Live-action three-dimensional model construction method, system and equipment based on digital city

Abstract

The invention discloses a live-action three-dimensional model construction method, system and device based on a digital city, and the method comprises the steps: obtaining the multi-source sensing data of a city, and carrying out the data preprocessing, and the multi-source sensing data comprises city image data, point cloud data and GIS data; converting the preprocessed multi-source sensing data in different formats into a preset standard format, and performing registration and fusion of the multi-source sensing data to obtain fused multi-source data; performing data optimization on the fused multi-source data, and constructing a city-level three-dimensional real scene model according to the optimized fused multi-source data and a preset three-dimensional space analysis tool; the preset three-dimensional space analysis tool provides space analysis and three-dimensional space visualization functions; according to the invention, efficient and accurate real-scene three-dimensional data of a city is provided, and three-dimensional and visual space information is provided.

Description

Technical Field

[0001] The present application relates to the field of digital city technology, and in particular to a method, system and device for constructing a real-scene three-dimensional model based on a digital city. Background Art

[0002] With the continuous advancement of information technology, digital cities have been widely used around the world as a modern means of urban planning, management and services. Most of the existing digital city technologies are based on two-dimensional map data. Although they meet the visualization needs of urban information to a certain extent, with the acceleration of urbanization and the increasing complexity of urban management needs, the limitations of two-dimensional map data are gradually emerging: it cannot provide three-dimensional and intuitive spatial information, which limits the depth and accuracy of urban planning and management. In addition, due to the limitation of data dimension, two-dimensional map data cannot fully express the complexity of urban space, especially in scenes such as high-rise buildings and complex terrain.

[0003] Based on this, how to provide efficient and accurate real-life three-dimensional data of the city has practical application value and significance. Summary of the invention

[0004] In order to provide efficient and accurate real-life three-dimensional data of a city and provide stereoscopic and intuitive spatial information, the present application provides a method, system and device for constructing a real-life three-dimensional model based on a digital city.

[0005] In the first aspect, the invention objective of the present application is achieved by adopting the following technical solutions: The method for constructing a real-scene three-dimensional model based on a digital city includes: Acquire multi-source sensor data of the city and perform data preprocessing, wherein the multi-source sensor data includes city image data, point cloud data and GIS data; convert the preprocessed multi-source sensor data in different formats into a preset standard format and perform multi-source sensor data registration and fusion to obtain fused multi-source data; The fused multi-source data is optimized, and a city-level three-dimensional real scene model is constructed based on the optimized fused multi-source data and a preset three-dimensional space analysis tool; the preset three-dimensional space analysis tool provides space analysis and three-dimensional space visualization functions.

[0006] The steps of constructing the city-level three-dimensional real scene model include: Terrain modeling, building modeling, texture mapping, road and vegetation modeling operations are performed based on the city image data, point cloud data and GIS data to obtain the corresponding city initial model and model integration elements; the city initial model is optimized, and model integration is performed based on all the corresponding model integration elements to obtain the city-level three-dimensional real-scene model.

[0007] By adopting the above technical scheme, in order to provide efficient and accurate real-life three-dimensional data of the city and provide three-dimensional and intuitive spatial information, the present application optimizes and efficiently processes the generation and analysis of real-life three-dimensional data; specifically, in the stage of generating real-life three-dimensional data: the present application collects and extracts different types of sensor raw data through preset sensor equipment in the city (such as lidar, panoramic camera, drone, etc.), that is, obtains multi-source sensor data of the city. In actual application, aerial photogrammetry or satellite remote sensing can be used to obtain city image data, LiDAR scanning technology can be used to obtain point cloud data, and geographic information system (GIS) data can be collected to obtain GIS data of the target city; the present invention performs data preprocessing (including data cleaning) on ​​multi-source sensor data, such as image matching and aerial triangulation of city image data, and point cloud The data is filtered and classified, and the GIS data is cleaned and structured and other pre-processing operations are performed to obtain multi-source sensor data with higher accuracy; then, in order to facilitate the subsequent visualization of the real-life 3D data in the intelligent terminal or platform in the 3D space and improve the consistency and interoperability of the data, the original multi-source sensor data in different data formats are uniformly converted (where the preset standard format is the platform standard format of the specified 3D model display terminal) to obtain multi-source sensor data in a standard format, and then the multi-source sensor data of the digital city after the conversion is registered and fused. The data registration and fusion process includes the precise fusion of image data and point cloud data, and then data optimization is performed on the obtained fused multi-source data, such as mesh simplification, texture compression, etc., so that the real-life 3D data can achieve a balance between visual quality and storage efficiency.

[0008] In the spatial data analysis stage: This application provides rich spatial analysis and visualization functions for city-level three-dimensional models through three-dimensional spatial analysis tools. Spatial analysis includes distance measurement, area calculation, volume estimation, etc. In actual application, combined with visualization tools, it can provide users with a smooth three-dimensional interactive experience. Through terrain modeling, building modeling, texture mapping, road and vegetation modeling operations, it provides efficient and accurate real-life three-dimensional data of three-dimensional spatial models. Therefore, this application achieves the purpose of providing efficient and accurate real-life three-dimensional data of cities and providing three-dimensional and intuitive spatial information.

[0009] In a preferred example of the present application: the method further comprises constructing a component-level three-dimensional reality model; The steps of constructing the component-level three-dimensional real scene model include: Based on the determined construction parameter requirements and construction accuracy of the target component, the corresponding point cloud data is obtained and processed, and the surface mesh is reconstructed; Performing detailed modeling, adding textures, and setting material effects for the target component using detailed modeling software, wherein the detailed modeling includes adding fine structures and / or modeling small components for the target component to obtain a component-level three-dimensional real-scene model; The component-level three-dimensional real scene model is optimized in terms of details and performance in a balanced manner according to the construction parameter requirements and construction accuracy, and a multi-precision version of the component-level three-dimensional real scene model is created.

[0010] By adopting the above technical solution, the component-level 3D real-scene model is a high-precision 3D model based on urban components (such as doors and windows, bridges, street lights, road obstacles, etc.), which can reflect the detailed size structure and appearance characteristics of the components; the component parameter requirements are parameters such as model resolution and model material requirements, and the component accuracy represents the degree of refinement of the model in terms of geometry, texture and material, which determines the realism and scope of application of the model; the power supply data is a set of 3D coordinate points obtained by means of laser scanning and other means, and the point cloud data processing includes operations such as denoising, filtering, and registration to improve data quality and the accuracy of subsequent 3D modeling; the surface mesh is the conversion of pre-processed power supply data into a continuous surface mesh to form a preliminary 3D model; the detail modeling software includes CAD software or 3D modeling software, and the detail modeling software is a software tool used to perform fine modeling of the preliminary 3D model (such as 3ds Max, Blender, Maya, etc.); specifically, detail modeling is to add more sophisticated structures and small components on the basis of the preliminary model, texture and material effects are to add real texture maps and material settings to the model to make it more visually realistic, and balanced optimization is to optimize the performance of the model while maintaining the details of the model to ensure that the component-level 3D model can be efficient and practical in different application scenarios or smart terminals under different bandwidth environments. At the same time, the present invention creates model versions with different accuracy levels to adapt to different application requirements and computing resources, which is conducive to improving the applicability of the model and enhancing the scene applicability performance.

