Air-ground integrated big data fusion method for urban measurement

Abstract

The invention discloses an air-ground integrated big data fusion method for urban measurement, and relates to the technical field of data processing. The method comprises the following steps: acquiring space-land integrated big data of a city; uniformly calibrating the air-ground integrated big data to a city standard coordinate system to obtain low space-time deviation data; obtaining a first conflict entity and a second conflict entity based on the low space-time deviation data; correcting spatial positions of the first conflict entity and the second conflict entity in the low space-time deviation data to obtain a conflict-free big data model; and constructing a long-short-term memory network based on the conflict-free big data model. According to the method, the problem of low fusion precision caused by non-uniform space-time reference and large format precision difference of multi-source data is solved by integrating space-ground integrated big data such as urban space point cloud, a sensor network and a remote sensing image, and the data consistency and reliability are improved and information faults and semantic conflicts are avoided by eliminating space-time deviation and entity conflicts.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and specifically to an air-ground integrated big data fusion method for urban measurement. Background Art

[0002] In today's urban development and planning, comprehensive and accurate city measurement is crucial for achieving refined management, precise decision-making, and efficient resource allocation. Urban measurement requires comprehensive consideration of multiple factors, including a city's spatial structure, functional layout, population mobility, and infrastructure. With the advancement of technology, single-source data is no longer sufficient to meet the complex demands of urban measurement. Fusion of data collected from various platforms, including the air and ground, can provide more comprehensive and accurate urban information, enabling deeper insights into urban operating patterns and development trends, and providing strong support for sustainable urban development.

[0003] However, due to the lack of unified spatiotemporal benchmarks across air-ground data collection platforms, significant differences in data format and accuracy exist, and there is a lack of effective mechanisms for feature alignment and semantic association of multi-source heterogeneous data. This results in information gaps and semantic conflicts in the fused data. For example, there are deviations in time synchronization and spatial coordinate conversion between aerial remote sensing data and ground sensor data, making it difficult to form a complete spatiotemporal sequence for data on urban spatial structure and infrastructure. Furthermore, the lack of real-time processing capabilities for massive amounts of data prevents the rapid integration of multi-source data to construct urban dynamic models, leading to fragmented urban vital sign diagnosis and making it difficult to support accurate decision-making in critical scenarios such as rainstorm waterlogging warnings and public emergency response. Summary of the Invention

[0004] The purpose of this application is to provide an air-ground integrated big data fusion method for urban measurement, so as to solve the technical problem that special events in cities cannot be accurately identified due to the low accuracy of air-ground integrated big data fusion.

[0005] To achieve the above objectives, this application provides the following technical solutions: A space-ground integrated big data fusion method for urban measurement, including: Acquire the city's integrated air-land big data; the integrated air-land big data includes at least the city's spatial point cloud data, the city's sensor network data, and the city's remote sensing image data; The integrated space-ground big data is uniformly calibrated to a city standard coordinate system to obtain data with low spatiotemporal deviation; the city standard coordinate system is at least used to eliminate the spatiotemporal deviation of different source data in the integrated space-ground big data; Based on the low spatiotemporal deviation data, a first conflicting entity and a second conflicting entity are obtained; the first conflicting entity is any entity in the low spatiotemporal deviation data; the second conflicting entity is any entity in the low spatiotemporal deviation data that spatially overlaps with the first conflicting entity; the first conflicting entity and the second conflicting entity have different data sources; Correcting the spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data to obtain a conflict-free large data model; Based on the conflict-free big data model, a long short-term memory network is constructed; the long short-term memory network is used to identify special events in the city.

[0006] As a specific solution in the technical solution of this application, the acquisition of urban space-land integrated big data includes: Scanning an urban area using a drone equipped with a laser radar to obtain spatial point cloud data of the city; Acquiring remote sensing image data of the city based on low-altitude remote sensing satellites; Based on fixed sensors deployed on major roads and public areas in the city, sensor network data of the city is obtained.

[0007] As a specific solution in the technical solution of this application, the city standard coordinate system includes a primary reference network, a secondary structure network, a third-level coverage network and a fourth-level dynamic network; The primary benchmark network is a network composed of multiple benchmark points in the urban space; the benchmark points are deeply buried municipal benchmark points or anchor points of cross-river bridges; at least three benchmark points are arranged in each administrative unit in the city to form a spatial triangle constraint; The secondary structure network is a spatial point cloud network composed of landmark buildings in the urban space; The three-level coverage network is the spatial position network of buildings in the first-level reference network and the second-level structure network extracted from the remote sensing image data based on an image feature extraction algorithm; The four-level dynamic network is a local coordinate system network constructed based on the synchronous positioning and map construction technology with each fixed sensor as an anchor point.

[0008] As a specific solution in the technical solution of this application, the unified calibration of the air-ground integrated big data to the city standard coordinate system to obtain low spatiotemporal deviation data includes: Mapping each spatial point cloud in the secondary structure network to the primary reference network to obtain a first fusion network; Based on the first fusion network, a plurality of elevation columns are obtained; the elevation columns are used to represent corresponding buildings in the city; Based on each elevation column, obtaining building projection information corresponding to each elevation column; the projection direction of the building projection is vertically downward; the building projection information at least includes the horizontal distance between each building projection; Based on the three-level coverage network, a projection offset of each building projection is obtained; the projection offset is used to at least represent the horizontal distance difference between the building projection and the corresponding building in the three-level coverage network; Based on the projection offset, adjusting the horizontal position of each elevation column in the first fusion network to obtain a second fusion network; Based on the four-level dynamic network, obtaining motion trajectory information of targets in each local coordinate system network; the targets include pedestrians and vehicles; The motion trajectory information is back-projected to the second fusion network to obtain the low spatiotemporal deviation data.

