Multi-level detail level single model generation method and system based on beidou grid code

Abstract

This invention provides a method and system for generating multi-level detail (LOD) individual models based on BeiDou grid codes. The method includes: acquiring and parsing BeiDou grid coding information to obtain all maximum-level grid codes; utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, deriving a hierarchical tile index file with a multi-scale LOD structure based on the prefix of each maximum-level grid code; for each sub-tile, transforming the WGS84 geodetic coordinates within the sub-tile's coverage area into coordinates compatible with the target 3D rendering engine via a coordinate transformation link including a local tangent plane coordinate system to obtain the sub-tile's local vertex data; constructing a transformation matrix for parent-child hierarchical connection; batch generating B3DM tile files based on the sub-tile's local vertex data and the transformation matrix, assigning a batch ID attribute to each individual model within the file under a given individualization style, and storing the individual model metadata corresponding to the batch ID in a batch table.

Description

Technical Field

[0001] This invention relates to the field of 3D model data processing technology, and in particular to a method and system for generating multi-level detail-level single-unit models based on BeiDou grid codes. Background Technology

[0002] With the rapid development of applications such as digital twin cities, low-altitude economy, and autonomous driving, the real-time visualization and efficient rendering capabilities of massive 3D model data have become a key technological bottleneck restricting the application and development of these scenarios. Traditional 3D model generation methods mainly rely on manual modeling or two-dimensional vector stretching. The former suffers from low automation, high cost, and long cycle time, making it difficult to meet the rapid visualization needs of massive mesh data. The latter, while capable of batch generating building bases, lacks direct support for spatial encoding and cannot utilize the hierarchical characteristics of the encoding itself to automatically construct multi-level levels of detail (LOD). Furthermore, although current mainstream 3D model formats (such as 3D Tiles) support LOD, their generation process usually requires the pre-preparation of model data with multiple levels of precision. A method has not yet been developed to reverse-engineer and generate 3D model datasets with parent-child hierarchical relationships from discrete, massive spatial encodings.

[0003] In recent years, BeiDou grid codes have received widespread attention in the field of spatial data organization and management. For example, in 2016, Sun Zhongqiu et al. systematically elaborated on the theoretical framework of the GeoSOT-3D ellipsoidal subdivision framework, demonstrating its feasibility in constructing true 3D data models. In 2020, Wu Yulong et al. proposed using BeiDou grid location codes to establish an emergency equipment data identification and service system, forming a grid-based emergency management framework and supporting the entire process of emergency command and rescue decision-making. In 2021, Li Lin et al. pointed out that traditional GIS+BIM modeling methods face bottlenecks in data organization and aggregation due to the privatization of object ID systems. BeiDou grid code technology provides an open big data architecture based on spatiotemporal coding, which can organize multi-source heterogeneous spatiotemporal data modeling through grid association. In 2025, Liu Longhao et al., addressing the challenges of spatiotemporal coupling modeling of multi-source elements in complex battlefield environments, proposed a multi-source element modeling and risk quantification method based on spatiotemporal grid subdivision, which can meet the real-time response requirements of UAV swarm trajectory planning. However, the above studies mainly focus on the application of BeiDou grid codes in two-dimensional identification, index organization, emergency management and refined management of airspace resources, and have not yet touched on the visualization of three-dimensional models.

[0004] In fact, BeiDou grid codes provide a globally unified, multi-scale spatial partitioning and encoding scheme. Their encoding prefixes naturally imply spatial scale and inclusion relationships, making them theoretically very suitable for driving the automatic generation of multi-level LOD models. However, there is currently a lack of automated methods for generating 3D Tiles datasets from massive amounts of discrete BeiDou grid codes. This deficiency makes it difficult to meet the practical application requirements for loading efficiency and rendering performance in loading and rendering large-scale 3D models, further restricting the in-depth application of BeiDou grid codes in large-scale 3D spatial information fields such as digital twin cities and low-altitude economy. Summary of the Invention

[0005] To address the problem that the loading and rendering efficiency of existing large-scene 3D models is insufficient to meet the needs of practical applications, this invention proposes a method and system for generating multi-level detail hierarchy single-unit models based on BeiDou grid codes. It utilizes the characteristics of BeiDou grid coding to efficiently derive the parent hierarchy structure, and improves the compactness of the LOD tree and the scheduling efficiency of the rendering engine by using spatial sorting and parent grouping and packaging strategies.

