A multi-source heterogeneous data fusion method of a city-level two-three-dimensional integrated model

Abstract

The application discloses a kind of city level two three-dimensional integration model multi-source heterogeneous data fusion method, comprising: obtaining multi-source heterogeneous data from different data sources and structure, form the data set of consistent space-time reference;The data after space-time registration is processed with semantic association, the semantic mapping and associated relationship between different data elements are established, and the integrated data set in semantics is formed;After the data after space-time registration and semantic association processing, the city level full space digital map of two three-dimensional integration covering city ground and underground is constructed;Based on the city level full space digital map, the thematic layer for city management is constructed to provide visual display framework, support the business scene construction and analysis application for city management, not only realize holographic digital expression from static structure to dynamic operation, but also through integrated spatial analysis algorithm and semantic reasoning capability, can support the depth analysis and intelligent decision of various business scenes of city.

Description

Technical Field

[0001] This invention relates to the field of city-level model data fusion technology, specifically to a method for fusion of multi-source heterogeneous data of a city-level integrated 2D and 3D model. Background Technology

[0002] With the deepening of the construction of digital twin cities and smart cities, urban management has evolved from a static depiction of the three-dimensional morphology of the earth's surface to the need to construct a dynamic, three-dimensional, and intelligent digital twin that integrates all spatial elements of the earth's surface, above ground, and underground, and incorporates real-time monitoring data from the Internet of Things. This full-space digital twin city is a key infrastructure for realizing refined urban management, scientific decision-making, and intelligent emergency response, and it plays an indispensable supporting role, especially for cutting-edge scenarios such as low-altitude economy, underground space development and utilization, and urban safety operation.

[0003] However, current mainstream urban 3D modeling and data management technologies have significant limitations in achieving this goal, making it difficult to support true integrated perception and analysis of the entire space. The main shortcomings of existing technologies are reflected in the following three aspects:

[0004] (1) Existing technologies mostly focus on a single spatial dimension. High-precision real-world 3D maps generated by technologies such as oblique photography and lidar are usually limited to the surface and above-ground buildings, lacking the expression of underground space. The vertical separation of above-ground, surface and underground data makes it impossible to conduct cross-dimensional integrated query, analysis and visualization under a unified spatial framework, forming data silos.

[0005] (2) The data involved in urban digital twins have multi-source heterogeneous characteristics, including static spatial models, vector thematic data, and dynamic IoT sensor data. Most existing systems manage single-type data and lack a unified and standardized framework to effectively integrate these multimodal data. Summary of the Invention

[0006] The purpose of this invention is to provide a multi-source heterogeneous data fusion method for a city-level integrated 2D and 3D model, in order to solve the technical problems of spatial dimension fragmentation, data silos, and isolated data modalities in the existing technology, which create fusion barriers.

[0007] To solve the above-mentioned technical problems, the present invention specifically provides the following technical solution:

[0008] This invention provides a method for fusing multi-source heterogeneous data in a city-level integrated 2D / 3D model, comprising the following steps:

[0009] Acquire multi-source heterogeneous data from different data sources and structures. The multi-source heterogeneous data includes at least: above-ground real-scene 3D maps, underground 3D geological maps, underground pipeline data, building information models, satellite remote sensing data, and Internet of Things dynamic monitoring data generated by oblique photogrammetry or lidar technology.

[0010] The acquired multi-source heterogeneous data are spatiotemporally registered and unified to the same spatial coordinate system and temporal reference to form a data set with consistent spatiotemporal reference.

[0011] Semantic association processing is performed on the spatiotemporally registered data to establish semantic mapping and association relationships between different data elements, forming a semantically interoperable integrated data set;

[0012] The data, after being processed by spatiotemporal registration and semantic association, are integrated in a unified manner according to a pre-set unified data fusion framework and standards to construct a city-level full-space digital map covering both above-ground and underground areas, integrating two and three dimensions.

[0013] Based on the aforementioned city-level full-space digital map, a thematic layer for urban management is constructed to provide a visualization framework, and spatial analysis algorithms are integrated to support the construction and analysis of business scenarios for urban management.

[0014] As a preferred embodiment of the present invention, multi-source heterogeneous data is acquired and integrated. This multi-source heterogeneous data includes real-scene 3D maps, underground 3D geological maps, underground pipeline network data, underground public space data, satellite remote sensing data, basic surface geographic data, and IoT dynamic monitoring data, including:

[0015] High-resolution images and point cloud data of the urban surface and above-ground structures are obtained by scanning with vehicle-mounted LiDAR, and the high-precision real-scene 3D map is generated after processing.

[0016] Based on engineering survey borehole data, ground-penetrating radar detection data, and historical geological data, a three-dimensional underground geological map represented by a Grid / Tin hybrid data model is constructed.

[0017] Obtain pipeline as-built drawings and detection data from urban infrastructure management departments to construct a three-dimensional vector diagram of the underground pipeline network data;

[0018] By integrating building information modeling, civil defense engineering drawings, and underground space surveying data, a three-dimensional map of the underground public space data is constructed.

[0019] Acquire multi-temporal and multi-spectral satellite remote sensing image data as a dynamic base map and supplementary information source for the aforementioned basic geographic data of the Earth's surface;

[0020] It can access IoT sensor networks deployed in buildings, pipelines, geological monitoring points, and urban environments to acquire the IoT dynamic monitoring data in real time or near real time.