[0011] In a preferred example of the present application: the three-dimensional spatial analysis tool includes a visual area analysis algorithm, and the method includes: Perform data preprocessing on the acquired three-dimensional scene data and construct an octree spatial index structure; Based on the constructed octree spatial index structure, starting from a preset observation point, rays are projected along multiple different directions to obtain ray projection paths; the rays are used to simulate line of sight propagation; Based on the ray projection path, traverse the tree nodes of the octree, determine and obtain the intersection point between the ray and the object in the scene of the three-dimensional scene data, and obtain the visible area starting from the preset observation point; and / or, The visual area analysis algorithm introduces a multi-threaded parallel computing algorithm and a GPU acceleration algorithm. The multi-threaded parallel computing algorithm processes the interaction between rays and nodes in parallel according to the hierarchical structure of the octree, and the GPU acceleration algorithm performs the calculation tasks of ray projection and intersection in parallel.

[0012] By adopting the above technical solutions, the octree spatial index structure is a data structure for efficiently managing and querying three-dimensional spatial data. By recursively dividing the space into eight subspaces until a preset granularity level is reached, it helps to quickly locate and retrieve data in a specific area; the visual area analysis algorithm is an algorithm for calculating the area that can be directly seen by a given observation point; the multi-threaded parallel computing algorithm speeds up the computing speed by executing multiple threads on multiple processor cores at the same time; the GPU acceleration algorithm utilizes the powerful parallel computing capability of the graphics processing unit (GPU) to accelerate specific types of data processing tasks to improve the efficiency of graphics and image processing of large amounts of data; specifically, the present invention can quickly locate and retrieve big data in a specific area by constructing an octree spatial index structure, thereby improving data processing efficiency, and greatly improves the efficiency of ray casting and intersection calculation through multi-threaded parallel computing algorithms and GPU acceleration algorithms. Through precise ray casting and intersection calculation, the visible area can be accurately simulated and determined, providing a scientific basis for decision-making in the fields of urban planning and architectural design.

[0013] In a preferred example of the present application: the three-dimensional space analysis tool includes a space path planning algorithm, and the method includes: Discretize the three-dimensional space of the designated area where the city-level three-dimensional real scene model is located into grid units, and calculate and associate a corresponding travel cost for each grid unit in combination with preset terrain constraints; Based on the gridded three-dimensional space and the travel cost of the grid cells, a preset heuristic search algorithm is used to perform heuristic path search, and the heuristic function value of each grid cell is calculated, and the search range is gradually expanded until the optimal path from the starting point to the target point is found; and / or, During the path planning process, the spatial path planning algorithm combines the dynamic obstacle avoidance mechanism to detect dynamic obstacles in the scene in real time, and calculates the potential impact of dynamic obstacles on the current path based on the position and movement trend of the dynamic obstacles; When a dynamic obstacle is detected that may block the current path, a real-time path replanning instruction is triggered. Based on the current position, target position and updated obstacle information, the preset heuristic search algorithm is used again to search the path and obtain a new optimal path.

[0014] By adopting the above technical solution, the grid unit is a plurality of cubic or hexahedral units after discretizing the three-dimensional space, the passage cost is the passing difficulty value calculated for each grid unit according to factors such as terrain and obstacles, and the heuristic search algorithm includes the A* algorithm when evaluating the path planning; the terrain constraint condition is a constraint condition pre-set according to terrain characteristics, road type and obstacles, etc.; the heuristic function value is used to evaluate the potential value of each grid unit, which includes the actual cost from the starting point to the current unit and the estimated cost from the current unit to the end point; the present invention is provided with a spatial path planning algorithm and a dynamic obstacle avoidance mechanism, wherein the dynamic obstacle avoidance mechanism detects and moves obstacles in the corresponding scene in real time, which is conducive to ensuring the safety and feasibility of the path; by discretizing the three-dimensional space into grid units, the terrain and obstacles can be represented more accurately, and the accuracy of path planning can be improved; the passage cost calculation fully considers the influence of actual environmental factors on path planning, and guides the search direction by using the heuristic search algorithm, which significantly improves the search efficiency, and the dynamic obstacle avoidance mechanism enhances the safety and real-time performance of the planned path.

[0015] In a preferred example of the present application: the method further includes: The spatial path planning algorithm is combined with a multi-objective optimization algorithm. During the path planning process, a custom path planning strategy is generated based on the path planning requirements input by the user; the custom path planning strategy calculates the weighted coefficients of the objective functions of each optimization goal according to the different optimization goals obtained, and evaluates and determines the final path planning solution; the final path planning solution includes the optimal path corresponding to different optimization goals, and the different optimization goals include but are not limited to the shortest distance, the shortest time and the lowest energy consumption; or, The method comprises: Distributed file repository for storing raw multi-source sensor data, large model files, and custom spatial index structures; Based on the hybrid database strategy, the spatial data is stored in the preset spatial database, while the metadata and attribute data are stored in the preset relational database; Based on the preset distributed cache system, frequently accessed data and query results are cached in the memory. According to the access frequency and importance of the data, an intelligent data tiered storage strategy is executed to automatically migrate data between high-speed storage and large-capacity storage.

[0016] By adopting the above technical solution, the multi-objective optimization algorithm is an algorithm that considers multiple optimization objectives (such as distance, time, energy consumption, etc.) at the same time when planning a path, and calculates the comprehensive optimal path through weighted coefficients; the weighted coefficients of the objective function are used to evaluate the weight values ​​of the importance of different optimization objectives, and the comprehensive objective function value is calculated through the weighted coefficients, so as to facilitate the generation of personalized path planning strategies according to different user needs, meet the specific needs of different users, and improve the reliability and comprehensive performance of path planning; further, in terms of data management of real-life three-dimensional data, the present invention is also provided with a distributed file library and a custom spatial index structure. When storing data, a distributed file system (such as HDF S) stores original multi-source sensor data and large model files to ensure high availability and scalability of data, stores spatial data in spatial databases (such as PostGIS) to support complex geographic spatial queries, and stores metadata and attribute data in relational databases (such as PostgreSQL) to support efficient structured data management; the hybrid database strategy of the present invention optimizes the storage and management of different types of data, improves the efficiency of data query and processing, supports multiple data access methods, meets the needs of different application scenarios, and can also apply intelligent data tiered storage strategies to reasonably utilize storage resources, reduce storage costs, improve storage efficiency, and ensure the long-term stable operation of the system.