[0006] In a first aspect, the present invention provides a method for generating a multi-level detail-level single-unit model based on BeiDou grid codes, comprising:

[0007] Obtain and parse the BeiDou grid coding information to obtain all maximum level grid codes;

[0008] By utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, a hierarchical tile index file with a multi-scale LOD structure is constructed based on the prefix of each maximum level grid code.

[0009] For each sub-tile in the hierarchical tile index file, the WGS84 geodetic coordinates within the area covered by the sub-tile are transformed into coordinates adapted to the target 3D rendering engine through a coordinate transformation link that includes a local tangent plane coordinate system to obtain the local vertex data of the sub-tile; and a transformation matrix is ​​constructed based on the origin offset of the sub-tile and its parent tile in the local tangent plane coordinate system for use in parent-child hierarchical connection.

[0010] B3DM tile files are generated in batches based on the local vertex data and transformation matrix of the sub-tiles. Each individual model is assigned a batch ID attribute under a given individualization style, and the metadata of the individual model corresponding to the batch ID is stored in the batch table.

[0011] Furthermore, leveraging the hierarchical prefix inheritance characteristic of BeiDou grid codes, a hierarchical tile index file with a multi-scale LOD structure is constructed based on the prefix of each maximum-level grid code, including:

[0012] Utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, prefixes are extracted from each maximum-level grid code to derive the parent grid codes for each level. After deduplication, a unique set of grid codes for each level is obtained, and a global bounding box is calculated based on the spatial range corresponding to each grid code. Then, the grid codes within each level are sorted according to their spatial location. Based on the global bounding box, the parent-child relationship is determined to group and package all grid codes belonging to the same parent grid, forming batch data merged by parent level. Finally, based on the nesting relationship between the batch data and the levels, a hierarchical tile index file with a multi-scale LOD structure is generated.

[0013] Furthermore, the transformation of the WGS84 geodetic coordinates within the area covered by the sub-tile into coordinates compatible with the target 3D rendering engine via a coordinate transformation link including a local tangent plane coordinate system includes:

[0014] The WGS84 geodetic coordinates within the area covered by the sub-tile are sequentially converted into geocentric and geofixed coordinates and ENU tangent plane coordinates based on the local origin of the sub-tile. Then, the ENU coordinate axes are mapped to coordinates adapted to the target 3D rendering engine to obtain the local vertex data of the sub-tile.

[0015] Furthermore, before performing coordinate transformation, it also includes configuring general parameters and mode parameters; wherein, the general parameters include at least one of the following: input file path, output directory, maximum number of grids contained in each B3DM file, and number of parallel worker threads; the mode parameters include at least one of the following: LOD level derivation enable flag, LOD minimum level, and singleton mode enable flag.

[0016] Furthermore, the configuration information for the individualized style includes at least one of the following options: solid face or wireframe display mode, face color, transparency, whether to use a non-lighting material, whether to display border lines, border line color, and border line width.

[0017] Secondly, this invention provides a multi-level detail hierarchy single-unit model generation system based on BeiDou grid codes, comprising:

[0018] The data parsing module is used to obtain and parse the BeiDou grid coding information to obtain all the maximum level grid codes;

[0019] The LOD level derivation module is used to utilize the hierarchical prefix inheritance characteristics of Beidou grid codes to derive and construct a hierarchical tile index file with a multi-scale LOD structure based on the prefix of each maximum level grid code.

[0020] The coordinate transformation module is used to transform the WGS84 geodetic coordinates within the coverage area of ​​each sub-tile in the hierarchical tile index file into coordinates adapted to the target 3D rendering engine through a coordinate transformation link that includes a local tangent plane coordinate system to obtain the local vertex data of the sub-tile; and to construct a transformation matrix based on the origin offset of the sub-tile and its parent tile in the local tangent plane coordinate system for use in parent-child hierarchical connection.

[0021] The single-unit model building module is used to batch generate B3DM tile files based on the local vertex data and transformation matrix of sub-tiles, and assign a batch ID attribute to each single-unit model under a given single-unit style, while storing the single-unit model metadata corresponding to the batch ID in the batch table.

[0022] Thirdly, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the method as described in the first aspect.

[0023] Fourthly, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in the first aspect.

