network.

[0081] In summary, steps S1-1 to S1-3 first partition the original discrete water depth points according to the channel spatial boundary, separating the water depth point data inside and outside the channel. Then, the water depth point data inside the channel is densified by Kriging interpolation to obtain a high-density water depth point set to solve the problem of sparse lateral measurement points in single beams. Finally, the densified water depth point set is used to complete the water depth assignment of the channel edge to generate the channel edge hard line. Combined with the original measurement points inside and outside the channel, slope lines and slope toe lines are constructed respectively. The entire partitioning-densification-multiple hard line extraction process can accurately delineate the terrain boundary inside and outside the channel. Relying on multiple sets of hard lines with water depth to form layered terrain constraints, it can effectively isolate the water depth data of deep and shallow areas near the channel, laying a solid data foundation for subsequent triangulation modeling to avoid the problems of excessive smoothing of slope terrain and loss of terrain details.

[0082] To address the challenges in constructing DEMs of abruptly changed terrain along waterways, this invention proposes an optimization strategy based on the TIN construction process, employing three hard-line constraints. The core idea is to extract waterway edge lines with water depth values ​​from the depth point data as hard-line constraints. Simultaneously, to reduce the interference of the waterway edge lines on the terrain inside and outside the waterway, two additional hard-line constraints with water depth values ​​are constructed on each side, jointly constraining the construction of the TIN model. As a possible implementation, in the above embodiment, steps S1-3 may specifically include the following steps:

[0083] S1-3-1. Divide the channel edge line into equal-distance segments to obtain several edge line segments;

[0084] like Figure 5-6 As shown, the channel edge line, serving as a location benchmark for abrupt topographic changes, is one of the hard break lines. After interpolating and densifying the water depth points within the channel, the water depth point closest to the channel edge line is retrieved segment by segment, and the water depth value of this point is used to assign values ​​to the channel edge line, thereby obtaining the hard break line of the channel edge line with water depth values. To improve the accuracy of the assignment, the channel edge line is first divided into several small segments of a set length. The segment length needs to be adapted to the density of the densified water depth points to ensure that there are enough water depth points in each segment as the basis for assignment.

[0085] S1-3-2. Based on the encrypted water depth point set, a spatial index is constructed using the nearest neighbor matching algorithm;

[0086] Using the encrypted depth point set generated by Kriging interpolation as the data source, a spatial index structure is built by employing the nearest neighbor matching algorithm. A spatial retrieval mapping relationship is established based on the planar coordinates of discrete depth points, which significantly reduces the retrieval time for matching the center point of the subsequent edge line segment with the nearest depth point and improves the automated processing efficiency of the overall assignment of channel edge lines.

[0087] S1-3-3. Based on the spatial index and several edge segments, use Euclidean distance to find the target water depth point, and assign values ​​to the channel edge to obtain the initial assigned channel edge hard break line.

[0088] like Figure 6-7 As shown, each edge segment is traversed and its center point is extracted. Based on the constructed spatial index, within a limited search radius, the Euclidean distance formula is used. Find the water depth point closest to the center point of each edge segment and assign that water depth to the current edge segment. The search area is limited to the interior of the channel to ensure that the assignment conforms to the fundamental principle that the hard break line of the channel edge should be determined by the inner water depth point.

[0089] S1-3-4. Perform continuity detection on the initially assigned hard break line of the waterway edge to obtain the hard break line of the waterway edge.

[0090] After all the edge segments have been assigned values, the continuity of the entire channel edge hard break line is checked to eliminate local abnormal jumps that may be caused by matching different data sources, ensuring the rationality of water depth changes along the line, and finally obtaining a channel edge hard break line with continuous water depth values, logically reasonable, and with effective shape constraint capabilities.

[0091] S1-3-5. Node density is increased on the channel edge line using the search radius threshold, and the nearest neighbor points are matched and assigned values ​​based on the water depth data outside the channel and the water depth data inside the channel to generate the slope line and the slope toe line respectively.