[0017] In a preferred example of the present application: the three-dimensional spatial analysis tool includes a terrain profile analysis algorithm, The terrain modeling includes generating a digital elevation model based on GIS data and creating terrain textures; The terrain profile analysis algorithm includes: Extracting elevation data points along a specified path from the digital elevation model; According to the complexity of terrain changes, an adaptive sampling algorithm is used to determine the sampling density and sampling points, and the elevation data points on the path are sampled to obtain a series of discrete elevation values; According to the elevation values ​​between adjacent sampling points, the bilinear interpolation algorithm is used to calculate the elevation between adjacent sampling points to obtain a continuous and smooth terrain profile curve; Based on a preset geological database, the stratigraphic structure and lithology distribution information of the corresponding position in the profile is determined and integrated with the topographic profile curve to obtain a complete topographic profile containing the stratigraphic structure and lithology distribution information.

[0018] By adopting the above technical scheme, the digital elevation model (DEM) is a three-dimensional terrain model generated based on geographic information system (GIS) data, which represents the elevation information of the earth's surface. The terrain profile analysis algorithm is an algorithm for generating a terrain profile on a specified path; the bilinear interpolation algorithm is an interpolation algorithm for estimating the value of an unknown point between two known points, and is often used to generate a smooth terrain profile curve; the present invention extracts high-precision elevation data points and dynamically adjusts the density to improve the calculation efficiency while ensuring the sampling accuracy. By using the bilinear interpolation algorithm, a continuous and smooth terrain profile curve is generated, the accuracy and visual effect of the terrain profile are improved, and at the same time, by integrating the underlying structure and lithology distribution information in the geological database, a more comprehensive geological analysis effect is improved.

[0019] In a preferred example of the present application: the method further includes: Based on the city-level 3D real scene model and the component-level 3D real scene model, the building detail information inside the building is integrated to construct a comprehensive rescue environment model; wherein the building detail information includes internal passages, stairs and emergency exits; In the comprehensive rescue environment model, an optimized heuristic search algorithm is used to calculate the optimal rescue route based on real-time traffic data and accident information; the accident information includes a fire spread path predicted based on building structure and building material information; For the interior of buildings, a three-dimensional path planning algorithm is used to combine the complex factors of the vertical and horizontal movements of personnel to generate an approach route that meets rescue needs; and targeted route plans are formulated based on the characteristics of different types of rescue equipment.

[0020] By adopting the above technical scheme, a comprehensive rescue environment model is formed, and the city-level real-life 3D model is integrated with the component-level real-life 3D model, which includes detailed information on the interior of the building and is suitable for rescue planning and decision-making. The real-time traffic data is road traffic status data acquired in real time; the accident information includes detailed information on difficult accidents such as fire and collapse, and improved heuristic search algorithms such as the A* algorithm; the approach route that meets the rescue needs is the safest and fastest approach route. The present invention combines the optimized heuristic search algorithm and the three-dimensional path planning algorithm, and can specify targeted route plans according to the characteristics of different types of rescue equipment, thereby improving the travel efficiency and safety of the rescue equipment, and is conducive to providing comprehensive information support for rescue decisions, and improving the scientificity and effectiveness of rescue operations.

[0021] In the second aspect, the invention objective of the present application is achieved by adopting the following technical solutions: A real-scene 3D model construction system based on a digital city, used to execute the real-scene 3D model construction method based on a digital city as described above, the system comprising: Data acquisition module, used to obtain multi-source sensor data of the city; A data processing terminal is used to pre-process the multi-source sensor data of the city, wherein the multi-source sensor data includes city image data, point cloud data and GIS data; convert the pre-processed multi-source sensor data in different formats into a preset standard format and perform multi-source sensor data registration and fusion to obtain fused multi-source data; The fused multi-source data is optimized, and a city-level 3D real scene model is constructed according to the optimized fused multi-source data and a preset 3D space analysis tool; the preset 3D space analysis tool provides space analysis and 3D space visualization functions; wherein the step of constructing the city-level 3D real scene model includes performing terrain modeling, building modeling, texture mapping, road and vegetation modeling operations based on the city image data, point cloud data and GIS data to obtain a corresponding city initial model and model integration elements; the city initial model is optimized, and model integration is performed based on all corresponding model integration elements to obtain the city-level 3D real scene model; A visualization interactive tool is used to display the city-level three-dimensional real-scene model and perform three-dimensional interaction.

[0022] By adopting the above technical scheme, in order to provide efficient and accurate real-life three-dimensional data of the city and provide three-dimensional and intuitive spatial information, the present application optimizes and efficiently processes the generation and analysis of real-life three-dimensional data; specifically, in the stage of generating real-life three-dimensional data: the present application collects and extracts different types of sensor raw data through preset sensor equipment in the city (such as lidar, panoramic camera, drone, etc.), that is, obtains multi-source sensor data of the city. In practical applications, the city image data can be obtained through aerial photogrammetry or satellite remote sensing, point cloud data can be obtained through LiDAR scanning technology, and geographic information system (GIS) data can be collected to obtain GIS data of the target city; the present invention performs data preprocessing (including data cleaning) on ​​the multi-source sensor data, such as image matching and aerial triangulation of the city image data, filtering and classification of the point cloud data, cleaning and structuring of the GIS data and other preprocessing operations to obtain multi-source sensor data with higher accuracy; then, in order to facilitate the subsequent visualization of the real-life three-dimensional data in the intelligent terminal or platform for three-dimensional spatial display, and improve the consistency and interoperability of the data, The original multi-source sensor data of different data formats are uniformly converted (where the preset standard format is the platform standard format of the specified three-dimensional model display terminal) to obtain multi-source sensor data in a standard format, and then the multi-source sensor data of the digital city after the conversion format is data aligned and fused. The data alignment and fusion process includes the precise fusion of image data and point cloud data, and then the obtained fused multi-source data is optimized, such as mesh simplification, texture compression, etc., so that the real-life three-dimensional data strikes a balance between visual quality and storage efficiency; in the spatial data analysis stage: the present application provides rich spatial analysis and visualization functions for the city-level three-dimensional model through a three-dimensional spatial analysis tool. The spatial analysis includes distance measurement, area calculation, volume estimation, etc. In actual application, combined with the visualization tool, it can provide users with a smooth three-dimensional interactive experience. Through terrain modeling, building modeling, texture mapping, road and vegetation modeling operations, it provides efficient and accurate real-life three-dimensional data of the three-dimensional space model, so that the present application achieves the purpose of providing efficient and accurate real-life three-dimensional data of the city and providing three-dimensional and intuitive spatial information.