[0024] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the method described in the first aspect.

[0025] The beneficial effects of this invention are as follows:

[0026] This invention presents a method and system for generating multi-level detail hierarchical single-unit models based on BeiDou grid codes, achieving automated and multi-precision reconstruction of 3D Tiles datasets from discrete BeiDou grid encoding. Utilizing the prefix characteristics of BeiDou grid codes, a multi-scale, multi-level LOD tree structure is constructed, solving the problem of traditional methods requiring pre-preparation of multi-level model data. Based on a batch partitioning strategy of "spatial sorting first, then parent-level grouping, and finally packaging by maximum number of units," the compactness of the LOD tree and the scheduling efficiency of the rendering engine are improved. Simultaneously, it supports batch generation of single-unit models, generating an independent Batch ID for each grid and storing metadata such as spatial range and level, enabling precise single-unit attribute querying. Combined with style parameters, the model's visual expressiveness is enhanced. This invention overcomes the bottleneck of traditional 3D model generation methods in reconstructing multi-scale, single-unit semantic models from discrete encoding, improving the automation level and spatial computation efficiency of BeiDou grid data from encoding to visualization. It has significant application value for rapid scene construction and intelligent decision-making in fields such as digital twin cities and low-altitude economy. Attached Figure Description

[0027] Figure 1 This is one of the flowcharts illustrating a method for generating a multi-level detail-level single-unit model based on BeiDou grid codes, provided in an embodiment of the present invention.

[0028] Figure 2 The second flowchart illustrates a method for generating multi-level detail-level single-unit models based on BeiDou grid codes, as provided in an embodiment of the present invention.

[0029] Figure 3 This refers to the BeiDou grid-coded data stored in the database used in this invention;

[0030] Figure 4 This is a visualization of the multilevel level of detail (LOD) and monolithic model of the present invention;

[0031] Figure 5 A schematic diagram of the structure of a multi-level detail hierarchy monolith model generation system based on BeiDou grid code provided in this embodiment of the invention;

[0032] Figure 6 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. 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.

[0034] like Figure 1 As shown, this embodiment of the invention provides a method for generating a multi-level detail-level single-unit model based on BeiDou grid codes, including the following steps:

[0035] S101: Obtain and parse the BeiDou grid coding information to obtain all maximum level grid codes;

[0036] S102: Utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, a hierarchical tile index file with a multi-scale LOD structure is constructed by deriving from the prefix of each maximum level grid code.

[0037] S103: For each sub-tile in the hierarchical tile index file, the WGS84 geodetic coordinates within the area covered by the sub-tile are transformed into coordinates adapted to the target 3D rendering engine through a coordinate transformation link that includes a local tangent plane coordinate system to obtain the local vertex data of the sub-tile; and a transformation matrix is ​​constructed based on the origin offset of the sub-tile and its parent tile in the local tangent plane coordinate system for use in parent-child hierarchical connection.

[0038] S104: Generate B3DM tile files in batches based on the local vertex data and transformation matrix of the sub-tiles, and assign a batch ID attribute to each individual model in the batch under a given individualization style, while storing the metadata of the individual model corresponding to the batch ID in the batch table.

[0039] This invention utilizes the hierarchical prefix inheritance characteristic of BeiDou grid codes to derive the Level of Detail (LOD) hierarchy. Without relying on additional spatial indexes or manual annotation, it automatically identifies the spatial scale and nesting relationships of each grid code directly from the encoding prefix, thus unifying the generation logic of multi-level detail models with the inherent structure of BeiDou grid codes. This mechanism allows massive amounts of discrete BeiDou grid codes to be automatically organized into 3D data units with a clear LOD hierarchy according to the hierarchical order implied by their encoding prefixes. This avoids the cumbersome steps of independently constructing multi-level indexes or performing geometric subdivisions required in traditional methods, significantly reducing the preprocessing complexity from raw data to 3D Tiles datasets. Furthermore, a coordinate transformation link including a local tangent plane coordinate system is introduced to accurately transform the original WGS84 geodetic coordinates into coordinates adapted to the target 3D rendering engine (such as Cesium). This ensures the numerical stability of large-scale geographic coordinate transformations and effectively reduces geometric deformation caused by ellipsoid curvature, thus providing a highly consistent coordinate reference for subsequent B3DM tile file generation. Ultimately, leveraging the aforementioned automated methods, standardized B3DM tile files can be generated in batches, forming an end-to-end automated pipeline from BeiDou grid codes to 3D Tiles datasets. For large-scale 3D models, this method achieves on-demand loading and progressive rendering: when the viewpoint is far away, only low-precision tiles corresponding to coarse-grained grid codes with shorter encoding prefixes are loaded; when the viewpoint is closer, it automatically switches to high-precision tiles corresponding to fine-grained grid codes with longer encoding prefixes. This LOD scheduling strategy, deeply bound to the BeiDou grid code hierarchy, fundamentally reduces the amount of data initially loaded, decreases memory usage and drawing calls during real-time rendering, and thus meets the stringent requirements for loading efficiency and rendering performance in large-scale 3D spatial information applications such as digital twin cities and low-altitude economy.