[0021] As a preferred embodiment of the present invention, the underground three-dimensional geological map represented by the Grid / Tin hybrid data model includes:

[0022] For the aforementioned underground three-dimensional geological map, in the layered geological body areas with relatively uniform and regular properties and structures within the model, a regular grid data structure is used for representation;

[0023] In areas with complex structures or requiring precise geometric depiction, such as model boundaries, stratigraphic contact surfaces, faults, and intrusions, an irregular triangular mesh data structure is used for representation.

[0024] By using a preset data structure and indexing mechanism, the regular grid and the irregular triangular network are seamlessly connected to form the Grid / Tin hybrid data model.

[0025] As a preferred embodiment of the present invention, the acquired multi-source heterogeneous data is spatiotemporally registered and unified to the same spatial coordinate system and temporal reference, forming a data set with consistent spatiotemporal reference, including:

[0026] A unified three-dimensional spatial reference framework is established with the urban plane coordinate system and elevation benchmark as the core. The real-scene three-dimensional map, basic surface geographic data, underground pipe network data, underground public space data and satellite remote sensing imagery are uniformly transformed into the unified three-dimensional spatial reference framework through the coordinate transformation parameter method.

[0027] For the underground three-dimensional geological map and BIM model stored in relative coordinates, three-dimensional spatial registration and mosaicking are performed using their known control points, feature points or their association with the above-ground model, so as to accurately anchor them into the three-dimensional unified reference frame.

[0028] The dynamic monitoring data is assigned a unified timestamp, and its spatial location information is associated and bound with the corresponding entity under the unified three-dimensional spatial reference frame.

[0029] For multi-time series data such as satellite remote sensing data, a unified time series benchmark is established, and data from different time phases are aligned and organized according to the acquisition time to form a data cube with a time dimension.

[0030] After unifying the spatial reference, the spatial topology relationships of adjacent or related data layers are checked and corrected to form a data set consistent with the spatiotemporal reference.

[0031] As a preferred embodiment of the present invention, semantic association processing is performed on the spatiotemporally registered data to establish semantic mapping and association relationships between different data elements, forming a semantically interoperable integrated data set, including:

[0032] Based on the spatiotemporal benchmark-consistent data set, a cross-domain unified semantic ontology library is constructed to define standardized descriptions of the types, attributes, relationships, and constraints of various data elements.

[0033] Using the unified semantic ontology library, automatic semantic parsing and annotation are performed on various data elements after spatiotemporal registration, assigning standardized semantic tags and attribute sets to entities such as buildings and roads in the real-world 3D map, strata and faults in the underground 3D geological map, pipelines and wells in the underground pipeline network data, and sensors in the dynamic monitoring data.

[0034] Based on the semantic tags, establish semantic mapping relationships between data elements describing the same real-world entity from different data sources;

[0035] Using the semantic mapping relationship as edges and data elements with semantic labels as nodes, a city-level semantic association graph spanning above-ground and underground areas is constructed, forming the core carrier of an integrated data set that is semantically interoperable.

[0036] As a preferred embodiment of the present invention, the data after spatiotemporal registration and semantic association processing is integrated into a unified data fusion framework and standard according to a preset framework, including:

[0037] Based on the spatiotemporally consistent data set and the semantic association graph, spatial data services and semantic query services are established.

[0038] The real-world 3D map is used as the upper surface layer, and the underground 3D geological map, underground pipeline map and underground public space are used as the underground layer for integrated and visualized rendering to construct a two-dimensional integrated scene.

[0039] Establish a real-time data channel between the dynamic monitoring data and the corresponding semantic entities in the integrated 2D / 3D scene, and dynamically inject the dynamic monitoring data as real-time tags and overlay them on the corresponding static model.

[0040] Based on the aforementioned integrated 2D / 3D scene and dynamic data, an integrated fusion framework serving different application fields is generated on demand.

[0041] As a preferred embodiment of the present invention, a city-level full-space digital map covering both above-ground and underground areas, integrating two-dimensional and three-dimensional elements, is constructed, including:

[0042] Based on the integrated fusion framework, a true three-dimensional form and spatial relationship of urban surface topography, above-ground buildings, underground geological structure, underground pipe network and underground space are constructed;

[0043] The semantic association graph is used to establish multi-level attribute information, business rules and dynamic association relationships covering all types of entities in the space.

[0044] Through real-time data channels, real-time monitoring data from the Internet of Things is continuously integrated and visualized to reflect the dynamic changes in urban operations;

[0045] Obtain spatiotemporal benchmarks and topological relationships consistent with internal data, and construct a city-level full-space digital map based on a unified spatiotemporal data foundation.

[0046] As a preferred embodiment of the present invention, based on the aforementioned city-level full-space digital map, a thematic layer for urban management is constructed to provide a visualization framework, including:

[0047] Based on the city-level full-space digital map, the full-space data is logically sliced ​​and physically organized according to a preset thematic classification system to form multiple independently managed and overlayable thematic data layers; the thematic classification system includes at least eight categories: administrative boundaries, roads and traffic, buildings, place names and addresses, thematic facilities, vegetation cover, water systems and lakes, and terrain slope.

[0048] For specific business scenarios in urban management, thematic layers containing real-time status information are dynamically generated by calling the semantic association graph and real-time data channels.

[0049] Spatial analysis algorithms are used to deeply integrate with the city-level full-space digital map for visualization, enabling real-time rendering in a two-dimensional and three-dimensional integrated scene.