[0023] In the third aspect, the invention objective of the present application is achieved by adopting the following technical solutions: A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for constructing a real-life three-dimensional model based on a digital city are implemented.

[0024] In a fourth aspect, the invention objective of the present application is achieved by adopting the following technical solutions: A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for constructing a real-scene three-dimensional model based on a digital city.

[0025] In summary, the present application includes at least one of the following beneficial technical effects: 1. Collect and extract different types of sensor raw data through preset sensor equipment in the city (such as laser radar, panoramic camera, drone, etc.), that is, obtain multi-source sensor data of the city. In practical application, aerial photogrammetry or satellite remote sensing can be used to obtain city image data, LiDAR scanning technology can be used to obtain point cloud data, and geographic information system (GIS) data can be collected to obtain GIS data of the target city; the present invention performs data preprocessing (including data cleaning) on ​​multi-source sensor data, such as image matching and aerial triangulation of city image data, filtering and classification of point cloud data, cleaning and structuring of GIS data, and other preprocessing operations to obtain multi-source sensor data with higher accuracy; then, in order to facilitate the subsequent visualization of real-scene three-dimensional data in an intelligent terminal or platform for three-dimensional spatial display, and to improve the consistency and interoperability of data, the original multi-source sensor data of different data formats are uniformly converted (where the preset standard format is a specified three-dimensional model display terminal The method comprises the following steps: first, converting the multi-source sensor data of the digital city into a standard format, and then performing data registration and fusion on the multi-source sensor data of the digital city after the conversion. The data registration and fusion process includes the precise fusion of image data and point cloud data. Then, data optimization is performed on the obtained fused multi-source data, such as mesh simplification, texture compression, etc., so that the real-life three-dimensional data can achieve a balance between visual quality and storage efficiency. In the spatial data analysis stage: the present application provides a rich spatial analysis and visualization function for the city-level three-dimensional model through a three-dimensional spatial analysis tool. The spatial analysis includes distance measurement, area calculation, volume estimation, etc. In practical applications, the combination of the visualization tool can provide users with a smooth three-dimensional interactive experience. Through terrain modeling, building modeling, texture mapping, road and vegetation modeling operations, efficient and accurate real-life three-dimensional data of the three-dimensional spatial model is provided. Thus, the present application achieves the purpose of providing efficient and accurate real-life three-dimensional data of the city and providing three-dimensional and intuitive spatial information. 2. The multi-objective optimization algorithm is an algorithm that considers multiple optimization objectives (such as distance, time, energy consumption, etc.) at the same time during path planning, and calculates the comprehensive optimal path through weighted coefficients; the weighted coefficients of the objective function are used to evaluate the weight values ​​of the importance of different optimization objectives, and the comprehensive objective function value is calculated through the weighted coefficients, which is conducive to generating personalized path planning strategies according to different user needs, meeting the specific needs of different users, and improving the reliability and comprehensive performance of path planning; further, in terms of data management of real-life three-dimensional data, the present invention is also provided with a distributed file library and a custom spatial index structure. When storing data, a distributed file system (such as HDFS) is used to store the original Multi-source sensor data and large model files ensure high availability and scalability of data, store spatial data in spatial databases (such as PostGIS), support complex geographic spatial queries, store metadata and attribute data in relational databases (such as PostgreSQL), and support efficient structured data management; the hybrid database strategy of the present invention optimizes the storage and management of different types of data, improves the efficiency of data query and processing, supports multiple data access methods, meets the needs of different application scenarios, and can also apply intelligent data tiered storage strategies to reasonably utilize storage resources, reduce storage costs, improve storage efficiency, and ensure the long-term stable operation of the system; 3. Comprehensive rescue environment model, which integrates the city-level real-life 3D model with the component-level real-life 3D model, contains detailed information about the interior of the building, and is suitable for rescue planning and decision-making. Real-time traffic data is road traffic status data acquired in real time; accident information includes detailed information about difficult accidents such as fire and collapse, and improved heuristic search algorithms such as the A* algorithm; the approach route that meets rescue needs is the safest and fastest approach route. The present invention combines the optimized heuristic search algorithm and the 3D path planning algorithm, and can specify targeted route plans according to the characteristics of different types of rescue equipment, thereby improving the travel efficiency and safety of rescue equipment, and is conducive to providing comprehensive information support for rescue decisions, and improving the scientificity and effectiveness of rescue operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a method for constructing a real-scene three-dimensional model based on a digital city in one embodiment of the present application; Figure 2 This is a platform application interface diagram of a method for constructing a real-scene three-dimensional model of a digital city in one embodiment of the present application; Figure 3 is another flow chart of a method for constructing a real-scene three-dimensional model based on a digital city in one embodiment of the present application; Figure 4 It is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The present application is further described in detail below in conjunction with the accompanying drawings.

[0028] In one embodiment, if Figure 1 As shown, the present application discloses a method for constructing a real-scene three-dimensional model based on a digital city, which specifically includes the following steps: S1: Acquire multi-source sensor data of the city and perform data preprocessing. The multi-source sensor data includes urban image data, point cloud data and GIS data; convert the preprocessed multi-source sensor data in different formats into a preset standard format and perform multi-source sensor data registration and fusion to obtain fused multi-source data.

[0029] In this embodiment, data preprocessing includes data cleaning, data geometry correction, data radiation correction and data format conversion; obtaining urban image data through satellite remote sensing, aerial photography and other means, using laser scanners and other equipment to obtain urban point cloud data, and obtaining urban terrain, roads, buildings and other geographic information data from the geographic information system; data registration is to align data from different sources using a registration algorithm (such as the ICP algorithm), and data fusion is to merge the aligned multi-source data.

[0030] S2: Optimize the fused multi-source data, and build a city-level 3D real scene model based on the optimized fused multi-source data and the preset 3D spatial analysis tool; the preset 3D spatial analysis tool provides spatial analysis and 3D spatial visualization functions.

[0031] The steps to construct a city-level 3D reality model include: Based on the city image data, point cloud data and GIS data, terrain modeling, building modeling, texture mapping, road and vegetation modeling operations are carried out to obtain the corresponding city initial model and model integration elements; the city initial model is optimized, and the model is integrated based on all the corresponding model integration elements to obtain a city-level three-dimensional real-scene model.