[0040] Based on the above embodiments, such as Figure 2As shown, this embodiment of the invention further provides a method for generating a multi-level detail-level single-unit model based on BeiDou grid codes, the steps of which are as follows:

[0041] S201, Beidou grid code reading and parsing.

[0042] Specifically, the system first reads the BeiDou grid encoding information in JSONL format from a specified path. It then parses and verifies the encoding length of each level of the three-dimensional BeiDou grid line by line, and determines the validity of the root node. The 2D code lengths and 3D code lengths for levels 1-10 are 4 and 7, 6 and 10, 7 and 12, 9 and 15, 11 and 18, 12 and 20, 14 and 23, 16 and 26, 18 and 29, and 20 and 32 bits, respectively. 2D code refers to two-dimensional grid code, and 3D code refers to three-dimensional grid code.

[0043] Based on this, the string structure is parsed to obtain the corresponding level, and the maximum level among all grid codes is recorded, providing a foundation for constructing multi-level detail (LOD). The processing flow is shown in Table 1.

[0044] Table 1

[0045]

[0046] The input data is a JSONL or plain text file (e.g., generated based on the "BeiDou Grid Location Code" (GB / T 39409-2020) standard) segmented. Figure 3 As shown in the diagram, the format is as follows: {"code":"N05H3A035"} or N05H3A035. Correctness is determined by the grid code length, and the decoding result is shown as {code, level, min: {lngDegree,latDegree, elevation}, max: {...}}. lngDegree represents longitude, latDegree represents latitude, and elevation represents altitude. If the input data volume is extremely large (e.g., tens of millions of grid codes), this step uses a streaming method to avoid loading all data into memory at once. This invention does not elaborate on the internal decoding process; the decoding output includes the WGS84 latitude, longitude, and altitude bounding boxes for each grid.

[0047] S202, Parameter settings in the single-model style selection.

[0048] Specifically, it is mainly used for the individual unit processing of BeiDou grid subdivision model encoding, including general parameters, mode parameters and individual unit style configuration parameters.

[0049] General parameters include input file path, output directory, maximum number of grids per B3DM file, and number of parallel worker threads, providing basic configuration and flexible adjustment space for model generation. The input file path specifies whether to use a JSONL file or a plain text file containing the data; the output directory defines where the processing results are saved; the maximum number of grids per B3DM file controls the subdivision level of the file, with a default value of 500; and the number of parallel worker threads sets the number of threads for parallel processing to improve efficiency, with a default value of 4. Additionally, built-in preset styles and custom styles can be included. The former provides various predefined appearance configurations; the latter allows users to specify personalized display styles via JSON files.

[0050] Mode parameters include whether to enable LOD level derivation, the minimum LOD level, and whether to enable individual unitization mode, which directly affect the display effect and performance optimization after model generation. Regarding "whether to enable LOD level derivation," this function is enabled by default and is used to dynamically adjust the model's level of detail based on viewpoint distance. The minimum LOD level sets the simplification level of the model at the farthest view distance, with a default value of 1. Regarding "whether to enable individual unitization mode," when enabled, this function generates a unique Batch Table and Batch ID attribute for each grid (or feature), supporting individual unitization attribute queries in visualization environments such as Cesium. This function is disabled by default.