[0009] As a specific solution in the technical solution of the present application, mapping each spatial point cloud in the secondary structure network to the primary reference network to obtain a first fusion network includes: Based on each spatial point cloud, multiple outliers are obtained; Delete each outlier point from each spatial point cloud to obtain multiple corrected spatial point clouds; Based on the iterative closest point algorithm, each reference point and each corrected space point cloud are formed into a plurality of matching point pairs; in the matching point pairs, each reference point is a source point set, and each corrected space point cloud is a target point set; Based on each matching point pair, a first fusion network is obtained.

[0010] As a specific solution in the technical solution of this application, the method of obtaining multiple elevation columns based on the first fusion network includes: Obtaining the surface curvature of each cloud point in the first fusion network based on Delaunay triangulation technology; Based on the surface curvature of each cloud point, multiple surface fitting results are obtained; Based on each surface fitting result, multiple elevation columns are obtained.

[0011] As a specific solution in the technical solution of this application, obtaining the projection offset of each building projection based on the three-level coverage network includes: Based on the projections of each building, a target building projection is obtained; the target building projection is any building projection for which a projection offset has not been obtained among the projections of each building; Based on the projection of the target building, a fixed offset is obtained; the fixed offset is the offset of the projection of the target building in the three-level coverage network; Based on the fixed offset, a projection offset of the target building projection is obtained.

[0012] As a specific solution in the technical solution of the present application, obtaining the first conflicting entity and the second conflicting entity based on the low spatiotemporal deviation data includes: Based on the low spatiotemporal deviation data, a first entity and a second entity are obtained; the first entity is any entity in the low spatiotemporal deviation data; the second entity is any entity in the low spatiotemporal deviation data that spatially overlaps with the first entity; and the first entity and the second entity have different data sources; Slicing the first entity and the second entity according to a spatial grid and a time window, and obtaining a spatial overlap area and a spatial overlap duration; the spatial overlap area is an area where the first entity and the second entity spatially overlap; the spatial overlap duration is a duration where the first entity and the second entity spatially overlap; Based on the spatial overlap area and the spatial overlap duration, obtaining a spatiotemporal overlap value; the spatiotemporal overlap value is used to at least characterize the size of the area and duration of the spatial overlap between the first entity and the second entity; If the spatiotemporal overlap value satisfies a preset condition, the first entity and the second entity are respectively used as a first candidate entity and a second candidate entity; the preset condition includes that the spatiotemporal overlap value is less than a first preset value; A first conflicting entity and a second conflicting entity are acquired based on the first candidate entity and the second candidate entity.

[0013] As a specific solution in the technical solution of the present application, obtaining the first conflicting entity and the second conflicting entity based on the first candidate entity and the second candidate entity includes: Based on different data sources in the low spatiotemporal deviation data, a plurality of confidence probabilities are obtained; the confidence probabilities are probabilities that the first candidate entity and the second candidate entity are two conflicting entities in the corresponding data sources; Based on the trust function theory, the conflict probability and the non-conflict probability are obtained from each confidence probability; the sum of the conflict probability and the non-conflict probability is equal to 100%; Based on the conflict probability and the non-conflict probability, obtaining a probability difference; the probability difference is equal to the conflict probability minus the non-conflict probability; If the probability difference is greater than a second preset value, the first candidate entity is determined to be a first conflicting entity, and the second candidate entity is determined to be a second conflicting entity.

[0014] As a specific solution in the technical solution of the present application, the spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data are corrected to obtain a conflict-free big data model, including: Using the position parameters of the first conflicting entity and the second conflicting entity as particle dimensions, randomly generating a particle swarm within an allowed adjustment range, each particle representing a potential correction position; Calculate the three-dimensional Euclidean distance between the corrected position corresponding to each particle and the positions of the first conflicting entity and the second conflicting entity in each data source, and perform weighted averaging based on the credibility of each data source obtained in the evidence fusion stage to obtain the distance error; Obtain the nearest triangle edge or triangle face in the Delaunay triangulation network for the corrected position of each particle, calculate the vertical distance from the corrected position to the triangle edge or triangle face, and if the corrected position belongs to a preset category entity and the vertical distance exceeds the allowable offset threshold of the category entity, then use the vertical distance as the topology violation degree; Constructing a fitness function based on the distance error and the topology violation; the fitness function is the sum of the distance error and the topology violation; Based on the fitness function, outputting the optimal position as a correction result; The spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data are corrected based on the correction result to obtain a conflict-free large data model.

[0015] Compared with the prior art, the present invention has the following advantages: This application integrates urban spatial point clouds, sensor networks, remote sensing images, and other integrated space-ground big data to address the low fusion accuracy issues caused by inconsistent spatiotemporal benchmarks and large differences in format accuracy across multiple data sources. By eliminating spatiotemporal biases and entity conflicts, it improves data consistency and reliability, avoiding information gaps and semantic conflicts. Furthermore, big data models built on conflict-free data can accurately identify special events, providing support for scenarios such as urban rainstorm and flood warnings and public emergency response. This helps facilitate refined urban management, precise decision-making, and efficient resource allocation, meeting the measurement needs of complex cities. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of an air-ground integrated big data fusion method for urban measurement proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0018] The terms "first", "second", etc. in the description of the embodiments of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. For example, the first conflicting entity and the second conflicting entity proposed below belong to different entities. It should be understood that the names used in this way can be interchanged where appropriate so that the embodiments described herein can be implemented in an order other than that shown or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or modules is not necessarily limited to those steps or modules clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. The division of modules that appears in the embodiments of the present application is only a logical division. In actual applications, there may be other division methods. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not performed. In addition, the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, and the indirect coupling or communication connection between modules can be electrical or other similar forms, which are not limited in the embodiments of the present application. Moreover, the modules or sub-modules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed into multiple circuit modules, and some or all of the modules may be selected according to actual needs to achieve the purpose of the embodiment of the present application.