[0050] As a preferred embodiment of the present invention, a spatial analysis algorithm is used to perform deep visualization integration with the city-level full-space digital map, including:

[0051] In the integrated 2D and 3D scene, a graphical interactive tool is provided, allowing users to define the parameters required for analysis directly in the 3D scene by clicking, dragging, and drawing.

[0052] Based on the defined parameters, the corresponding spatial analysis service is invoked in real time to perform calculations, and the intermediate process and the final result are rendered in real time to the integrated 2D and 3D scene.

[0053] The profile analysis results are generated in real time and dynamically drawn along the profile path to create a two-dimensional profile map containing above-ground buildings, underground pipe networks and geological structures, which is then highlighted in conjunction with the three-dimensional scene.

[0054] The results of the vision analysis are used to draw the line of sight and occlusion in three-dimensional space in real time using visible and invisible visual codes.

[0055] The buffer analysis results are generated and overlaid in real time in three-dimensional space in the form of a semi-transparent three-dimensional volume.

[0056] As a preferred embodiment of the present invention, an integrated spatial analysis algorithm is provided to support the construction and analysis of business scenarios for urban management, including:

[0057] Based on the full-space geometry and topology of the integrated two-dimensional and three-dimensional model, the structural characteristics of urban space are analyzed, and combined with real-time monitoring data and physical laws, urban dynamic processes are simulated and predicted.

[0058] Based on the semantic association graph and historical data, intelligent reasoning and situational assessment are performed;

[0059] When a facility triggers an alarm, it automatically analyzes other facilities that are directly or indirectly related to it in the semantic association graph to assess the risk of cascading effects.

[0060] The system calculates the operational health of the region in real time and displays it as a visual heatmap.

[0061] Compared with the prior art, the present invention has the following advantages:

[0062] This invention constructs a unified data fusion framework and standard, integrating heterogeneous data from multiple sources, including oblique photogrammetry 3D real-world data, underground geological and pipeline models, BIM, satellite remote sensing, and IoT dynamic monitoring, under a unified spatiotemporal benchmark. This results in the geometric registration and semantic association of these data, constructing a comprehensive 2D / 3D digital twin model covering both above-ground and underground urban areas. This model not only achieves a holographic digital representation from static structure to dynamic operation but also, through integrated spatial analysis algorithms and semantic reasoning capabilities, supports in-depth analysis and intelligent decision-making in various business scenarios such as urban planning, emergency command, safety early warning, and facility operation and maintenance. It fundamentally solves the data silos and application barriers inherent in traditional 3D modeling technologies, providing core technical support for the scientific governance and efficient operation of smart cities. Attached Figure Description

[0063] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0064] Figure 1Flowchart of the multi-source heterogeneous data fusion method for a city-level integrated 2D / 3D model provided in this embodiment of the invention

[0065] Figure 2 The flowchart illustrates the algorithm for the multi-source heterogeneous data fusion method provided in this embodiment of the invention. Detailed Implementation

[0066] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] like Figure 1 and Figure 2 As shown, this invention provides a method for fusing multi-source heterogeneous data in a city-level integrated 2D / 3D model, comprising the following steps:

[0068] Acquire multi-source heterogeneous data from different data sources and structures. The multi-source heterogeneous data includes at least: above-ground real-scene 3D maps, underground 3D geological maps, underground pipeline data, building information models, satellite remote sensing data, and Internet of Things dynamic monitoring data generated by oblique photogrammetry or lidar technology.

[0069] The acquired multi-source heterogeneous data are spatiotemporally registered and unified to the same spatial coordinate system and temporal reference to form a data set with consistent spatiotemporal reference.

[0070] Semantic association processing is performed on the spatiotemporally registered data to establish semantic mapping and association relationships between different data elements, forming a semantically interoperable integrated data set;

[0071] In this embodiment, through semantic association processing, isolated model elements, such as a building, a pipeline, or a sensor, are associated with real-world entities with business significance, constructing an interoperable semantic association graph that enables the system to support complex knowledge-based analysis.

[0072] The data, after being processed by spatiotemporal registration and semantic association, are integrated in a unified manner according to a pre-set unified data fusion framework and standards to construct a city-level full-space digital map covering both above-ground and underground areas, integrating two and three dimensions.

[0073] In this embodiment, the constructed full-space digital model can easily generate various thematic layers and deeply integrate advanced spatial analysis algorithms. This loosely coupled, service-oriented architecture greatly facilitates the rapid construction and flexible expansion of business scenarios in different fields such as urban planning, emergency management, low-altitude flight, and public safety.

[0074] Based on the aforementioned city-level full-space digital map, a thematic layer for urban management is constructed to provide a visualization framework, and spatial analysis algorithms are integrated to support the construction and analysis of business scenarios for urban management.

[0075] This invention deeply integrates above-ground 3D real-world maps, underground geological and pipeline models, BIM models, remote sensing images, and dynamic data through a unified spatiotemporal registration and semantic association framework. It establishes a city-level full-space digital map covering everything from the air to the ground, with unified geometry and semantics. This completely solves the core problems of data silos and integration barriers in existing technologies, providing a unique and complete spatial data foundation for digital twin cities.

[0076] Acquire and integrate multi-source heterogeneous data, including real-scene 3D maps, underground 3D geological maps, underground pipeline network data, underground public space data, satellite remote sensing data, basic surface geographic data, and IoT dynamic monitoring data, including:

[0077] High-resolution images and point cloud data of the urban surface and above-ground structures are obtained by scanning with vehicle-mounted LiDAR, and the high-precision real-scene 3D map is generated after processing.