[0032] Specifically, data optimization includes data compression and data noise reduction. By optimizing the fusion of multi-source data, the data quality and processing efficiency are improved. The city-level three-dimensional real-scene model covers a high-precision three-dimensional model of the entire city or a specific area, including buildings, roads, terrain and other elements.

[0033] In this embodiment, the city-level three-dimensional real-scene model is designed in a real-scene three-dimensional data platform. In the real-scene three-dimensional data platform, in terms of real-scene three-dimensional data management, it also includes a distributed file library (referring to a distributed file system, such as HDFS) for storing original multi-source sensor data and large model files, and a custom spatial index structure, such as an improved R-tree and a quadtree, to accelerate spatial queries; based on a hybrid database strategy, the spatial data is stored in a preset spatial database, and the metadata and attribute data are stored in a preset relational database.

[0034] Based on the preset distributed cache system (such as Redis), frequently accessed data and query results are cached in the memory. According to the access frequency and importance of the data, an intelligent data tiered storage strategy is executed to automatically migrate data between high-speed storage (such as SSD) and large-capacity storage (such as HDD). In order to support data version control and collaborative editing, a 3D data version management system based on Git is also designed in the real-life 3D data platform, allowing multiple users to edit and update models at the same time.

[0035] Specifically, in the real-life 3D data platform, machine learning and deep learning algorithms are also integrated for scene understanding and predictive analysis. For example, we use convolutional neural networks for building semantic segmentation to identify the structure and function of buildings; and use time series analysis to predict urban development trends. In order to support real-time analysis of large-scale data, the real-life 3D data platform uses a distributed computing framework (such as Apache Spark) to decompose complex analysis tasks into multiple computing nodes for parallel execution; and develops a wealth of visualization tools, including heat maps, contour maps, 3D bar graphs, etc., to intuitively display the analysis results; these visualization tools are based on WebGL technology and can achieve a smooth 3D interactive experience in a Web browser, which is conducive to providing a user-friendly operating interface, supporting the visualization of 3D data and the application of spatial analysis tools, such as 3D data display functions. Figure 2 shown.

[0036] In one embodiment, a method for constructing a real-scene 3D model based on a digital city includes constructing a component-level 3D real-scene model; the steps of constructing the component-level 3D real-scene model include: S31: Based on the determined construction parameter requirements and construction accuracy of the target component, the corresponding point cloud data is acquired and processed, and the surface mesh is reconstructed.

[0037] In this embodiment, the construction parameter requirements are to determine the model's accuracy, resolution, material requirements and other parameters based on actual application requirements; point cloud data is a set of three-dimensional coordinate points obtained through laser scanning and other means, and is used to construct the basic data of the three-dimensional model; surface mesh reconstruction is to convert the processed point cloud data into a continuous surface mesh to form a preliminary three-dimensional model.

[0038] Specifically, point cloud data processing includes denoising, filtering and registration, among which denoising is to remove noise points in point cloud data; filtering is to use filtering algorithms (such as Gaussian filtering, median filtering) to smooth point cloud data and reduce the impact of irregular points; registration is to align point cloud data obtained by multiple scans to ensure data consistency; surface mesh reconstruction includes triangulation algorithm, Poisson reconstruction and mesh optimization. The triangulation algorithm is to use the Delaunay triangulation algorithm to convert point cloud data into a triangular mesh; Poisson reconstruction is to use the Poisson surface reconstruction algorithm to generate a smooth surface mesh; mesh optimization optimizes the generated surface mesh to reduce redundant vertices and facets.

[0039] S32: Performing detailed modeling, adding textures, and setting material effects for the target component using detailed modeling software. The detailed modeling includes adding fine structures and / or modeling small components for the target component to obtain a component-level three-dimensional real-scene model.

[0040] In this embodiment, the detail modeling software includes CAD software and 3D modeling software (such as 3ds Max, Blender, Maya, etc.); the detail modeling is to add more refined structures and small components on the basis of the preliminary model, such as adding doors, windows, handrails, decorations and other fine structures on the basis of the preliminary model to improve the realism and detail performance of the model; the texture and material effects add real texture mapping and material settings to the model, among which the texture mapping is to add real texture mapping to the model surface, and the material setting is to set the material properties of the model according to the physical properties of the actual material, such as reflectivity, transparency, etc.

[0041] S33: Optimize the balance between details and performance of the component-level 3D real scene model according to the construction parameter requirements and construction accuracy, and create a multi-precision version of the component-level 3D real scene model.

[0042] In this embodiment, the level of detail (LOD) technology is applied to create multiple models with different precision versions, and the appropriate model version can be automatically selected according to the application scenario. For example, the low-precision version is used for scenes that require fast loading and rendering, such as mobile devices and low-configuration computers; the high-precision version is used for scenes that require high-detail performance, such as professional design and virtual reality applications; performance optimization includes texture compression and material optimization using texture compression technology.

[0043] In one embodiment, the three-dimensional spatial analysis tool includes a viewshed analysis algorithm, a spatial path planning algorithm, and a terrain profile analysis algorithm.

[0044] The visual shed analysis algorithm includes the following steps: S101: Preprocess the acquired three-dimensional scene data and construct an octree spatial index structure.

[0045] In this embodiment, data preprocessing includes cleaning the collected three-dimensional scene data (such as point cloud data, three-dimensional models, etc.), removing noise points, and filling missing data; according to the preprocessed three-dimensional scene data, selecting a suitable root node, and recursively dividing the space into eight subspaces according to the construction principle of the octree until the preset minimum spatial size or maximum depth is reached. In this process, the spatial range of each node and the object information contained therein are recorded for subsequent quick query.

[0046] S102: Based on the constructed octree spatial index structure, starting from a preset observation point, rays are projected along multiple different directions to obtain ray projection paths; the rays are used to simulate line of sight propagation.

[0047] In this embodiment, the positions of one or more observation points are determined according to analysis requirements.

[0048] S103: Based on the ray projection path, traverse the tree nodes of the octree, determine and obtain the intersection points of the ray and the objects in the scene of the three-dimensional scene data, and obtain the visible area starting from the preset observation point.

[0049] In this embodiment, the octree is traversed starting from the root node, along the path of the ray, and traversing the octree layer by layer until the smallest spatial unit that the ray may intersect with is found, and all intersection points of the ray and the object are collected to form the boundary of the visible area starting from the observation point, thereby determining the complete visible area.