[0051] Individual Model Style Settings: Primarily used for the visual representation of individual models. Style configuration controls the visual appearance of the generated model, including rendering mode (selectable as solid face or wireframe display), face color (hexadecimal string), transparency (value range 0~1), whether to use unlit materials, whether to display border lines, border line color, and border line width, etc., to meet different display needs and enhance the visual expressiveness of individual models. These style configuration parameters are written into the B3DM's glTF material, ensuring the model displays as expected in the 3D scene. In practical applications, double-sided rendering is achieved using the KHR_materials_unlit extension, suitable for scenes that do not require lighting calculations. When the transparency setting is less than 1, alphaMode must be configured to 'BLEND' to correctly handle transparency effects. Wireframe mode is enabled by setting the glTF primitive mode to 1 (LINES), suitable for displaying the model's structural framework. For border line rendering, it is recommended to use a separate primitive and offset it by 10cm along the U-axis of the ENU to avoid z-fighting and ensure the border lines are clearly visible.

[0052] S203, Level of Detail (LOD) Derivation.

[0053] Specifically, leveraging the hierarchical prefix inheritance characteristic of BeiDou grid codes, the LOD model can automatically derive all parent grid codes from fine-grained BeiDou grid codes, completing the generation of multi-scale, multi-level individual models. This process includes reading and processing JSONL data, deriving parent grid codes and deduplicating them, calculating global bounding boxes, grouping and packaging data by spatial location and parent to generate batches, and then constructing a nested tileset.json structure. This method ensures the compactness of the LOD tree and rendering efficiency, while improving computational efficiency based on the rules of BeiDou grid codes, requiring no additional data input. Specifically, as follows:

[0054] First, read the JSONL line by line, extract the valid grid codes and determine the maxLevel, obtaining all encoded information codes[]. Second, derive the parent level by prefix truncation, calculating L{N-1} / L{N-2} / ... / L{minLevel} for each L{N} code, while simultaneously deduplicating the parent codes to obtain a unique set of grid codes for each level, i.e., gridsByLevel = { level: [{code, min, max}]}. Next, calculate the global bounding box, i.e., calculate the perimeter and maximum and minimum height information by traversing all levels of grids. Then, sort by spatial location (longitude first, then latitude), and group by parent level (ensuring each batch is completely within a parent tile), and package by maxGridsPerTile to generate batchesByLevel = { level:[[grid, grid, ...], ...]}. Furthermore, generate a nested tileset.json, determining the parent-child relationship based on the global bounding box, nesting children level by level starting from L{minLevel}, and refining each tile: REPLACE, based on geometricError, completes the level classification standard based on visual experience; finally, B3DM is generated in parallel to ensure that each batch generates a corresponding B3DM file and completes the output of the results, namely L{minLevel} / 0000.b3dm, ..., L{maxLevel} / 0000.b3dm, ..., the LOD tileset.json structure is shown in Table 2;

[0055] Table 2

[0056]

[0057] Based on the hierarchical prefix inheritance characteristic of BeiDou grid codes, the first N characters of a child code are exactly its parent code. For example, a level 10 grid code is 32 characters long; the first 29 characters yield a level 9 parent code, the first 26 characters yield a level 8 parent code, and so on. The system first reads all grid codes to determine the maximum level (e.g., level 10). Then, for each grid code, the prefix is ​​extracted level by level from the maximum level downwards to generate the parent code for each level, and the `decodeGridCode` function is used to obtain the corresponding WGS84 bounding box (latitude, longitude, and elevation range). To prevent duplication, a dictionary (Map) is used to deduplicate the grid codes at each level, ultimately obtaining a unique set of grid codes and their spatial range for each level. This process does not require reading any external data and is completed entirely based on the encoding rules of BeiDou grid codes, improving computational efficiency.

[0058] When generating B3DM tiles, mesh packing and batch partitioning are used to combine multiple meshes into a single tile file to reduce the number of files and improve loading efficiency. To address the issue that simply following the reading order or random grouping can result in a tile containing spatially distant meshes, leading to an excessively large bounding box and compromising the spatial clustering of the Level of Dimension (LOD), a strategy of "spatial sorting first, then grouping by parent level, and finally packing by maximum number" is adopted. Specifically: For each level, all meshes at that level are first sorted according to the latitude and longitude of their center point (longitude first, then latitude), ensuring that spatially adjacent meshes are adjacent in the list. Then, for the Level 1 mesh (global scale), it is directly divided into groups of maxGridsPerTile. For levels 2 and above, each grid is first grouped according to its parent code (i.e., the code corresponding to the previous level) to ensure that grids under the same parent are grouped together as much as possible. Then, within each parent group, they are packaged according to maxGridsPerTile to ensure that the spatial range is strictly limited to the parent grid and has good spatial proximity, thereby ensuring the compactness of the LOD tree and rendering efficiency.