[0078] In this embodiment, the steps for constructing a high-precision real-scene 3D map are as follows:

[0079] High-density point cloud and panoramic images of building facades and ground along urban roads are quickly acquired by vehicle-mounted LiDAR. Five-view high-overlap images of building tops and complex areas are acquired by drone oblique photography. For blind spots that cannot be covered by vehicle-mounted LiDAR and drones, ground laser scanning or handheld mobile mapping equipment are used to fill blind spots and ensure the integrity of urban spatial geometry and texture information, and acquire multi-platform, multi-angle point cloud data.

[0080] After processing point cloud data from multiple platforms and angles, the corresponding point features between the image and the point cloud are extracted. The positioning and orientation parameters of the image and the lidar point cloud data are optimized and calculated as a whole to achieve accurate registration between the image and the point cloud and generate a color point cloud with accurate geographic coordinates.

[0081] The system automatically classifies colored points to construct an irregular triangular mesh, establishes an initial white model of urban buildings and the ground surface, and maps image fragments from the optimal perspective onto the corresponding triangular mesh surface. Through texture balancing and edge editing, it eliminates color differences and stitching seams between different images, generating a high-fidelity 3D map with realistic colors and geometric accuracy.

[0082] Based on engineering survey borehole data, ground-penetrating radar detection data, and historical geological data, a three-dimensional underground geological map represented by a Grid / Tin hybrid data model is constructed.

[0083] In this embodiment, the steps for constructing the underground three-dimensional geological map are as follows:

[0084] By collecting and digitizing historical geological survey reports, hydrogeological data, and regional geological maps using engineering survey borehole data, ground-penetrating radar data, and historical geological data, a multi-source geological database covering the study area is constructed.

[0085] The coordinates of the stratigraphic boundary points corresponding to each borehole and profile are extracted to generate discrete point sets for reconstruction of each layer. For stratigraphic interfaces with relatively continuous layered distribution and gentle changes, the grid layer is calculated by radial basis function interpolation function to ensure the continuity and smoothness of the macroscopic undulation morphology of the regional layer.

[0086] Irregular triangular meshes are used for fine geometric modeling. By constraining the Delaunay triangulation function, a high-precision TIN surface model that can accurately fit complex morphologies such as structural abrupt changes, pinch-outs, and faults is constructed to ensure the geometric fidelity of key geological structures. Three-dimensional solid models are then used to reconstruct the three-dimensional underground geological map.

[0087] Obtain pipeline as-built drawings and detection data from urban infrastructure management departments to construct a three-dimensional vector diagram of the underground pipeline network data;

[0088] By integrating building information modeling, civil defense engineering drawings, and underground space surveying data, a three-dimensional map of the underground public space data is constructed.

[0089] Acquire multi-temporal and multi-spectral satellite remote sensing image data as a dynamic base map and supplementary information source for the aforementioned basic geographic data of the Earth's surface;

[0090] It can access IoT sensor networks deployed in buildings, pipelines, geological monitoring points, and urban environments to acquire the IoT dynamic monitoring data in real time or near real time.

[0091] In this embodiment, a systematic data acquisition channel covering all dimensions of the city, including above-ground, surface, underground, and dynamic sensing, was planned. It not only utilizes vehicle-mounted LiDAR to acquire high-precision surface images, but also innovatively integrates professional geological exploration data such as engineering survey boreholes and ground-penetrating radar to construct underground models, ensuring the professionalism and authority of core spatial data. At the same time, by combining the macroscopic coverage of multispectral remote sensing with the microscopic real-time perception of IoT sensors, a multi-scale, multi-precision complementary data system was formed, ranging from centimeter-level fine modeling to kilometer-level macroscopic monitoring, and from static structures to dynamic states. This provides reliable data for building a high-fidelity digital twin city.

[0092] The underground three-dimensional geological map represented by the Grid / Tin hybrid data model includes:

[0093] For the aforementioned underground three-dimensional geological map, in the layered geological body areas with relatively uniform and regular properties and structures within the model, a regular grid data structure is used for representation;

[0094] In areas with complex structures or requiring precise geometric depiction, such as model boundaries, stratigraphic contact surfaces, faults, and intrusions, an irregular triangular mesh data structure is used for representation.

[0095] By using a preset data structure and indexing mechanism, the regular grid and the irregular triangular network are seamlessly connected to form the Grid / Tin hybrid data model.

[0096] In this embodiment, the steps for constructing the Grid / Tin hybrid data model are as follows:

[0097] Using the aforementioned underground 3D geological map, areas with uniform internal properties, continuous layered distribution, and gentle morphological changes are formed into Grid regions using a regular grid data structure. For areas requiring precise geometric depiction, such as model boundaries, stratigraphic pinch-out points, fault fracture zones, intrusive rock contact zones, and karst development areas, an irregular triangular mesh data structure is used to form Tin regions. At the boundary between Grid and Tin regions, a buffer zone of a certain width is set for smooth transition of the data structure.

[0098] The grid nodes on the boundary of the Grid region are used as hard constraint boundaries for the triangulation of the Tin region, ensuring that the boundary edges of Tin and Grid are completely coincident; and a unified node index table is established at the intersection of Grid and Tin, where the boundary nodes of Tin directly reference the boundary nodes of Grid, forming shared edges and nodes.