[0050] S104: The visual area analysis algorithm introduces a multi-threaded parallel computing algorithm and a GPU acceleration algorithm. The multi-threaded parallel computing algorithm processes the interaction between rays and nodes in parallel according to the hierarchical structure of the octree, and the GPU acceleration algorithm performs the calculation tasks of ray projection and intersection in parallel.

[0051] In this embodiment, multi-threaded parallel computing refers to assigning the interaction tasks between rays and nodes to multiple threads according to the hierarchical structure of the octree, and each thread is responsible for processing a part of the rays; GPU acceleration utilizes the powerful parallel processing capabilities of the GPU to transfer computationally intensive tasks such as ray casting and intersection calculation to the GPU for execution.

[0052] Specifically, by constructing an octree spatial index structure, data in a specific area can be quickly located and retrieved, reducing unnecessary data search time. The application of multi-threaded parallel computing and GPU acceleration technology has greatly improved the efficiency of ray casting and intersection calculation, allowing the visual domain analysis to be completed in a short time.

[0053] In one embodiment, the spatial path planning algorithm includes the following steps: S111: discretizing the three-dimensional space of the designated area where the city-level three-dimensional real scene model is located into grid units, and calculating and associating a corresponding travel cost for each grid unit in combination with preset terrain constraints.

[0054] In this embodiment, the travel cost is a difficulty value calculated for each grid unit based on factors such as terrain and obstacles, and is used for evaluation during path planning.

[0055] Specifically, according to the needs of the application scenario, an appropriate grid unit size is selected, and the three-dimensional space of the specified area is divided into multiple grid units so that the size of each unit is consistent; then the travel cost is calculated for each grid unit based on factors such as terrain features (such as slope, height difference), road type (such as highways, ordinary roads, sidewalks), obstacles (such as buildings, trees), etc.; the associated travel cost is to associate the calculated travel cost value with the corresponding grid unit to form a travel cost matrix.

[0056] S112: Based on the gridded three-dimensional space and the travel cost of the grid cells, a preset heuristic search algorithm is used to perform a heuristic path search, and the heuristic function value of each grid cell is calculated, and the search range is gradually expanded until the optimal path from the starting point to the target point is found.

[0057] In this embodiment, the heuristic search algorithm is the A* algorithm, and the heuristic function value is used to evaluate the potential value of each grid unit, including the actual cost from the starting point to the current unit and the estimated cost from the current unit to the end point.

[0058] Specifically, the starting point and end point of the path planning are determined, and an initial open list and a closed list are created; the open list is used to store the grid cells to be searched, and the closed list is used to store the grid cells that have been searched. Heuristic path search includes heuristic function value calculation, extended search range and path reconstruction, where the heuristic function value calculation is for each grid cell, the formula for calculating its heuristic function value is: f(n)= g(n)+ h(n)f(n)=g(n)+h(n), where g(n) is the actual cost from the starting point to the current cell and h(n) is the estimated cost from the current cell to the end point; extended search range is to select the cell with the smallest heuristic function value from the open list, move it from the open list to the closed list, and expand its neighboring unsearched cells; path reconstruction is to backtrack from the target point to the starting point when the target point is added to the closed list to construct the optimal path.

[0059] and / or, S113: During the path planning process, the spatial path planning algorithm is combined with the dynamic obstacle avoidance mechanism to detect dynamic obstacles in the scene in real time, and calculate the potential impact of the dynamic obstacles on the current path based on the position and movement trend of the dynamic obstacles.

[0060] In this embodiment, dynamic obstacle detection in the scene is performed in real time using sensors such as lidar and cameras to obtain data such as the position, speed, and direction of the dynamic obstacles. The potential impact on the current path is calculated as follows: based on the current state and historical trajectory of the dynamic obstacle, its future movement trend is predicted, and it is evaluated whether it will hinder the path and affect the safety of the path.

[0061] S114: When a dynamic obstacle is detected that may block the current path, a real-time path replanning instruction is triggered. Based on the current position, target position and updated obstacle information, a preset heuristic search algorithm is used again to search for a path to obtain a new optimal path.

[0062] In this embodiment, when a dynamic obstacle is detected that may block the current path, a path replanning instruction is triggered; when the path replanning instruction is received, the travel cost matrix and the heuristic function value are updated according to the current dynamic obstacle information, and the heuristic search algorithm is used to re-search the path based on the updated search conditions, and a new optimal path is selected.

[0063] Furthermore, the spatial path planning algorithm is combined with a multi-objective optimization algorithm. During the path planning process, a custom path planning strategy is generated based on the path planning requirements obtained from user input; the custom path planning strategy calculates the weighted coefficients of the objective functions of each optimization objective based on the different optimization objectives obtained, and evaluates and determines the final path planning scheme; the final path planning scheme includes the optimal path corresponding to different optimization objectives, and the different optimization objectives include but are not limited to the shortest distance, the shortest time, and the lowest energy consumption.

[0064] Among them, the multi-objective optimization algorithm is an algorithm that considers multiple optimization objectives (such as distance, time, energy consumption, etc.) at the same time during the path planning process, and calculates the comprehensive optimal path through weighted coefficients; the weighted coefficient of the objective function is used to evaluate the weight value of the importance of different optimization objectives, and the comprehensive objective function value is calculated through the weighted coefficient. The hybrid database strategy refers to storing different types of data in different databases, such as spatial databases specifically used to store and manage spatial data, such as PostGIS; relational databases used to store metadata and attribute data, such as PostgreSQL; distributed cache systems are used to cache frequently accessed data and query results to improve data access speed, such as Redis.

[0065] The terrain profile analysis algorithm includes the following steps: S121: Extracting elevation data points along a specified path from the digital elevation model.

[0066] Specifically, terrain modeling includes generating digital elevation models based on GIS data and creating terrain textures; a digital elevation model (DEM) is a three-dimensional terrain model generated based on geographic information system (GIS) data, which represents the elevation information of the earth's surface; elevation data points include each coordinate point on the path and its corresponding elevation value; terrain profile analysis algorithms are algorithms used to generate terrain profiles on a specified path, which are often used in geological surveys, urban planning and other fields; in actual applications, users define one or more specified paths through an interface or API, and the paths can be straight lines or curves.

[0067] S122: According to the complexity of the terrain changes, an adaptive sampling algorithm is used to determine the sampling density and sampling points, and the elevation data points on the path are sampled to obtain a series of discrete elevation values.

[0068] Specifically, the adaptive sampling algorithm is an algorithm that dynamically adjusts the sampling density according to the complexity of terrain changes; the terrain complexity can be determined by calculating indicators such as the elevation change rate and slope on the path; adaptive sampling is to dynamically adjust the sampling density according to the terrain complexity: increase the sampling density in areas with larger terrain changes, and reduce the sampling density in areas with smaller terrain changes.