[0019] To address the technical problem raised in the background art that due to the low accuracy of integrated air-ground big data fusion, it is impossible to accurately identify special urban events, this application proposes an integrated air-ground big data fusion method for urban measurement. The integrated air-ground big data fusion method for urban measurement includes steps S100 to S500.

[0020] Step S100: Obtaining the city's space and land integrated big data.

[0021] In this embodiment, the air-ground integrated big data includes at least the spatial point cloud data of the city, the sensor network data of the city, and the remote sensing image data of the city.

[0022] In the embodiments of this application, there are no restrictions on the sources of integrated air-ground big data. For example, spatial point cloud data can be collected through airborne / ground laser scanning by professional teams, or obtained from government platforms, commercial companies, and scientific research sharing channels. Sensor network data can come from municipally deployed transportation, environmental protection, and other sensors, enterprise-operated IoT devices, and information open to public data platforms. Remote sensing image data includes free satellite imagery (such as Landsat and Sentinel), commercial satellite data, drone aerial photography, and aerial photogrammetry results. This data can be either open source information that can be downloaded for free, or data read from databases, obtained through collaboration, or purchased.

[0023] To accurately acquire the three core data types of integrated air-ground data (i.e., urban spatial point cloud data, urban sensor network data, and urban remote sensing image data) through clear and operational technical means, ensuring the pertinence and effectiveness of data sources and providing high-quality raw data for subsequent fusion processing, in one embodiment of the present application, step S100, acquiring the integrated air-ground data of the city, includes steps S110 to S130.

[0024] Step S110: Scanning the urban area based on the drone equipped with the laser radar to obtain spatial point cloud data of the city.

[0025] In this embodiment, a multi-rotor drone or fixed-wing drone equipped with a laser radar (e.g., LiDAR) can be used. Based on the terrain complexity and accuracy requirements of the urban area, a flight route (e.g., a grid or ribbon pattern) can be planned, along with a set flight altitude (typically 50-200 meters, with lower altitudes for higher accuracy requirements) and a scanning frequency (e.g., 100-500 points / m2). During flight, the LiDAR emits laser pulses toward the ground, receives reflected signals, calculates distance, and generates point cloud data containing three-dimensional (X, Y, Z) coordinates covering surface features such as buildings, roads, and vegetation. After the flight, point cloud processing software (e.g., Cloud Compare) is used to stitch the data together and reduce noise.

[0026] It should be understood that ground images contain rich visual information (e.g., building facade textures, road signs, vegetation colors, etc.), which can be combined with spatial point cloud data collected by drone lidar to help identify the semantic attributes of each entity in the point cloud data (e.g., distinguishing between buildings, roads, vegetation, etc.), thereby improving the interpretability of the point cloud data. Ground images can also provide texture information for the three-dimensional model constructed from point cloud data, making the model more consistent with the real scene. At the same time, for details that may be missed in point cloud data (e.g., small facilities, surface signs, etc.), ground images can serve as a supplementary data source, enhancing the ability of integrated air-ground big data to depict local details of the city, and providing more comprehensive basic data for subsequent fusion steps such as calibration of the city's standard coordinate system and entity conflict detection. Based on this, in this embodiment, ground images of the city can also be obtained based on drones equipped with cameras.

[0027] Specifically, in order to obtain ground images of the city, a high-resolution (for example, more than 20 million pixels) RGB camera can be installed on the drone to capture ground images at a preset overlap rate (for example, a heading overlap rate of more than 70% and a lateral overlap rate of more than 60%) to obtain detailed information such as building facades, road signs, pedestrians and vehicles, which can be used to assist in semantic recognition or texture mapping of point cloud data.

[0028] Step S120: acquiring remote sensing image data of the city based on low-altitude remote sensing satellites.

[0029] In this embodiment, a low-altitude remote sensing satellite with an orbital altitude below 1000 kilometers (e.g., Gaofen-7 or WorldView-3) can be selected. The image resolution (e.g., 0.5 to 5 meters, with fine-grained urban measurements requiring resolutions within 1 meter) and spectral bands (visible light, near-infrared, etc., for distinguishing vegetation, buildings, and water bodies) can be selected as needed. A data subscription request is submitted through the satellite operator (e.g., the Resource Satellite Application Center or Digital Globe), specifying the latitude and longitude range of the urban area and the capture time (avoiding periods of cloud cover). After the satellite captures the image as instructed, it generates a remote sensing image containing geographic coordinates. After pre-processing such as radiometric and geometric correction, it is provided to the user. Users can further process the image using specialized software (e.g., ENVI or ERDAS) to extract macroscopic features such as urban land use types and building density.

[0030] Step S130: Acquire sensor network data of the city based on fixed sensors deployed on main roads and public areas of the city.

[0031] In this embodiment, different types of fixed sensors can be deployed according to functional requirements at major urban intersections, bridges, squares, parks, and other public areas. For example, radar sensors or video analysis equipment can be installed to collect real-time data such as traffic flow, speed, and congestion status; PM2.5 sensors, noise sensors, and temperature and humidity sensors can be deployed to regularly upload monitoring data; and IoT sensors can be embedded in streetlights, trash cans, and other equipment to collect information such as energy consumption and usage status. Sensors can be connected to the city data center via wired or wireless networks. Data is stored in real time by timestamp, forming a structured database (e.g., MySQL) or a time series database (e.g., InfluxDB) for subsequent fusion processing and reflection of the city's dynamic operating status.

[0032] Step S200: uniformly calibrating the air-ground integrated big data to the city standard coordinate system to obtain low spatiotemporal deviation data.

[0033] In this embodiment, the city standard coordinate system is used to at least eliminate the spatiotemporal deviation of different source data in the space-ground integrated big data. Based on this, in this embodiment, the city standard coordinate system includes a primary reference network, a secondary structure network, a third coverage network, and a fourth dynamic network.