[0099] Construct a hybrid spatial index that covers the entire model, indexing both the regular grid blocks of the Grid and the triangular grid blocks of the Tin;

[0100] The processed Grid and Tin blocks are encapsulated into a unified hybrid data model object, providing a unified access interface to the outside world, thus forming a Grid / Tin hybrid data model.

[0101] In this embodiment, a Grid / Tin hybrid model is adopted. The Tin model is used to accurately fit the stratigraphic contact surface, fault line, intrusion boundary, etc., which greatly improves the geometric fidelity of the three-dimensional geological model and makes it more realistically reflect the actual situation of underground space.

[0102] In this embodiment, a Grid / Tin hybrid model is adopted, using regular grids in homogeneous internal regions and irregular triangular meshes in complex boundaries. This enables the model to achieve high-precision geometric representation of key geological structures while maintaining efficient storage and rapid spatial analysis capabilities for massive amounts of data.

[0103] The acquired multi-source heterogeneous data are spatiotemporally registered and unified to the same spatial coordinate system and temporal reference, forming a data set with a consistent spatiotemporal reference, including:

[0104] A unified three-dimensional spatial reference framework is established with the urban plane coordinate system and elevation benchmark as the core. The real-scene three-dimensional map, basic surface geographic data, underground pipe network data, underground public space data and satellite remote sensing imagery are uniformly transformed into the unified three-dimensional spatial reference framework through the coordinate transformation parameter method.

[0105] For the underground three-dimensional geological map and BIM model stored in relative coordinates, three-dimensional spatial registration and mosaicking are performed using their known control points, feature points or their association with the above-ground model, so as to accurately anchor them into the three-dimensional unified reference frame.

[0106] The dynamic monitoring data is assigned a unified timestamp, and its spatial location information is associated and bound with the corresponding entity under the unified three-dimensional spatial reference frame.

[0107] For multi-time series data such as satellite remote sensing data, a unified time series benchmark is established, and data from different time phases are aligned and organized according to the acquisition time to form a data cube with a time dimension.

[0108] After unifying the spatial reference, the spatial topology relationships of adjacent or related data layers are checked and corrected to form a data set consistent with the spatiotemporal reference.

[0109] In this embodiment, coordinate transformation is combined with three-dimensional spatial registration and mosaicking. By utilizing known relationships, such as the positional correspondence between building pile foundations and bearing layers, it ensures that the above-ground buildings and their foundation soil layers, as well as the underground pipelines and their geological environment, fit together seamlessly in three-dimensional space. This truly breaks down the vertical barrier from above ground to underground from a geometric perspective, providing a basis for analyzing the interaction between buildings and geology, the geological suitability of pipeline laying, and so on.

[0110] Semantic association processing is performed on the spatiotemporally registered data to establish semantic mapping and association relationships between different data elements, forming a semantically interoperable integrated data set, including:

[0111] Based on the spatiotemporal benchmark-consistent data set, a cross-domain unified semantic ontology library is constructed to define standardized descriptions of the types, attributes, relationships, and constraints of various data elements.

[0112] Using the unified semantic ontology library, automatic semantic parsing and annotation are performed on various data elements after spatiotemporal registration, assigning standardized semantic tags and attribute sets to entities such as buildings and roads in the real-world 3D map, strata and faults in the underground 3D geological map, pipelines and wells in the underground pipeline network data, and sensors in the dynamic monitoring data.

[0113] In this embodiment, by constructing a unified semantic ontology library across domains, a general vocabulary and grammatical rules are established for heterogeneous entities such as buildings, roads, strata, pipelines, and sensors. This is equivalent to providing a knowledge framework for computer systems to understand urban elements. By automatically semantically annotating all data elements, the original geometric models and monitoring readings are transformed into understandable objects with clear semantic labels, enabling the system not only to see the location and shape of objects, but also to understand their attributes.

[0114] Based on the semantic tags, establish semantic mapping relationships between data elements describing the same real-world entity from different data sources;

[0115] Using the semantic mapping relationship as edges and data elements with semantic labels as nodes, a city-level semantic association graph spanning above-ground and underground areas is constructed, forming the core carrier of an integrated data set that is semantically interoperable.

[0116] In this embodiment, a city-level semantic association graph is constructed, with standardized semantic tags as nodes and rich mapping relationships between entities as edges. It can automatically and quickly associate risk pipelines and affected key users along the graph, realizing risk transmission and deduction from monitoring to early warning, and from single point to system. This completely changes the traditional analysis mode that relies on manual experience and decentralized queries, and realizes automated and systematic association impact analysis based on knowledge graph.

[0117] The data, after spatiotemporal registration and semantic association processing, is integrated into a unified system according to a pre-defined unified data fusion framework and standards, including:

[0118] Based on the spatiotemporally consistent data set and the semantic association graph, spatial data services and semantic query services are established.

[0119] The real-world 3D map is used as the upper surface layer, and the underground 3D geological map, underground pipeline map and underground public space are used as the underground layer for integrated and visualized rendering to construct a two-dimensional integrated scene.

[0120] In this embodiment, by establishing spatial data services and semantic query services, the processed massive fusion data is transformed from a file package into an online callable service. By establishing a real-time data channel, dynamic monitoring data is injected into the static model, truly realizing millisecond-level fusion and real-time presentation of dynamic and static data.

[0121] Establish a real-time data channel between the dynamic monitoring data and the corresponding semantic entities in the integrated 2D / 3D scene, and dynamically inject the dynamic monitoring data as real-time tags and overlay them on the corresponding static model.