[0092] The interpolation results of the channel edge line differ between the inner and outer sides of the channel. Using only the channel edge line as a hard break line will affect the construction of the TIN triangulation network outside the channel. To avoid the cross-influence of water depth values ​​on both sides of the channel, two hard break lines (slope line and toe line) are created, extending outwards from the channel edge line as the reference. The hard break lines are obtained by connecting the water depth points closest to the channel edge line one by one, and the hard break lines are assigned values ​​according to the water depth values ​​of the connecting points. Following the principle of one value per side, the slope line is generated using the water depth points on the outer side of the channel, while the toe line is generated using the water depth points on the inner side of the channel.

[0093] The hard break lines on both sides (i.e., slope lines and toe lines) are directly extracted from the original single-beam bathymetry data to ensure data reliability. The core idea is to use the channel edge as a spatial reference, searching for the nearest true water depth points both inside and outside the channel, and then connecting these points sequentially to form break lines. This ensures that the water depth at each point on the line is supported by measured data. The slope line searches only for the nearest water depth point in the outer area of ​​the channel, while the toe line searches only for the nearest water depth point in the inner area of ​​the channel, ensuring that the two hard break lines represent water depth information for different terrain types. The slope lines and toe lines generated using this method retain the authenticity of the original data and provide accurate terrain boundary control for subsequent TIN construction. The specific implementation steps for each hard break line are as follows:

[0094] First, based on the channel edge line, the nodes of the channel edge line are densified according to a set interval, and the node data is exported as the position reference to be matched. The interval is mainly determined with reference to the survey line spacing d1. Then, a spatial index is constructed for the original single-beam depth points to improve the efficiency of nearest neighbor search. Then, according to the set search radius threshold, the node data will sequentially retrieve the nearest depth point from the depth measurement data on the corresponding channel side, such as... Figure 8 As shown. The setting of the search radius threshold r needs to comprehensively consider the survey line spacing d1 and the depth measurement point spacing d2 on the same survey line. Too small a value may result in no matching points, while too large a value may introduce interfering points.

[0095] For the nearest water depth point found, its coordinates and water depth value information are stored sequentially. After processing all the vertices in sequence, the stored water depth point information is connected in its original order to obtain a single-sided break line composed of real water depth points. Then, the water depth values ​​of the water depth point nodes are assigned to the break line according to the method of assigning values ​​to the channel edge line, thus obtaining a hard break line with water depth.

[0096] In summary, steps S1-3-1 to S1-3-5 divide the channel edge line into equal segments according to the appropriate density to obtain small segments of the edge line. A spatial index is constructed based on the dense water depth point set generated by Kriging interpolation inside the channel. Within the search range inside the channel, the nearest dense water depth point is matched segment by segment using the Euclidean distance formula to complete the initial assignment of the edge line. After continuity verification and elimination of abnormal jumps, a stable and reliable main hard line of the channel edge line is obtained. Subsequently, the reference nodes are densified according to the survey line spacing based on the channel edge line. A reasonable search radius is set, and the connection line of the original measured water depth points on the outer and inner sides of the channel is matched and assigned values ​​to generate two hard lines: the slope line and the slope toe line. This process takes into account Kriging interpolation encryption, nearest neighbor retrieval assignment, and continuity verification, avoiding the problems of sparse lateral measurement points in single beams and the bullseye effect of the inverse distance weighting method. It ensures that the water depth of the main hard line conforms to the real terrain inside the waterway, and the two hard lines rely on the original measured data to accurately restore the abrupt boundary of the waterway slope, forming a complete terrain-constrained hard line system. This provides accurate boundaries for subsequent constrained triangulation modeling, effectively preventing the waterway slope terrain from being excessively smoothed and distorted, and preserving the details of the real waterway cross section.

[0097] As one possible implementation, in the above embodiments, step S2 may specifically include the following steps:

[0098] Using the original discrete water depth data and the target hard break line as constraints, triangulation is performed based on the constrained Delaunay triangulation algorithm to construct a constrained Delaunay triangulation model. Specifically:

[0099] The previously processed single-beam bathymetry point set {P} (i.e., the original discrete water depth point data), the hard break line L0 of the channel edge, and the slope line L are combined. outerSlope line L inner As a constraint, the point set is triangulated using the constrained Delaunay triangulation algorithm to ensure that all hard breaks exist as triangle edges and are not flipped or deleted by the algorithm. The formula for calculating the constraint is as follows:

[0100]

[0101] Where e is any hard-broken line segment, E final This is the set of all edges of the final generated triangulated mesh. A high-fidelity constrained Delaunay triangulation model is constructed based on hard-break lines. This model uses hard-break lines as fixed boundaries, preventing the triangulated mesh from crossing terrain boundaries, accurately reproducing the abrupt contours of the channel slope, ensuring the correct terrain boundary structure, and providing a reliable mesh base for subsequent high-precision water depth model construction. The effect is as follows: Figure 9 As shown.