[0034] In this embodiment, the primary reference network is a network composed of multiple reference points in the urban space. The reference points are deeply buried municipal reference points or anchor points of cross-river bridges. At least three reference points are arranged in each administrative unit in the city to form a spatial triangle constraint. It should be understood that the spatial triangle constraint is a technical method in the fields of surveying and mapping, geographic information or engineering measurement that uses the geometric properties of triangles to locate and control the stability of spatial points. Specifically, in three-dimensional space, three non-collinear points can form a triangle, which has a uniquely determined spatial position and geometric shape (for example, side length, angle and area, etc.). By utilizing this characteristic, a stable "spatial reference frame" can be formed by accurately measuring the three reference points (that is, the three-dimensional coordinates of each reference point), that is, the spatial triangle constraint.

[0035] In this embodiment, the secondary structure network is a spatial point cloud network formed by landmark buildings in the urban space.

[0036] In this embodiment, the three-level coverage network is the spatial position network of buildings in the first-level reference network and the second-level structure network extracted from the remote sensing image data based on an image feature extraction algorithm.

[0037] In this embodiment, the four-level dynamic network is a local coordinate system network constructed based on the simultaneous positioning and mapping technology with each fixed sensor as an anchor point.

[0038] In this embodiment, step S200 calibrates the integrated air-ground big data to the standard city coordinate system, thereby obtaining data with low spatiotemporal deviation. The core goal is to eliminate spatiotemporal deviations caused by differences in data collection platforms (e.g., drones, near-Earth satellites, etc.) between different sources, laying the foundation for subsequent data fusion, conflict resolution, and event identification. Therefore, step S200, which calibrates the integrated air-ground big data to the standard city coordinate system to obtain data with low spatiotemporal deviation, includes steps S210 through S270.

[0039] Step S210: Mapping each spatial point cloud in the secondary structure network to the primary reference network to obtain a first fusion network.

[0040] In this embodiment, the first fused network is formed by mapping the spatial point clouds in the secondary structure network to the primary reference network. The core function of step S210 is to achieve spatial alignment of local high-precision data with the global reference frame, providing a unified spatial foundation for subsequent data fusion. Based on this, step S210 maps each spatial point cloud in the secondary structure network to the primary reference network to obtain the first fused network, which includes steps S211 to S214.

[0041] Step S211: Based on each spatial point cloud, obtain multiple outlier points.

[0042] In this embodiment, outliers are points in each spatial point cloud that differ significantly from the distribution of surrounding point clouds and do not conform to normal spatial characteristics. These outliers are often caused by unstable features (e.g., antennas on top of buildings, temporary objects, etc.) or data acquisition errors. They can interfere with the precise matching of the secondary structure network with the primary reference network. Therefore, they need to be deleted in step S212 to ensure the reliability of the subsequent correction of the spatial point cloud.

[0043] In this embodiment, any reasonable method can be used to obtain outliers. For example, when processing a spatial point cloud, a circle with a radius of 10 meters (in other embodiments, 5 meters or 15 meters) is used as the neighborhood, with any cloud point as the center. The average distance between each cloud point and the other cloud points in its neighborhood is calculated. If the average distance corresponding to a cloud point is greater than three times the average of the average distances corresponding to all cloud points, the cloud point is determined to be an outlier.

[0044] Step S212: Delete each outlier point from each spatial point cloud to obtain a plurality of corrected spatial point clouds.

[0045] It should be noted that deleting certain cloud points (i.e., outliers) from the corresponding spatial point cloud is a mature technology and will not be described in detail here.

[0046] Step S213: forming a plurality of matching point pairs between each reference point and each corrected space point cloud based on an iterative closest point algorithm.

[0047] In this embodiment, each reference point in the matching point pair is a source point set, and each corrected space point cloud is a target point set.

[0048] It's important to note that the Iterative Closest Point (ICP) algorithm is a classic 3D point cloud registration algorithm. Its core goal is to find the optimal rigid body transformation (i.e., rotation matrix and translation vector) between two point clouds from different viewpoints through iterative optimization, ensuring that they coincide as closely as possible after alignment. Based on this, in this embodiment, step S213, which uses the Iterative Closest Point algorithm to form multiple matching point pairs between each reference point and each corrected spatial point cloud, includes steps S213a through S213c.

[0049] Step S213a: Based on each reference point and each corrected space point cloud, obtain multiple iterative matching point pairs.

[0050] In this embodiment, the iterative matching point pair is a point pair formed by the reference point and the cloud point closest to each corrected space point cloud.

[0051] Step S213b: Based on all iterative matching point pairs, the objective function is solved by the least squares method to obtain the rotation matrix and translation vector that minimize the sum of the squares of the distances between the source point set and the target point set.

[0052] In this embodiment, the objective function formula is:

[0053] Where n represents the number of iterative matching point pairs; Represents the coordinates corresponding to the points in the target point set in the i-th iterative matching point; Represents the coordinates corresponding to the points in the source point set in the i-th iterative matching point; represents the rotation matrix; represents the translation vector; It means to find the square of the distance between two coordinate points; It means finding the rotation matrix and translation vector that minimizes the sum of the squares of the distances between the source point set and the target point set.

[0054] Step S213c: Transform the source point set according to the calculated rotation matrix and translation vector, re-find the nearest point and update the iterative matching point pair, and repeat the calculation process of the above step S213b until the change in the sum of the squares of the distances of the two iterations is less than the set threshold or the iteration reaches the maximum number of iterations, at which time a stable matching point pair is obtained.

[0055] In the embodiment of the present application, the above threshold value can be set according to requirements, for example, the above threshold value can be 0.1 or 0.2.

[0056] Step S214: Obtain a first fusion network based on each matching point pair.

[0057] In this embodiment, through iterative optimization of the ICP algorithm, multiple precise matching point pairs are ultimately generated between the source point set (i.e., each reference point) and the target point set (i.e., each corrected spatial point cloud). This provides a spatial transformation basis for subsequently mapping the secondary structure network onto the primary reference network to form the first fused network. In other words, in this embodiment, the first fused network is obtained by aligning each reference point in the primary reference network with each matching point in the secondary structure network.