[0122] Based on the aforementioned integrated 2D / 3D scene and dynamic data, an integrated fusion framework serving different application fields is generated on demand.

[0123] In this embodiment, urban space is organized into an upper surface layer and an underground layer at the visualization level, and the two are seamlessly switched and transparently displayed through an integrated rendering engine, which can intuitively view the precise spatial relationship between building foundations and underground soil and rock layers and pipelines.

[0124] Construct a city-level, full-space digital map covering both above-ground and underground areas, integrating two-dimensional and three-dimensional elements, including:

[0125] Based on the integrated fusion framework, a true three-dimensional form and spatial relationship of urban surface topography, above-ground buildings, underground geological structure, underground pipe network and underground space are constructed;

[0126] The semantic association graph is used to establish multi-level attribute information, business rules and dynamic association relationships covering all types of entities in the space.

[0127] Through real-time data channels, real-time monitoring data from the Internet of Things is continuously integrated and visualized to reflect the dynamic changes in urban operations;

[0128] Obtain spatiotemporal benchmarks and topological relationships consistent with internal data, and construct a city-level full-space digital map based on a unified spatiotemporal data foundation.

[0129] In this embodiment, the city-level full-space digital map has an internally consistent spatiotemporal reference and topological relationship. Each entity is an intelligent object that integrates precise geometry, rich attributes and real-time status, and can perform knowledge-based reasoning and state-based simulation, and can dynamically reflect the city's operational status.

[0130] Based on the aforementioned city-level full-space digital map, a thematic layer for urban management is constructed to provide a visualization framework, including:

[0131] Based on the city-level full-space digital map, the full-space data is logically sliced ​​and physically organized according to a preset thematic classification system to form multiple independently managed and overlayable thematic data layers; the thematic classification system includes at least eight categories: administrative boundaries, roads and traffic, buildings, place names and addresses, thematic facilities, vegetation cover, water systems and lakes, and terrain slope.

[0132] In this embodiment, through logical slicing and physical organization, thematic layers covering eight basic categories and even more subdivided fields can be quickly generated. This multi-faceted approach ensures absolute consistency of all thematic data in terms of spatial benchmarks, data timeliness, and semantic definitions.

[0133] For specific business scenarios in urban management, thematic layers containing real-time status information are dynamically generated by calling the semantic association graph and real-time data channels.

[0134] Spatial analysis algorithms are used to deeply integrate with the city-level full-space digital map for visualization, enabling real-time rendering in a two-dimensional and three-dimensional integrated scene.

[0135] In this embodiment, thematic layers containing real-time status can be dynamically generated based on semantic association maps and real-time data channels. For example, a real-time safety status map of urban lifelines can be generated, which integrates real-time data such as pipeline pressure, flow, and leakage monitoring, greatly improving the city's safety operation guarantee capability and emergency response efficiency.

[0136] The spatial analysis algorithm is used to perform deep visualization integration with the city-level full-space digital map, including:

[0137] In the integrated 2D and 3D scene, a graphical interactive tool is provided, allowing users to define the parameters required for analysis directly in the 3D scene by clicking, dragging, and drawing.

[0138] Based on the defined parameters, the corresponding spatial analysis service is invoked in real time to perform calculations, and the intermediate process and the final result are rendered in real time to the integrated 2D and 3D scene.

[0139] The profile analysis results are generated in real time and dynamically drawn along the profile path to create a two-dimensional profile map containing above-ground buildings, underground pipe networks and geological structures, which is then highlighted in conjunction with the three-dimensional scene.

[0140] The results of the vision analysis are used to draw the line of sight and occlusion in three-dimensional space in real time using visible and invisible visual codes.

[0141] The buffer analysis results are generated and overlaid in real time in three-dimensional space in the form of a semi-transparent three-dimensional volume.

[0142] In this embodiment, by providing a graphical interactive tool that allows direct clicking, dragging, and drawing in a 3D scene, operations such as defining profile paths, line-of-sight directions, and buffer zones are simplified, completely eliminating the complex parameter input process. This allows planning, emergency response, and management personnel without a strong GIS background to easily initiate complex 3D spatial analyses, truly realizing the business-oriented popularization of professional analysis tools.

[0143] Integrating spatial analysis algorithms, it supports the construction and analysis of business scenarios for urban management, including:

[0144] Based on the full-space geometry and topology of the integrated two-dimensional and three-dimensional model, the structural characteristics of urban space are analyzed, and combined with real-time monitoring data and physical laws, urban dynamic processes are simulated and predicted.

[0145] Based on the semantic association graph and historical data, intelligent reasoning and situational assessment are performed;

[0146] When a facility triggers an alarm, it automatically analyzes other facilities that are directly or indirectly related to it in the semantic association graph to assess the risk of cascading effects.

[0147] The system calculates the operational health of the region in real time and displays it as a visual heatmap.

[0148] In this embodiment, the spatial analysis algorithm flow is as follows:

[0149] Input layer: Input multi-source heterogeneous fusion data, provide logical relationships between entities through semantic association graph, and perform process simulation and situational awareness of entities based on real-time monitoring data;

[0150] Based on monitoring data and physical models, we extrapolate and analyze the geometric problems of "what, where, and how" through full-space structural analysis, and perform calculations based on geometry and topology.

[0151] Output layer: Through dynamic thematic layers, real-time 3D visualization and other methods, it is transformed into an intuitive situation map, and all outputs are gathered and used to serve specific urban management business scenarios, such as planning review, emergency command, safety early warning, facility operation and maintenance, etc., to realize the value closed loop from data to decision-making.