[0069] S123: According to the elevation values ​​between adjacent sampling points, a bilinear interpolation algorithm is used to calculate the elevations between adjacent sampling points to obtain a continuous and smooth terrain profile curve.

[0070] In this embodiment, the bilinear interpolation algorithm is an interpolation algorithm used to estimate the value of an unknown point between two known points.

[0071] S124: Based on a preset geological database, the stratigraphic structure and lithology distribution information of the corresponding position in the profile is determined and integrated with the topographic profile curve to obtain a complete topographic profile including the stratigraphic structure and lithology distribution information.

[0072] In this embodiment, the geological database is a database that stores geological information (such as stratigraphic structure and lithology distribution); by integrating the stratigraphic structure and lithology distribution information in the geological database, a topographic profile containing rich geological information is generated, providing a more comprehensive geological analysis.

[0073] In one embodiment, if Figure 3 As shown, the method for constructing a real-scene three-dimensional model based on a digital city also includes: S10: Based on the city-level 3D real-scene model and the component-level 3D real-scene model, the building detail information inside the building is integrated to construct a comprehensive rescue environment model; the building detail information includes internal passages, stairs and emergency exits.

[0074] In this embodiment, the building detail information includes detailed information such as passages, stairs, emergency exits, etc. inside the building; the accident information includes specific information of accidents such as fire and collapse.

[0075] Specifically, detailed information such as passages, stairs, emergency exits, etc. inside the building is integrated into the city-level three-dimensional real-scene model to form a comprehensive rescue environment model.

[0076] S20: In the comprehensive rescue environment model, an optimized heuristic search algorithm is used to calculate the optimal rescue route based on real-time traffic data and accident information; the accident information includes the fire spread path predicted based on the building structure and building material information.

[0077] Specifically, the optimized heuristic search algorithm is an improved A* algorithm, which obtains road traffic condition data in real time through traffic monitoring systems, GPS and other means; accident information includes fire spread paths, collapse areas, etc. predicted based on building structure and building material information.

[0078] S30: For the interior of the building, a three-dimensional path planning algorithm is used to combine the complex factors of the vertical and horizontal movement of personnel to generate an approach route that meets the rescue needs; and a targeted route plan is formulated based on the characteristics of different types of rescue equipment.

[0079] In this embodiment, the rescue equipment characteristics are the characteristics and requirements of different types of rescue equipment (such as fire trucks, ambulances, special vehicles, etc.); vertical movement such as stairs and elevators, horizontal movement such as corridors and rooms; route complexity factors include factors such as channel layout, stair locations, emergency exits, etc. inside the building; and rescue needs include safety needs and speed needs, so as to facilitate the provision of comprehensive and safe path planning technology.

[0080] Specifically, an optimized heuristic search algorithm is used, combined with real-time traffic data and accident information, to calculate the optimal rescue route from the starting point to the end point. During planning, the fire spread path can also be predicted based on the building structure and building material information to ensure that the rescue route avoids dangerous areas. The targeted route plan is a targeted rescue route plan formulated according to the characteristics of different types of rescue equipment, which improves the travel efficiency and safety of the rescue equipment and ensures the smooth progress of the rescue mission.

[0081] It should be understood that the serial numbers of the steps in the above embodiments do not imply a sequence of execution. The execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0082] In one embodiment, a real-scene 3D model construction system based on a digital city is provided. The real-scene 3D model construction system based on a digital city corresponds to the real-scene 3D model construction method based on a digital city in the above embodiment.

[0083] The real-life 3D model construction system based on the digital city includes a data acquisition module, a data processing terminal, and a visualization interactive tool. The detailed description of each functional module is as follows: Data acquisition module, used to obtain multi-source sensor data of the city; The data processing terminal is used to pre-process the multi-source sensor data of the city, which includes urban image data, point cloud data and GIS data; convert the pre-processed multi-source sensor data in different formats into a preset standard format and perform multi-source sensor data registration and fusion to obtain fused multi-source data; The fused multi-source data is optimized, and a city-level 3D real scene model is constructed based on the optimized fused multi-source data and the preset 3D space analysis tool; the preset 3D space analysis tool provides spatial analysis and 3D space visualization functions; wherein, the steps of constructing the city-level 3D real scene model include terrain modeling, building modeling, texture mapping, road and vegetation modeling operations based on city image data, point cloud data and GIS data, and obtaining the corresponding city initial model and model integration elements; the city initial model is optimized, and the model is integrated based on all the corresponding model integration elements to obtain the city-level 3D real scene model; A visualization and interactive tool used to display city-level 3D real-scene models and conduct 3D interactions.

[0084] For the specific limitations of the real-life three-dimensional model construction system based on digital cities, please refer to the limitations of the real-life three-dimensional model construction method based on digital cities mentioned above, which will not be repeated here; each module in the above-mentioned real-life three-dimensional model construction system based on digital cities can be implemented in whole or in part through software, hardware and a combination thereof; each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0085] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store city-level real-scene three-dimensional models and component-level real-scene three-dimensional models. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the steps of the method for constructing a real-scene three-dimensional model based on a digital city are implemented.

[0086] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of a method for constructing a real-life three-dimensional model based on a digital city are implemented.

[0087] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of a method for constructing a real-scene three-dimensional model based on a digital city are implemented.

[0088] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0089] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0090] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the features thereof may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for constructing a real-scene three-dimensional model based on a digital city, characterized in that: include: Acquire multi-source sensor data of the city and perform data preprocessing, wherein the multi-source sensor data includes city image data, point cloud data and GIS data; Converting the pre-processed multi-source sensor data in different formats into a preset standard format and performing multi-source sensor data registration and fusion to obtain fused multi-source data; Optimizing the fused multi-source data, and constructing a city-level three-dimensional real scene model according to the optimized fused multi-source data and a preset three-dimensional spatial analysis tool; The preset three-dimensional space analysis tool provides space analysis and three-dimensional space visualization functions; The steps of constructing the city-level three-dimensional real scene model include: Perform terrain modeling, building modeling, texture mapping, road and vegetation modeling operations based on the city image data, point cloud data and GIS data to obtain corresponding city initial models and model integration elements; The initial city model is optimized, and model integration is performed based on all corresponding model integration elements to obtain the city-level three-dimensional real-scene model.