[0059] S204, Coordinate Transformation and Matrix Construction.

[0060] Specifically, coordinate transformation is crucial for ensuring the accurate spatial positioning of the 3D model. An algorithm achieves a high-precision end-to-end transformation from WGS84 geodetic coordinates to ECEF geocentric-fixed coordinates, then to ENU local tangent plane coordinates, and finally to glTF Y-Up coordinates. The 3D rendering engine in this embodiment is adapted to glTF Y-Up coordinates. By maintaining double-precision floating-point numbers in the calculation process, the precision loss caused by premature conversion to single-precision is avoided. The axis mapping step ensures the correct orientation of the model in 3D earth engines such as Cesium. The entire coordinate transformation process guarantees the spatial accuracy and high-precision display of the model. Specifically, as follows:

[0061] First, for each grid cell, WGS84 coordinates of the corner points are generated based on their minimum / maximum latitude / longitude and elevation. Second, the `wgs84ToEcef` function is called to convert them to ECEF coordinates. In this process, generating eight ECEF corner points converts the WGS84 bounding box (latitude / longitude + elevation range) of the BeiDou grid code into vertex coordinates of a 3D grid, providing the basic geometric data for subsequent coordinate transformations (ECEF→ENU→glTF Y-Up). Simultaneously, the connection order of the eight vertices is defined using triangle indexing, dividing the grid into 12 triangles (6 faces, 2 per face), and following the right-hand rule to ensure that the normals of each face face outwards. The abstract grid code is instantiated into a correctly rendered 3D box, supporting subsequent style settings (color, border, transparency) and single-unit processing. Next, the center point (ECEF coordinates) of the entire tile is calculated, along with the ENU basis vector at that center point. Then, the coordinates of each corner point in the ENU coordinate system are calculated using the ECEF coordinates and ENU basis vectors. Finally, glTF data was generated and axis mapping was performed to ensure the correct orientation of the model in 3D Earth engines such as Cesium. The WGS84 geodetic coordinate system uses (lat, lon, alt) to represent position (latitude, longitude, and ellipsoidal height); the ECEF geocentric Earth-fixed coordinate system uses (X, Y, Z) geocentric rectangular coordinates; and the ENU local tangent plane coordinate system uses (E, N, U) tangent plane rectangular coordinates (in meters) to describe position information. The coordinate transformation aims to convert from the WGS84 geodetic coordinate system (lat, lon, alt) to the ECEF geocentric Earth-fixed coordinate system (X, Y, Z) and then to the ENU local tangent plane coordinate system (E, N, U). This mapping ensures the correct orientation of the model in 3D Earth engines such as Cesium.

[0062] Matrix construction refers to the need to construct a transformation matrix for each child tile in LOD mode. This matrix transforms the ENU coordinates within the child tile to the ENU coordinate system of the parent tile, facilitating the connection between parent and child layers. Since the offset of the child tile center relative to the parent tile center can be directly calculated in the ENU coordinate system, this transformation matrix is ​​a pure translation matrix, requiring no rotation or scaling, simplifying calculations and improving numerical stability. Regarding ENU basis vector consistency, B3DM vertices use global ENU axes, and the offset of `tile.transform` also uses global ENU axis projection, ensuring a consistent transformation chain. This results in smoother transitions between different layers, improving overall rendering quality and user experience.

[0063] S205: Batch B3DM Generation and Monolithic Model Construction.

[0064] Specifically, after the coordinate transformation is completed, each batch (i.e., a set of meshes) will be converted into a B3DM tile file. The generation process includes: constructing glTF binary data (GLB), encapsulating the Feature Table and Batch Table, and writing it to a file. The construction of glTF requires generating vertex positions, normals, and indices, and rendering and visualizing the materials according to the individual style settings in step S202.

[0065] In this embodiment of the invention, planar normals are used, with one normal per triangle, shared by all vertices to ensure consistent face color and uniform lighting on the box surface. In batch-mode, a Batch ID attribute needs to be assigned to each triangle, and metadata (such as mesh code, bounding box, and hierarchy) for each mesh is stored in the Batch Table. Based on this, in rendering engines that support 3D tiles, such as Cesium, clicking on any mesh will retrieve its complete attribute information, achieving true individualization. The process for generating a single B3DM tile is shown in Table 3.