[0058] Step S220: Based on the first fusion network, obtain multiple elevation columns.

[0059] In this embodiment, the elevation columns are used to characterize the corresponding buildings in the city. It should be noted that the purpose of obtaining the elevation columns is to provide an accurate three-dimensional reference for correcting the projection distortion (i.e., projection offset) of the satellite image in step S240. In the embodiments of the present application, multiple elevation columns can be obtained based on the first fusion network using any reasonable method. For example, step S220, which obtains multiple elevation columns based on the first fusion network, includes steps S221 to S223.

[0060] Step S221: Obtain the surface curvature of each cloud point in the first fusion network based on the Delaunay triangulation technology.

[0061] It's important to understand that Delaunay triangulation (DT) is a point-based geometric subdivision technique widely used in fields such as geographic information systems, computer graphics, and finite element analysis. Its core principle is to efficiently represent the topological relationships of spatial point sets through a network of triangles that meet specific geometric criteria. Delaunay triangulation is a mature technique for obtaining the surface curvature of each cloud point in the first fused network and will not be discussed in detail here.

[0062] Step S222: obtaining a plurality of surface fitting results based on the surface curvature of each cloud point.

[0063] In this example, after the triangulation network is constructed, a minimum curvature surface fitting method is used. The curvature of the surface at each feature point is calculated by taking the weighted average of the curvatures of all triangles in the neighborhood of that feature point. This ultimately yields a surface fitting result for the urban framework surface. This process ensures that the elevation and spatial position of the elevation columns are consistent with the actual buildings.

[0064] Step S223: Based on the surface fitting results, multiple elevation columns are obtained.

[0065] In this embodiment, the center point coordinates of the elevation column correspond one-to-one to the geometric center coordinates of the bottom of the building, so that each elevation column uniquely represents the corresponding building in the city.

[0066] Step S230: Based on each elevation column, obtain building projection information corresponding to each elevation column.

[0067] In this embodiment, the projection direction of the building projection is vertically downward. The building projection information at least includes the horizontal distance between each building projection.

[0068] Step S240: Based on the three-level coverage network, the projection offset of each building projection is obtained.

[0069] In this embodiment, the projection offset is at least used to characterize the size of the horizontal distance difference between the building projection and the corresponding building in the three-level coverage network. That is to say, in this embodiment, the purpose of executing step S240 is to obtain the size of the horizontal distance difference between the building projection and the corresponding building in the three-level coverage network, and then provide key data support for the subsequent adjustment of the building projection position and the realization of more accurate data fusion. In the process of air-ground integrated big data fusion of urban measurement, this step plays an important role in calibrating satellite images and ensuring the spatial consistency of data from different data sources. Based on this, step S240, based on the three-level coverage network, obtains the projection offset of each building projection, including steps S241 to S243.

[0070] Step S241: Based on the projections of each building, a target building projection is obtained.

[0071] In this embodiment, the target building projection is any building projection among the various building projections for which a projection offset has not been obtained. That is, in this embodiment, the projection offset of any building projection among the various building projections is obtained in the same manner as the projection offset of the target building projection.

[0072] Step S242: Obtain a fixed offset based on the target building projection.

[0073] In this embodiment, the fixed offset is the offset of the target building projected in the three-level coverage network. Specifically, the calculation formula of the fixed offset is as follows:

[0074] in, Indicates the fixed offset corresponding to the projection of the target building; Indicates the elevation of the elevation column corresponding to the projection of the target building; Indicates the satellite shooting angle; represents the tangent function; Indicates the satellite image scale.

[0075] Step S243: Based on the fixed offset, obtain the projection offset of the target building projection.

[0076] In this embodiment, step S243, obtaining the projection offset of the target building projection based on the fixed offset, includes step S243a and step S243b.

[0077] Step S243a: Obtain the projection offset direction.

[0078] In this embodiment, the projection offset direction is along the radial direction from the building outline point in the satellite image to the center point of the corresponding elevation column. This direction is determined by the central projection principle of satellite imagery, ensuring that the offset compensation is consistent with the physical cause of actual projection distortion.

[0079] Step S243b: Based on the fixed offset and the projection offset direction, obtain the projection offset of the target building projection. Specifically, based on the fixed offset and the projection offset direction, the calculation formula for obtaining the projection offset of the target building projection is as follows:

[0080] in, Indicates the projection offset corresponding to the projection of the target building; Indicates the fixed offset corresponding to the projection of the target building; Represents the vector from the building outline point in the satellite image to the center point of the elevation column in three-dimensional space (hereinafter referred to as the projection vector); represents the magnitude of the projection vector; represents the anti-zero coefficient, Can be any positive number close to zero, for example, It can be 0.01 or 0.001, etc.

[0081] Step S250: Based on the projection offset, adjust the horizontal position of each elevation column in the first fusion network to obtain a second fusion network.

[0082] In this embodiment, the second fusion network can be obtained by adjusting the horizontal position of each elevation column in the first fusion network according to the projection offset.

[0083] Step S260: Based on the four-level dynamic network, the motion trajectory information of the target in each local coordinate system network is obtained.

[0084] In this embodiment, the targets include pedestrians and vehicles.

[0085] Step S270: Back-projecting the motion trajectory information to the second fusion network to obtain the low spatiotemporal deviation data.

[0086] It's important to note that while low-temporal-spatial-bias data eliminates the temporal-spatial bias in integrated space-ground big data, it's still insufficient for supporting semantic understanding of the city model. This is because conflicting records of the same entity may exist in different data sources, and eliminating entity conflicts is necessary to enhance the model's event interpretability. Therefore, step S300 must be executed.

[0087] Step S300: Acquire a first conflicting entity and a second conflicting entity based on the low spatiotemporal deviation data.