2. The method for constructing a real-life three-dimensional model based on a digital city according to claim 1, characterized in that: The method also includes constructing a component-level three-dimensional reality model; The steps of constructing the component-level three-dimensional real scene model include: Based on the determined construction parameter requirements and construction accuracy of the target component, the corresponding point cloud data is obtained and processed, and the surface mesh is reconstructed; Performing detailed modeling, adding textures, and setting material effects for the target component using detailed modeling software, wherein the detailed modeling includes adding fine structures and / or modeling small components for the target component to obtain a component-level three-dimensional real-scene model; The component-level three-dimensional real scene model is optimized in terms of details and performance in a balanced manner according to the construction parameter requirements and construction accuracy, and a multi-precision version of the component-level three-dimensional real scene model is created.

3. The method for constructing a real-life three-dimensional model based on a digital city according to claim 1, characterized in that: The three-dimensional spatial analysis tool includes a visual area analysis algorithm, and the method includes: Perform data preprocessing on the acquired three-dimensional scene data and construct an octree spatial index structure; Based on the constructed octree spatial index structure, starting from a preset observation point, rays are projected along multiple different directions to obtain ray projection paths; the rays are used to simulate line of sight propagation; Based on the ray projection path, traverse the tree nodes of the octree, determine and obtain the intersection point between the ray and the object in the scene of the three-dimensional scene data, and obtain the visible area starting from the preset observation point; and / or, The visual area analysis algorithm introduces a multi-threaded parallel computing algorithm and a GPU acceleration algorithm. The multi-threaded parallel computing algorithm processes the interaction between rays and nodes in parallel according to the hierarchical structure of the octree, and the GPU acceleration algorithm performs the calculation tasks of ray projection and intersection in parallel.

4. The method for constructing a real-life three-dimensional model based on a digital city according to claim 1, characterized in that: The three-dimensional space analysis tool includes a space path planning algorithm, and the method includes: Discretize the three-dimensional space of the designated area where the city-level three-dimensional real scene model is located into grid units, and calculate and associate a corresponding travel cost for each grid unit in combination with preset terrain constraints; Based on the gridded three-dimensional space and the travel cost of the grid cells, a preset heuristic search algorithm is used to perform heuristic path search, and the heuristic function value of each grid cell is calculated, and the search range is gradually expanded until the optimal path from the starting point to the target point is found; and / or, During the path planning process, the spatial path planning algorithm combines the dynamic obstacle avoidance mechanism to detect dynamic obstacles in the scene in real time, and calculates the potential impact of dynamic obstacles on the current path based on the position and movement trend of the dynamic obstacles; When a dynamic obstacle is detected that may block the current path, a real-time path replanning instruction is triggered. Based on the current position, target position and updated obstacle information, the preset heuristic search algorithm is used again to search the path and obtain a new optimal path.

5. The method for constructing a real-scene three-dimensional model based on a digital city according to claim 4, characterized in that: The method further comprises: The spatial path planning algorithm is combined with a multi-objective optimization algorithm. During the path planning process, a custom path planning strategy is generated based on the path planning requirements input by the user; the custom path planning strategy calculates the weighted coefficients of the objective functions of each optimization goal according to the different optimization goals obtained, and evaluates and determines the final path planning solution; the final path planning solution includes the optimal path corresponding to different optimization goals, and the different optimization goals include but are not limited to the shortest distance, the shortest time and the lowest energy consumption; or, The method comprises: Distributed file repository for storing raw multi-source sensor data, large model files, and custom spatial index structures; Based on the hybrid database strategy, the spatial data is stored in the preset spatial database, while the metadata and attribute data are stored in the preset relational database; Based on the preset distributed cache system, frequently accessed data and query results are cached in the memory. According to the access frequency and importance of the data, an intelligent data tiered storage strategy is executed to automatically migrate data between high-speed storage and large-capacity storage.

6. The method for constructing a real-scene three-dimensional model based on a digital city according to claim 1, characterized in that: The three-dimensional spatial analysis tool includes a terrain profile analysis algorithm, The terrain modeling includes generating a digital elevation model based on GIS data and creating terrain textures; The terrain profile analysis algorithm includes: Extracting elevation data points along a specified path from the digital elevation model; According to the complexity of terrain changes, an adaptive sampling algorithm is used to determine the sampling density and sampling points, and the elevation data points on the path are sampled to obtain a series of discrete elevation values; According to the elevation values ​​between adjacent sampling points, the bilinear interpolation algorithm is used to calculate the elevation between adjacent sampling points to obtain a continuous and smooth terrain profile curve; Based on a preset geological database, the stratigraphic structure and lithology distribution information of the corresponding position in the profile is determined and integrated with the topographic profile curve to obtain a complete topographic profile containing the stratigraphic structure and lithology distribution information.

7. The method for constructing a real-scene three-dimensional model based on a digital city according to claim 2, characterized in that: The method further comprises: Based on the city-level 3D real scene model and the component-level 3D real scene model, the building detail information inside the building is integrated to construct a comprehensive rescue environment model; wherein the building detail information includes internal passages, stairs and emergency exits; In the comprehensive rescue environment model, an optimized heuristic search algorithm is used to calculate the optimal rescue route based on real-time traffic data and accident information; the accident information includes a fire spread path predicted based on building structure and building material information; For the interior of buildings, a three-dimensional path planning algorithm is used to combine the complex factors of the vertical and horizontal movements of personnel to generate an approach route that meets rescue needs; and targeted route plans are formulated based on the characteristics of different types of rescue equipment.

8. A real-scene three-dimensional model construction system based on a digital city, characterized in that: The system is used to execute the method for constructing a real-scene three-dimensional model based on a digital city according to any one of claims 1 to 7, and comprises: Data acquisition module, used to obtain multi-source sensor data of the city; A data processing terminal is used to pre-process the multi-source sensor data of the city, wherein the multi-source sensor data includes city image data, point cloud data and GIS data; convert the pre-processed multi-source sensor data in different formats into a preset standard format and perform multi-source sensor data registration and fusion to obtain fused multi-source data; The fused multi-source data is optimized, and a city-level 3D real scene model is constructed according to the optimized fused multi-source data and a preset 3D space analysis tool; the preset 3D space analysis tool provides space analysis and 3D space visualization functions; wherein the step of constructing the city-level 3D real scene model includes performing terrain modeling, building modeling, texture mapping, road and vegetation modeling operations based on the city image data, point cloud data and GIS data to obtain a corresponding city initial model and model integration elements; the city initial model is optimized, and model integration is performed based on all corresponding model integration elements to obtain the city-level 3D real scene model; A visualization interactive tool is used to display the city-level three-dimensional real-scene model and perform three-dimensional interaction.

9. A computer 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 computer program, the steps of the method for constructing a real-life three-dimensional model based on a digital city as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for constructing a real-life three-dimensional model based on a digital city as described in any one of claims 1 to 7 are implemented.

Images