[0102] As one possible implementation, in the above embodiments, step S3 may specifically include the following steps:

[0103] S3-1. Based on the constrained Delaunay triangulation model, the water depth value of the grid center point is obtained by using the natural neighborhood method combined with interpolation weights.

[0104] The CDT model can generate a regular grid model based on the natural neighborhood method. The natural neighborhood method relies on the Voronoi diagram and defines the interpolation weights based on the area ratio of neighboring cells. For each grid center point x to be interpolated, its water depth value is obtained by weighting the values ​​of the surrounding natural neighbor points. The formula for calculating the water depth value D(x) of each grid center point x is as follows:

[0105]

[0106] Where D(x) is the water depth value at each grid center point x, n is the total number of natural neighbor points around the grid center point x, and w i Let D(p) be the interpolation weight for the i-th neighboring point. i Let be the water depth value of the i-th natural neighbor point. When encountering hard discontinuities, the search neighborhood and area calculation will not cross the hard discontinuities, thus more realistically reflecting abrupt terrain changes.

[0107] S3-2. Based on the water depth value of the center point of the grid, generate a channel rule grid DEM and obtain the corresponding neighborhood measurement points and weights of the grid.

[0108] The water depth values ​​calculated from all grid center points are integrated into a continuous grid model to generate a channel regular grid DEM. At the same time, the neighboring measurement points p corresponding to each grid cell are recorded. i Interpolation weight w iKey parameters such as the total number of neighboring measurement points n are reserved to provide basic data for subsequent uncertainty calculations. At the same time, the required grid metadata fields (boundary box, grid origin, grid spacing, number of grid points, etc.) for the S-102 format are reserved.

[0109] S3-3. Use the neighborhood measurement points and weights corresponding to the grid to retrieve the depth measurement error of the natural neighboring points, and perform grid-by-grid calculations using the weights corresponding to the grid and the depth measurement error of the natural neighboring points to obtain the uncertainty of the depth measurement grid data.

[0110] According to the law of error propagation, for each sounding grid, the water depth value is obtained by calculating the area weighted average of natural neighborhood points. Therefore, its uncertainty is... The uncertainty should be calculated by propagating the error from natural neighborhood points using an area-weighted method. Natural neighborhood points, as vertices of the Delaunay triangle, correspond to the original measurement points. Their uncertainty primarily stems from the cumulative errors in data acquisition equipment accuracy and data processing; relevant accuracy information is recorded in the data attributes. Uncertainty of sounding grid data. The calculation formula is as follows:

[0111]

[0112] in, Let be the depth sounding error of the i-th natural neighbor point.

[0113] In summary, steps S3-1 to S3-3 rely on a constrained Delaunay triangulation model embedded with hard discontinuities in the waterway. The natural neighborhood method, based on Voronoi area weights, calculates the water depth value at the grid center point. The interpolation neighborhood calculation does not cross the hard discontinuities, reconstructing the abrupt topography of the waterway. Then, the water depth values ​​are integrated to generate a regular grid DEM of the waterway, and the corresponding neighborhood measurement points and interpolation weights for each grid are retained. Next, the original measurement point depth measurement error is retrieved, and the uncertainty is calculated grid-by-grid according to the error propagation law. This process accurately reconstructs the abrupt topography of the waterway slope, avoids excessive smoothing and distortion, and ensures the authenticity of the DEM topography representation. It also quantifies the vertical uncertainty of the water depth data grid-by-grid, achieving high-precision modeling and error tracing of waterway depth measurement data, supporting the construction of the S-102 standardized waterway depth measurement product and its application in navigation safety assessment.