[0088] In this embodiment, the first conflicting entity is any entity in the low spatiotemporal deviation data. The second conflicting entity is any entity in the low spatiotemporal deviation data that spatially overlaps with the first conflicting entity. The first conflicting entity and the second conflicting entity have different data sources.

[0089] It should be clear that in the embodiments of the present application, any reasonable method can be used to obtain the first conflict entity and the second conflict entity based on the low spatiotemporal deviation data. For example, any two spatial overlapping entities whose spatial overlapping area or spatial overlapping duration in the low spatiotemporal deviation data is less than the third preset value can be used as the first conflict entity and the second conflict entity. For example, if it is determined whether any two spatial overlapping entities are the first conflict entity and the second conflict entity based on the spatial overlapping area, the third preset value can be 0.3 times or 0.2 times the maximum spatial overlapping area (that is, the maximum spatial overlapping area that can be formed by the above two spatial overlapping entities), etc.; in this embodiment, if it is determined whether any two spatial overlapping entities are the first conflict entity and the second conflict entity based on the spatial overlapping duration, the third preset value can be 10 minutes or 20 minutes, etc.

[0090] In order to accurately obtain the first conflicting entity and the second conflicting entity, in one embodiment of the present application, step S300, based on the low spatiotemporal deviation data, obtains the first conflicting entity and the second conflicting entity, including steps S310 to S350.

[0091] Step S310: Acquire a first entity and a second entity based on the low spatiotemporal deviation data.

[0092] In this embodiment, the first entity is any entity in the low spatiotemporal deviation data, and the second entity is any entity in the low spatiotemporal deviation data that spatially overlaps with the first entity, and the first entity and the second entity have different data sources.

[0093] Step S320: Slice the first entity and the second entity according to the spatial grid and time window, and obtain the spatial overlapping area and spatial overlapping duration.

[0094] In this embodiment, the spatial overlap area is the area where the first entity and the second entity spatially overlap, and the spatial overlap duration is the duration where the first entity and the second entity spatially overlap.

[0095] It's important to note that for the same urban entity (e.g., a building, a cracked road, or flooded areas), records from different data sources (e.g., drone point clouds, satellite imagery, and ground-based sensor data) should theoretically have a high degree of temporal and spatial overlap. This is because they describe the same objectively existing urban feature. For example, a building scanned by a drone and captured by satellite imagery at the same time should roughly match in spatial location and time, resulting in a high degree of overlap. When different data sources show low overlap for what should theoretically be the same entity, it indicates significant discrepancies in the multi-source data's descriptions of that entity. This could be due to a recognition error in one data source (e.g., satellite imagery misidentifying a building's outline), semantic misalignment (e.g., different data sources defining "road" differently), or the misassociation of entities that are actually different. This lack of temporal and spatial consistency violates the fundamental assumption that multi-source data on the same entity should converge, and therefore can be preliminarily identified as an anomaly.

[0096] Step S330: Based on the spatial overlap area and the spatial overlap duration, obtain a spatiotemporal overlap value.

[0097] In this embodiment, the spatiotemporal overlap value is used to at least characterize the size and duration of the spatial overlap between the first entity and the second entity.

[0098] In this embodiment, any reasonable method can be used to obtain the spatiotemporal overlap value based on the spatial overlap area and the spatial overlap duration. For example, the spatiotemporal overlap value can be the sum or product of the values ​​of the spatial overlap area and the spatial overlap duration. In a specific embodiment of the present application, step S330: Based on the spatial overlap area and the spatial overlap duration, the calculation formula for obtaining the spatiotemporal overlap value is as follows:

[0099] in, represents the spatiotemporal overlap value corresponding to the first entity and the second entity; represents the average value of the spatial overlapping area of ​​the first entity and the second entity in each spatial slice; represents the maximum value of the spatial overlapping area of ​​the first entity and the second entity in each spatial slice; Indicates the duration of spatial overlap between the first entity and the second entity; Indicates the total duration of the low spatiotemporal deviation data. and There is no possibility of being zero. If there is an extreme case, a non-zero constant, such as 0.001, is added to the denominator of the fraction.

[0100] Step S340: If the spatiotemporal overlap value satisfies a preset condition, the first entity and the second entity are respectively used as a first candidate entity and a second candidate entity.

[0101] In this embodiment, the larger the spatiotemporal overlap value corresponding to the first entity and the second entity, the more likely it is that the first entity and the second entity are the same entity, and there is no anomaly between the two entities; the smaller the spatiotemporal overlap value corresponding to the first entity and the second entity, the more likely it is that the first entity and the second entity are not the same entity, and there is an anomaly between the two entities. That is, in this embodiment, the preset condition can be set according to needs. For example, the preset condition can be that the spatiotemporal overlap value is less than a first preset value. In the embodiment of the present application, the first preset value can be selected according to needs. For example, the first preset value can be 0.3, 0.4, or 0.5, etc.

[0102] Step S350: Based on the first candidate entity and the second candidate entity, obtain a first conflicting entity and a second conflicting entity.

[0103] As mentioned above, even if the first and second candidate entities meet the preset conditions, it's not certain that they conflict. Specifically, if either the first or second candidate entity is not a real entity but a virtual entity created by data anomalies, then the first and second candidate entities may not conflict. For example, drone lidar scanning may produce false point clouds due to haze, bird interference, or device vibration, which may be mistakenly identified by the algorithm as entities such as small buildings. Satellite imagery may also produce artifacts (such as light spots and streaks) due to cloud obstruction and sensor noise, resulting in the extraction of abnormal areas (such as road cracks) as entities. These candidate entities are merely artifacts of data noise and have no corresponding physical counterparts in the city. During feature extraction or clustering, algorithms may misclassify meaningless spatial point sets as entities due to improper parameters. For example, randomly distributed ground point clouds may be clustered as road cracks, or meaningless pixel blocks in satellite imagery may be misclassified as small facilities. These candidate entities are the result of algorithmic processing errors and have no corresponding physical counterparts. In other words, the purpose of step S350 is to accurately determine the true conflicting entity (i.e., the first conflicting entity and the second conflicting entity) from the candidate entities (i.e., the first candidate entity and the second candidate entity). Based on this, step S350, which obtains the first conflicting entity and the second conflicting entity based on the first candidate entity and the second candidate entity, includes steps S351 to S354.