[0152] This embodiment is not limited to the analysis of the static structure of urban space. More importantly, it can combine real-time monitoring data with physical law models to simulate and predict the dynamic processes of the city. This enables city managers to shift from passively responding to events that have already occurred to proactively anticipating risks, simulating solutions, and optimizing decisions. This significantly moves the management window forward and greatly improves the resilience and safety level of the city.

[0153] This invention constructs a unified data fusion framework and standard, integrating heterogeneous data from multiple sources, including oblique photogrammetry 3D real-world data, underground geological and pipeline models, BIM, satellite remote sensing, and IoT dynamic monitoring, under a unified spatiotemporal benchmark. This results in the geometric registration and semantic association of these data, constructing a comprehensive 2D / 3D digital twin model covering both above-ground and underground urban areas. This model not only achieves a holographic digital representation from static structure to dynamic operation but also, through integrated spatial analysis algorithms and semantic reasoning capabilities, supports in-depth analysis and intelligent decision-making in various business scenarios such as urban planning, emergency command, safety early warning, and facility operation and maintenance. It fundamentally solves the data silos and application barriers inherent in traditional 3D modeling technologies, providing core technical support for the scientific governance and efficient operation of smart cities.

[0154] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A method for fusing multi-source heterogeneous data in a city-level integrated 2D / 3D model, characterized in that, Includes the following steps: Acquire multi-source heterogeneous data from different data sources and structures. The multi-source heterogeneous data includes at least: above-ground real-scene 3D maps, underground 3D geological maps, underground pipeline data, building information models, satellite remote sensing data, and Internet of Things dynamic monitoring data generated by oblique photogrammetry or lidar technology. The acquired multi-source heterogeneous data are spatiotemporally registered and unified to the same spatial coordinate system and temporal reference to form a data set with consistent spatiotemporal reference. Semantic association processing is performed on the spatiotemporally registered data to establish semantic mapping and association relationships between different data elements, forming a semantically interoperable integrated data set; Semantic association processing is performed on the spatiotemporally registered data to establish semantic mapping and association relationships between different data elements, forming a semantically interoperable integrated data set, including: Based on the spatiotemporal benchmark-consistent data set, a cross-domain unified semantic ontology library is constructed to define standardized descriptions of the types, attributes, relationships, and constraints of various data elements. Using the unified semantic ontology library, automatic semantic parsing and annotation are performed on various data elements after spatiotemporal registration, assigning standardized semantic tags and attribute sets to entities such as buildings and roads in the real-world 3D map, strata and faults in the underground 3D geological map, pipelines and wells in the underground pipeline network data, and sensors in the dynamic monitoring data. Based on the semantic tags, establish semantic mapping relationships between data elements describing the same real-world entity from different data sources; Using the semantic mapping relationship as edges and data elements with semantic labels as nodes, a city-level semantic association graph spanning above-ground and underground areas is constructed, forming the core carrier of a semantically interoperable integrated data set. The data, after being processed by spatiotemporal registration and semantic association, are integrated in a unified manner according to a pre-set unified data fusion framework and standards to construct a city-level full-space digital map covering both above-ground and underground areas, integrating two and three dimensions. The data, after spatiotemporal registration and semantic association processing, is integrated into a unified system according to a pre-defined unified data fusion framework and standards, including: Based on the spatiotemporally consistent data set and the semantic association graph, spatial data services and semantic query services are established. The real-world 3D map is used as the upper surface layer, and the underground 3D geological map, underground pipeline map and underground public space are used as the underground layer for integrated and visualized rendering to construct a two-dimensional integrated scene. Establish a real-time data channel between the dynamic monitoring data and the corresponding semantic entities in the integrated 2D / 3D scene, and dynamically inject the dynamic monitoring data as real-time tags and overlay them on the corresponding static model. Based on the aforementioned integrated 2D / 3D scene and dynamic data, an integrated fusion framework serving different application fields is generated on demand. Based on the aforementioned city-level full-space digital map, a thematic layer for urban management is constructed to provide a visualization framework, and spatial analysis algorithms are integrated to support the construction and analysis of business scenarios for urban management.

2. The method for multi-source heterogeneous data fusion of a city-level integrated 2D / 3D model according to claim 1, characterized in that, Acquire and integrate multi-source heterogeneous data, including real-scene 3D maps, underground 3D geological maps, underground pipeline network data, underground public space data, satellite remote sensing data, basic surface geographic data, and IoT dynamic monitoring data, including: High-resolution images and point cloud data of the urban surface and above-ground structures are obtained by scanning with vehicle-mounted LiDAR, and the high-precision real-scene 3D map is generated after processing. Based on engineering survey borehole data, ground-penetrating radar detection data, and historical geological data, a three-dimensional underground geological map represented by a Grid / Tin hybrid data model is constructed. Obtain pipeline as-built drawings and detection data from urban infrastructure management departments to construct a three-dimensional vector diagram of the underground pipeline network data; By integrating building information modeling, civil defense engineering drawings, and underground space surveying data, a three-dimensional map of the underground public space data is constructed. Acquire multi-temporal and multi-spectral satellite remote sensing image data as a dynamic base map and supplementary information source for the aforementioned basic geographic data of the Earth's surface; It can access IoT sensor networks deployed in buildings, pipelines, geological monitoring points, and urban environments to acquire the IoT dynamic monitoring data in real time or near real time.