[0066] Table 3

[0067]

[0068] The handling of the border lines is somewhat special because border lines at the same depth as solid surfaces can cause Z-fighting (depth conflict). Therefore, by adding border lines and offsetting them along the U-axis (i.e., offsetting the border lines outward by 0.1 meters along the U-axis of the ENU coordinate system), we ensure that the borders are always visible and avoid the Z-fighting phenomenon. This step ensures that the generated 3D model has both rich visual effects and the ability to perform individual queries and attribute displays.

[0069] S206. Results Query, Retrieval, and Visualization.

[0070] Specifically, based on the B3DM tiles and tileset.json generated in step S205, the model is loaded and displayed in a 3D map platform (such as Cesium). Through interaction, the following operations are completed: visualization of models and grids in common formats such as GeoJSON, 3D map overlay, precise grid query, and export of search results (JSONL format). The visualization and query processes are highly responsive, supporting real-time retrieval of millions of grids, and meeting the application needs of large-scale 3D geographic information data. Specific results are as follows: Figure 4 As shown.

[0071] Based on the same inventive concept, such as Figure 5As shown, this embodiment of the invention also provides a multi-level detail level single-unit model generation system based on BeiDou grid code, characterized in that it includes: a data parsing module, a LOD level derivation module, a coordinate transformation module, and a single-unit model construction module.

[0072] Specifically, the data parsing module is used to obtain and parse the BeiDou grid coding information to obtain all maximum level grid codes; the LOD level derivation module is used to utilize the hierarchical prefix inheritance characteristic of the BeiDou grid codes to derive a hierarchical tile index file with a multi-scale LOD structure based on the prefix of each maximum level grid code; the coordinate transformation module is used to transform the WGS84 geodetic coordinates within the coverage area of ​​each sub-tile in the hierarchical tile index file into coordinates adapted to the target 3D rendering engine through a coordinate transformation link including a local tangent plane coordinate system to obtain the local vertex data of the sub-tile; and to construct a transformation matrix based on the origin offset of the sub-tile and its parent tile in the local tangent plane coordinate system for parent-child hierarchical connection; the individual model construction module is used to batch generate B3DM tile files based on the local vertex data and transformation matrix of the sub-tiles, and assign a batch ID attribute to each individual model in the given individualization style, while storing the individual model metadata corresponding to the batch ID in the batch table.

[0073] It should be noted that the multi-level detail level single-unit model generation system provided in this embodiment of the invention is for implementing the above method. Its specific functions can be referred to in the above method embodiments, and will not be repeated here.

[0074] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 6As shown, the electronic device may include: a processor 601, a communication interface 602, a memory 603, and a communication bus 604, wherein the processor 601, the communication interface 602, and the memory 603 communicate with each other through the communication bus 604. The processor 601 can call logical instructions in the memory 603 to generate a multi-level detail level single-unit model based on BeiDou grid codes. The method includes: acquiring and parsing BeiDou grid coding information to obtain all maximum level grid codes; utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, deriving based on the prefix of each maximum level grid code to construct a hierarchical tile index file with a multi-scale LOD structure; for each sub-tile in the hierarchical tile index file, transforming the WGS84 geodetic coordinates within the coverage area of ​​the sub-tile into coordinates adapted to the target 3D rendering engine through a coordinate transformation link including a local tangent plane coordinate system to obtain the local vertex data of the sub-tile; constructing a transformation matrix based on the origin offset of the sub-tile and its parent tile in the local tangent plane coordinate system for parent-child hierarchical connection; batch generating B3DM tile files based on the local vertex data and transformation matrix, assigning a batch ID attribute to each single-unit model in a given single-unitization style, and storing the single-unit model metadata corresponding to the batch ID in a batch table.

[0075] Furthermore, when the logical instructions in the aforementioned memory 603 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0076] This invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by a computer, the computer can execute a method for generating a multi-level detail-level monolithic model based on BeiDou grid codes provided in the above-described method embodiments.

[0077] This invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for generating a multi-level detail-level monolithic model based on BeiDou grid codes provided in the above-described method embodiments.