[0104] Step S351: Acquire multiple confidence probabilities based on different data sources in the low spatiotemporal deviation data.

[0105] In this embodiment, the confidence probability can be the probability that the first candidate entity is a real entity, and / or the probability that the second candidate entity is a real entity. Alternatively, the confidence probability can be the probability that the first candidate entity and the second candidate entity are conflicting entities in the corresponding data source. For any candidate entity, drone point clouds can determine the probability of the entity's existence based on point cloud density and geometric features; satellite imagery can determine the probability of the entity's existence based on spectral characteristics and texture analysis; and ground sensors can determine the probability of the entity's existence through cross-validation combined with data from neighboring devices.

[0106] Step S352: Obtain conflict probability and non-conflict probability from each confidence probability based on the trust function theory.

[0107] In this embodiment, the sum of the conflict probability and the non-conflict probability is equal to 100%.

[0108] It's important to note that trust function theory quantifies the degree of support for a proposition from different pieces of evidence using basic probability assignments. It then synthesizes multi-source evidence through orthogonality and rules, ultimately judging the credibility of the proposition based on the fusion results. Its core approach is to address uncertainty rather than randomness, allowing for an unknown range of evidence support for a proposition. This makes it suitable for scenarios involving conflicting, heterogeneous data from multiple sources. For example, consider a scenario where a road section in a city may have cracks. Observations from drones, satellites, and ground sensors provide this information. Trust function theory is needed to determine whether the cracks actually exist.

[0109] First, extract evidence from each data source: Drone point cloud: The scan revealed that the point cloud density in this area is 20 points / m² (>15 points / m²). According to the rule, the probability of "there is a crack" is 0.8 (that is, the probability of "there is a crack" in the following text). , the remaining 0.2 is uncertain).

[0110] Satellite imagery: Analyzing the spectral characteristics of the area, we found that the reflectivity of a specific band was abnormal (which is consistent with the characteristics of a crack), and calculated the confidence level of "the existence of a crack" to be 0.7 (that is, the confidence level of the crack in the following text). , the remaining 0.3 is uncertain).

[0111] Ground sensor: Only a single sensor detected weak vibration, and the cross-validation was insufficient, so the probability of judging "there is a crack" was 0.6 (that is, , the remaining 0.4 is uncertain).

[0112] At this point, the evidence sets from the three data sources are: 、 、 .

[0113] Then, the evidence is fused using the belief function theory: The first step is to integrate drone and satellite evidence: Overlap coefficient :Both support "there are cracks", the intersection is significant, and the calculation is =0.8×0.7+0.8×0.3+0.2×0.7=0.56+0.24+0.14=0.94.

[0114] Conflict coefficient : There is no contradiction between the two (neither denies the existence of cracks). =0.

[0115] Fusion probability : = / (1- )=0.94 / 1=0.94.

[0116] Step 2: Integrate the above results with ground sensor evidence: Overlap coefficient : =0.94×0.6+0.94×0.4+0.06×0.6=0.98.

[0117] Conflict coefficient : Still no opposition, =0.

[0118] Final fusion probability : = / (1- )=0.98.

[0119] In this embodiment, the probability that there are cracks on the road is 0.98 (equivalent to the conflict probability mentioned above), and the probability that there are no cracks on the road is 0.02 (equivalent to the non-conflict probability mentioned above).

[0120] Step S353: Obtain a probability difference based on the conflict probability and the non-conflict probability.

[0121] In this embodiment, the probability difference is equal to the collision probability minus the non-collision probability. In the computer field, obtaining the difference (i.e., the probability difference) between two values ​​(i.e., the collision probability and the non-collision probability) is a mature technology and will not be described in detail here.

[0122] Step S354: If the probability difference is greater than a second preset value, the first candidate entity is determined to be a first conflicting entity, and the second candidate entity is determined to be a second conflicting entity.

[0123] It should be noted that in this embodiment, the larger the probability difference, the more likely the first candidate entity and the second candidate entity are conflicting entities; the smaller the probability difference, the more likely the first candidate entity and the second candidate entity are non-conflicting entities. In this embodiment, the second preset value can be set as needed, for example, the second preset value can be 0.2 or 0.3.

[0124] Step S400: correcting the spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data to obtain a conflict-free large data model.

[0125] In this embodiment, if the first conflicting entity and the second conflicting entity are the same entity, it is necessary to make the first conflicting entity and the second conflicting entity close to each other to increase the spatiotemporal overlap value of the first conflicting entity and the second conflicting entity; if the first conflicting entity and the second conflicting entity are not the same entity, it is necessary to make the first conflicting entity and the second conflicting entity far away from each other to avoid the first conflicting entity and the second conflicting entity overlapping with each other.

[0126] In a specific embodiment of the present application, step S400 corrects the spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data to obtain a conflict-free large data model, including steps S410 to S460.

[0127] Step S410: Using the position parameters of the first conflicting entity and the second conflicting entity as particle dimensions, randomly generating a particle group within an allowable adjustment range, each particle representing a potential correction position.

[0128] In this embodiment, the corrected position is generated as follows: First, determine the particle dimension: the spatial position parameters of the first collision entity and the second collision entity (for example, three-dimensional coordinates (x, y, z)) are used as the particle dimension, that is, the dimension of each particle is 3.