[0114] As one possible implementation, in the above embodiments, step S4 may specifically include the following steps:

[0115] S4-1. Match the uncertainty of the channel rule grid DEM with the bathymetry grid data to obtain a matching dataset;

[0116] Based on the constrained Delaunay triangulation model, a regular grid DEM of the waterway is generated through natural neighborhood interpolation. At the same time, uncertainty data of the bathymetric grid at the same resolution is calculated grid by grid using the error propagation formula. The coordinate systems, number of rows and columns, and boundary ranges of the two types of regular grids are completely consistent. Spatial matching is completed pixel by pixel, and a matching dataset is established that corresponds one-to-one between the water depth value at the center point of each grid and its corresponding vertical position uncertainty. This provides matching paired data for the two-dimensional array required for subsequent storage in the S-102 standard.

[0117] S4-2. Based on the S-102 data specification, the matching dataset is encapsulated into the feature instance group of the S-102 product to obtain the rule grid of the two-dimensional array;

[0118] After obtaining the regular grid DEM, it is encoded according to the S-102 structure. S-102 data products follow the "S-100HDF5 Data Model and File Format" specification, with the specific structure as follows: Figure 10 As shown. The attributes of the feature instance group (BathymetricCoverage.01) of the S-102 product store the geographic location information of bathymetric coverage, including the bounding box, the longitude and latitude of the grid origin, the grid spacing, the number of points in the latitude and longitude direction, the start sequence, and the number of groups. The HDF5 dataset in the product includes a two-dimensional array defined according to the number of points in the latitude and longitude direction, i.e., a regular grid. Each discrete point in the grid stores bathymetric grid data containing water depth and uncertainty data. The attribute of the feature instance group is obtained according to the S-102 grid metadata fields preset in step S3-2. After the matching dataset of water depth values ​​and their vertical position uncertainty is output in step S4-1, it can be converted into an HDF5 dataset according to the S-102 standard.

[0119] S4-3. Verify the regular grid of the two-dimensional array and obtain the S-102 standard sounding surface product as the result of constructing a high-precision complex sounding surface model;

[0120] A full-dimensional quality check is performed on the two-dimensional array regular grid storing water depth and uncertainty. This check verifies the integrity of geographic metadata of the element instance group, the effect of water depth numerical terrain reconstruction, and the rationality of the distribution of grid uncertainty values, eliminating abnormal and distorted data. After passing the check, a compliant S-102 standard bathymetric surface HDF5 product is generated. This product is the final result of the high-precision complex bathymetric surface model and can be connected to the smart waterway platform as a standardized digital foundation, providing high-precision terrain data support with error traceability for navigation safety assurance, maintenance plan optimization, and waterway evolution analysis.

[0121] In summary, steps S4-1 to S4-3 first match the channel regular grid DEM with the bathymetry uncertainty raster at the same resolution pixel by pixel to form a matching dataset with a one-to-one correspondence between water depth and error. Then, according to the S-100 HDF5 and S-102 standards, the raster geographic information is extracted, filled into the attribute of the feature instance group, and a two-dimensional array regular grid storing water depth and uncertainty is constructed. Finally, the integrity and numerical rationality of the two-dimensional array grid data are verified to generate a standardized S-102 bathymetry surface model. This process relies on the interpolation results of the three hard-line constrained triangular network to fully preserve the details of the abrupt topographic changes on the channel slope. At the same time, each grid is equipped with a quantified vertical position uncertainty. The product format conforms to international maritime bathymetry standards and can be directly connected to the smart waterway platform as a standardized digital base. It can achieve high-precision waterway topographic reconstruction and support data error tracing, effectively supporting navigation safety assessment, dredging and maintenance plan preparation, and long-term evolution trend analysis of the waterway.

[0122] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0123] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0124] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0125] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for constructing a high-precision complex bathymetry surface model based on the S-102 standard, characterized in that, include: S1. Segment the original discrete water depth point data and extract the target hard break line; S2. Using the original discrete water depth point data and the target hard break line, construct a constrained Delaunay triangulation model; S3. Based on the constrained Delaunay triangulation model, the natural neighborhood method is used to generate a channel regular grid DEM, and the uncertainty of the bathymetry grid data is calculated grid by grid. S4. Based on the uncertainty of the channel rule grid DEM and the sounding grid data, obtain the construction results of the high-precision complex sounding surface model.

2. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 1, characterized in that, S1. Segment the original discrete water depth point data and extract the target hard break lines, including: The raw discrete water depth data is obtained and segmented according to spatial location to obtain water depth data inside the waterway and water depth data outside the waterway; Based on the water depth point data in the waterway, Kriging is used to interpolate and encrypt the data to generate an encrypted water depth point set. Based on the encrypted water depth point set, combined with the water depth point data outside the channel and the water depth point data inside the channel, the channel edge line is segmented and assigned values ​​to obtain the target hard break line. The target hard break line includes the channel edge line hard break line, the slope line and the slope toe line.

3. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 2, characterized in that, Based on the encrypted depth point set, combined with the external and internal depth point data of the waterway, the waterway edge line is segmented and assigned values ​​to obtain the target hard break line, including: The channel edge line is segmented at equal intervals to obtain several small edge line segments; A spatial index is constructed based on the encrypted water depth point set using the nearest neighbor matching algorithm; The target water depth point is found using Euclidean distance based on the spatial index and several edge segments, and the channel edge is assigned a value to obtain the initial assigned channel edge hard break line. The continuity of the initially assigned hard break line of the waterway edge is detected to obtain the hard break line of the waterway edge. The nodes of the channel edge are densified using a search radius threshold, and the nearest points are matched and assigned values ​​based on the water depth data outside the channel and the water depth data inside the channel, respectively generating the slope line and the slope toe line.

4. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 3, characterized in that, The search radius threshold is: Where d1 is the survey line spacing, r is the search radius threshold, and d2 is the depth measurement point spacing on the same survey line.

5. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 1, characterized in that, S2. Using the original discrete water depth point data and the target hard break line, construct a constrained Delaunay triangulation model, including: Using the original discrete water depth data and the target hard break line as constraints, triangulation is performed based on the constrained Delaunay triangulation algorithm to construct a constrained Delaunay triangulation model. The formula for calculating the constraints is as follows: Where e represents any hard-cut line segment, L0 represents the hard-cut line of the channel edge, and L outer For the slope line, L inner For the toe line, E final This is the set of all edges of the final generated triangular mesh.

6. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 1, characterized in that, S3. Based on the constrained Delaunay triangulation model, a channel regular grid DEM is generated using the natural neighborhood method, and the uncertainty of the bathymetry grid data is calculated grid by grid, including: Based on the constrained Delaunay triangulation model, the water depth value of the grid center point is obtained by using the natural neighborhood method combined with interpolation weights. Based on the water depth value at the center point of the grid, a channel rule grid DEM is generated, and the corresponding neighborhood measurement points and weights of the grid are obtained. The depth measurement error of natural neighboring points is retrieved using the neighborhood measurement points and weights corresponding to the grid, and the uncertainty of the depth measurement grid data is obtained by performing grid-by-grid calculations using the weights corresponding to the grid and the depth measurement errors of the natural neighboring points.

7. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 6, characterized in that, The formula for calculating the water depth at the center point of the grid is as follows: Where D(x) is the water depth value at each grid center point x, n is the total number of natural neighbor points around the grid center point x, and w i Let D(p) be the interpolation weight for the i-th neighboring point. i ) represents the water depth value of the i-th natural neighbor point.

8. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 6, characterized in that, The formula for calculating the uncertainty of the sounding grid data is: in, To measure the uncertainty of the grid data, n is the total number of natural neighborhood points around the grid center point x, and w i The interpolation weights for the i-th neighboring point are... Let be the depth sounding error of the i-th natural neighbor point.

9. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 1, characterized in that, S4. Based on the uncertainty of the channel rule grid DEM and the sounding grid data, obtain the construction results of a high-precision complex sounding surface model, including: The uncertainty of the waterway rule grid DEM is matched with that of the bathymetry grid data to obtain a matching dataset; Based on the S-102 data specification, the matching dataset is encapsulated into the feature instance group of the S-102 product to obtain the rule grid of the two-dimensional array; The regular grid of the two-dimensional array is verified, and the S-102 standard sounding surface product is obtained as the result of constructing a high-precision complex sounding surface model.

10. The method for constructing a high-precision complex bathymetry surface model based on the S-102 standard according to claim 9, characterized in that, The attribute storage geographic location information of the feature instance group of the S-102 product includes the bounding box range, grid origin latitude and longitude, grid latitude and longitude interval, number of points in latitude and longitude direction, starting sequence and number of instance groups.

Images