[0078] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0079] 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 them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating multi-level detail-level single-unit models based on BeiDou grid codes, characterized in that, include: Obtain and parse the BeiDou grid coding information to obtain all maximum level grid codes; By utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, a hierarchical tile index file with a multi-scale LOD structure is constructed based on the prefix of each maximum level grid code. For each sub-tile in the hierarchical tile index file, the WGS84 geodetic coordinates within the area covered by the sub-tile are transformed into coordinates adapted to the target 3D rendering engine through a coordinate transformation link that includes a local tangent plane coordinate system to obtain the local vertex data of the sub-tile; and a transformation matrix is ​​constructed based on the origin offset of the sub-tile and its parent tile in the local tangent plane coordinate system for use in parent-child hierarchical connection. B3DM tile files are generated in batches based on the local vertex data and transformation matrix of the sub-tiles. Each individual model is assigned a batch ID attribute under a given individualization style, and the metadata of the individual model corresponding to the batch ID is stored in the batch table.

2. The method for generating a multi-level detail-level single-unit model based on BeiDou grid code according to claim 1, characterized in that, Utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, a hierarchical tile index file with a multi-scale LOD structure is constructed based on the prefix of each maximum-level grid code, including: Utilizing the hierarchical prefix inheritance characteristic of BeiDou grid codes, prefixes are extracted from each maximum-level grid code to derive the parent grid codes for each level. After deduplication, a unique set of grid codes for each level is obtained, and a global bounding box is calculated based on the spatial range corresponding to each grid code. Then, the grid codes within each level are sorted according to their spatial location. Based on the global bounding box, the parent-child relationship is determined to group and package all grid codes belonging to the same parent grid, forming batch data merged by parent level. Finally, based on the nesting relationship between the batch data and the levels, a hierarchical tile index file with a multi-scale LOD structure is generated.

3. The method for generating a multi-level detail-level single-unit model based on BeiDou grid code according to claim 1, characterized in that, The process of transforming the WGS84 geodetic coordinates within the area covered by the sub-tile into coordinates compatible with the target 3D rendering engine via a coordinate transformation link that includes a local tangent plane coordinate system includes: The WGS84 geodetic coordinates within the area covered by the sub-tile are sequentially converted into geocentric and geofixed coordinates and ENU tangent plane coordinates based on the local origin of the sub-tile. Then, the ENU coordinate axes are mapped to coordinates adapted to the target 3D rendering engine to obtain the local vertex data of the sub-tile.

4. The method for generating a multi-level detail-level single-unit model based on BeiDou grid code according to claim 1, characterized in that, Before performing coordinate transformation, configuration of general parameters and mode parameters is also included; wherein, the general parameters include at least one of the following: input file path, output directory, maximum number of grids contained in each B3DM file, and number of parallel worker threads; the mode parameters include at least one of the following: LOD level derivation enable flag, LOD minimum level, and singleton mode enable flag.

5. The method for generating a multi-level detail-level single-unit model based on BeiDou grid code according to claim 1, characterized in that, The configuration information for the individual style includes at least one of the following options: solid face or wireframe display mode, face color, transparency, whether to use a non-lighting material, whether to display border lines, border line color, and border line width.

6. A multi-level detail hierarchy single-unit model generation system based on BeiDou grid code, characterized in that, include: The data parsing module is used to obtain and parse the BeiDou grid coding information to obtain all the maximum level grid codes; The LOD level derivation module is used to utilize the hierarchical prefix inheritance characteristics of Beidou grid codes to derive and construct a hierarchical tile index file with a multi-scale LOD structure based on the prefix of each maximum level grid code. The coordinate transformation module is used to transform the WGS84 geodetic coordinates within the coverage area of ​​each sub-tile in the hierarchical tile index file into coordinates adapted to the target 3D rendering engine through a coordinate transformation link that includes a local tangent plane coordinate system to obtain the local vertex data of the sub-tile; and to construct a transformation matrix based on the origin offset of the sub-tile and its parent tile in the local tangent plane coordinate system for use in parent-child hierarchical connection. The single-unit model building module is used to batch generate B3DM tile files based on the local vertex data and transformation matrix of sub-tiles, and assign a batch ID attribute to each single-unit model under a given single-unit style, while storing the single-unit model metadata corresponding to the batch ID in the batch table.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

Images