3. The method for multi-source heterogeneous data fusion of a city-level integrated 2D / 3D model according to claim 2, characterized in that, The subsurface 3D geological map, represented by the Grid / Tin hybrid data model, includes: For the aforementioned underground three-dimensional geological map, in the layered geological body areas with relatively uniform and regular properties and structures within the model, a regular grid data structure is used for representation; In areas with complex structures or requiring precise geometric depiction, such as model boundaries, stratigraphic contact surfaces, faults, and intrusions, an irregular triangular mesh data structure is used for representation. By using a preset data structure and indexing mechanism, the regular grid and the irregular triangular network are seamlessly connected to form the Grid / Tin hybrid data model.

4. The multi-source heterogeneous data fusion method for a city-level integrated 2D / 3D model according to claim 3, characterized in that, The acquired multi-source heterogeneous data are spatiotemporally registered and unified to the same spatial coordinate system and temporal reference, forming a data set with a consistent spatiotemporal reference, including: A unified three-dimensional spatial reference framework is established with the urban plane coordinate system and elevation benchmark as the core. The real-scene three-dimensional map, basic surface geographic data, underground pipe network data, underground public space data and satellite remote sensing imagery are uniformly transformed into the unified three-dimensional spatial reference framework through the coordinate transformation parameter method. For the underground three-dimensional geological map and BIM model stored in relative coordinates, three-dimensional spatial registration and mosaicking are performed using their known control points, feature points or their association with the above-ground model, so as to accurately anchor them into the three-dimensional unified reference frame. The dynamic monitoring data is assigned a unified timestamp, and its spatial location information is associated and bound with the corresponding entity under the unified three-dimensional spatial reference frame. For multi-time series data such as satellite remote sensing data, a unified time series benchmark is established, and data from different time phases are aligned and organized according to the acquisition time to form a data cube with a time dimension. After unifying the spatial reference, the spatial topology relationships of adjacent or related data layers are checked and corrected to form a data set consistent with the spatiotemporal reference.

5. The method for multi-source heterogeneous data fusion of a city-level integrated 2D / 3D model according to claim 1, characterized in that, Construct a city-level, full-space digital map covering both above-ground and underground areas, integrating two-dimensional and three-dimensional elements, including: Based on the integrated fusion framework, a true three-dimensional form and spatial relationship of urban surface topography, above-ground buildings, underground geological structure, underground pipe network and underground space are constructed; The semantic association graph is used to establish multi-level attribute information, business rules and dynamic association relationships covering all types of entities in the space. Through real-time data channels, real-time monitoring data from the Internet of Things is continuously integrated and visualized to reflect the dynamic changes in urban operations; Obtain spatiotemporal benchmarks and topological relationships consistent with internal data, and construct a city-level full-space digital map based on a unified spatiotemporal data foundation.

6. The multi-source heterogeneous data fusion method for a city-level integrated 2D / 3D model according to claim 5, characterized in that, Based on the aforementioned city-level full-space digital map, a thematic layer for urban management is constructed to provide a visualization framework, including: Based on the city-level full-space digital map, the full-space data is logically sliced ​​and physically organized according to a preset thematic classification system to form multiple independently managed and overlayable thematic data layers; the thematic classification system includes at least eight categories: administrative boundaries, roads and traffic, buildings, place names and addresses, thematic facilities, vegetation cover, water systems and lakes, and terrain slope. For specific business scenarios in urban management, thematic layers containing real-time status information are dynamically generated by calling the semantic association graph and real-time data channels. Spatial analysis algorithms are used to deeply integrate with the city-level full-space digital map for visualization, enabling real-time rendering in a two-dimensional and three-dimensional integrated scene.

7. The method for multi-source heterogeneous data fusion of a city-level integrated 2D / 3D model according to claim 6, characterized in that, The spatial analysis algorithm is used to perform deep visualization integration with the city-level full-space digital map, including: In the integrated 2D and 3D scene, a graphical interactive tool is provided, allowing users to define the parameters required for analysis directly in the 3D scene by clicking, dragging, and drawing; Based on the defined parameters, the corresponding spatial analysis service is invoked in real time to perform calculations, and the intermediate process and the final result are rendered in real time to the integrated 2D and 3D scene. The profile analysis results are generated in real time and dynamically drawn along the profile path to create a two-dimensional profile map containing above-ground buildings, underground pipe networks and geological structures, which is then highlighted in conjunction with the three-dimensional scene. The results of the vision analysis are used to draw the line of sight and occlusion in three-dimensional space in real time using visible and invisible visual codes. The buffer analysis results are generated and overlaid in real time in three-dimensional space in the form of a semi-transparent three-dimensional volume.

8. The method for multi-source heterogeneous data fusion of a city-level integrated 2D / 3D model according to claim 7, characterized in that, Integrating spatial analysis algorithms, it supports the construction and analysis of business scenarios for urban management, including: Based on the full-space geometry and topology of the integrated two-dimensional and three-dimensional model, the structural characteristics of urban space are analyzed, and combined with real-time monitoring data and physical laws, urban dynamic processes are simulated and predicted. Based on the semantic association graph and historical data, intelligent reasoning and situational assessment are performed; When a facility triggers an alarm, it automatically analyzes other facilities that are directly or indirectly related to it in the semantic association graph to assess the risk of cascading effects. The system calculates the operational health of the region in real time and displays it as a visual heatmap.

Images