[0129] Then, set the adjustment range: according to the position distribution of the conflicting entity (i.e., the first conflicting entity or the second conflicting entity) in the multi-source data, determine the allowable adjustment range of each dimension (for example, [ , ]、[ , ]、[ , ],in, Indicates the minimum range value of the X-axis in three-dimensional space; The maximum range value of the X axis in three-dimensional space; Indicates the minimum range value of the Y axis in three-dimensional space; The maximum range value of the Y axis in three-dimensional space; Indicates the minimum range value of the Z axis in three-dimensional space; The maximum range of the Z axis in three-dimensional space. The above range is three times the standard deviation of the position of the entity in each data source (in other embodiments, it can also be two times or four times).

[0130] Finally, generate a particle swarm: randomly generate a number of particles (for example, 40 or 50 particles) within the above adjustment range, each particle representing a potential correction position.

[0131] Step S420: Calculate the three-dimensional Euclidean distance between the corrected position corresponding to each particle and the positions of the first conflicting entity and the second conflicting entity in each data source, and perform weighted averaging based on the credibility of each data source obtained in the evidence fusion stage to obtain the distance error.

[0132] In this embodiment, the method for obtaining the distance error is as follows: First, the three-dimensional Euclidean distance is calculated: for each particle’s corrected position, the three-dimensional Euclidean distance between it and the original position of the first and second collision entities in various data sources (e.g., drone point clouds, satellite images, ground sensor network data, etc.) is calculated.

[0133] Finally, weighted averaging is performed: the credibility of each data source obtained in the evidence fusion phase of step S300 is obtained (for example, the credibility of the drone point cloud is 0.8, the credibility of the satellite image is 0.7, and the credibility of the ground sensor is 0.6), and the distance error is calculated using the formula:

[0134] in, Indicates distance error; Indicates the number of data sources; represents the credibility obtained in the evidence fusion stage corresponding to the j-th data source; Represents the three-dimensional Euclidean distance between the original positions of the first conflicting entity and the second conflicting entity in the j-th data source.

[0135] Step S430: Obtain the nearest triangle edge or triangle face in the Delaunay triangulation corresponding to the corrected position of each particle, calculate the vertical distance from the corrected position to the triangle edge or triangle face, and if the corrected position belongs to a preset category entity and the vertical distance exceeds the allowable offset threshold of the category entity, then use the vertical distance as the topology violation.

[0136] In this embodiment, the allowed deviation threshold may be preset according to requirements, for example, the allowed deviation threshold may be 0.5 meters or 0.6 meters.

[0137] Step S440: constructing a fitness function based on the distance error and the topology violation.

[0138] In this embodiment, the fitness function is the sum of the distance error and the topology violation.

[0139] Step S450: Based on the fitness function, output the optimal position as a correction result.

[0140] In this embodiment, an iterative search using the particle swarm optimization algorithm is used to find the corrected position that minimizes the fitness function. It should be noted that PSO (Particle Swarm Optimization) is an optimization algorithm based on swarm intelligence. It simulates the collaborative behavior of a flock of birds when foraging and searches for the optimal solution in the solution space. This is a mature technology and will not be described in detail here.

[0141] Step S460: Based on the correction result, the spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data are corrected to obtain a conflict-free large data model.

[0142] In this embodiment, the original positions of the first conflicting entity and the second conflicting entity are updated to the optimal corrected positions outputted in step S450 , thereby eliminating the conflict between the two entities caused by the spatial position deviation.

[0143] Step S500: constructing a long short-term memory network based on the conflict-free big data model.

[0144] In this embodiment, the Long Short-Term Memory (LSTM) network is used to identify special events in cities. LSTM is a recurrent neural network suitable for processing time series data, effectively capturing long-term data dependencies. The fused and corrected data is divided into multiple sequence samples according to time windows and input into the LSTM model for training. During training, the model learns the characteristic patterns of data under normal urban conditions. When the input sample data deviates significantly from this normal pattern, it determines that a special urban event may have occurred. For example, in a rainstorm flooding warning scenario, when abnormal changes are observed in the fused features of monitored rainfall data, drainage system flow data, and ground water level data, the LSTM model identifies the possibility of a flooding event and labels the relevant data. The labeling information includes key information such as the time, location, and type of event.

[0145] The proposed method for integrating space-ground big data for urban measurement, proposed in this application, addresses the problem of low fusion accuracy caused by inconsistent spatiotemporal benchmarks and large differences in format accuracy across multiple data sources by integrating urban spatial point clouds, sensor networks, and remote sensing images. By eliminating spatiotemporal biases and entity conflicts, it improves data consistency and reliability, avoiding information gaps and semantic conflicts. Furthermore, the big data model constructed based on conflict-free data can accurately identify special events, providing support for scenarios such as urban rainstorm and flood warnings and public emergency response, facilitating refined urban management, precise decision-making, and efficient resource allocation, meeting the needs of complex urban measurement.

[0146] It should be understood that computer-readable storage media in this application include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to: phase change memory, static random access memory, dynamic random access memory, other types of random access memory, read-only memory, electrically erasable programmable read-only memory, flash memory or other memory technology, read-only compact disc read-only memory, digital versatile disc or other optical storage, magnetic cassettes, disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media such as modulated data signals and carrier waves.

[0147] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0148] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the methods, devices and equipment described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0149] In the several embodiments provided in the embodiments of the present application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is merely a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or modules can be electrical, mechanical or other forms.

[0150] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of the present embodiment according to actual needs.

[0151] In addition, the functional modules in each embodiment of the present application can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into a module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium.

[0152] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0153] The computer program product includes one or more computer instructions. When the computer program is loaded and executed on a computer, the process or functions described in accordance with the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium that can be stored on a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, hard disk, or magnetic tape), an optical medium (e.g., a digital versatile disk), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0154] Although the embodiments of the present application have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions, and alterations may be made to these embodiments without departing from the principles of the present application.

Claims

1. A space-ground integrated big data fusion method for urban measurement, characterized by: include: Acquire the city's integrated air-land big data; the integrated air-land big data includes at least the city's spatial point cloud data, the city's sensor network data, and the city's remote sensing image data; The integrated space-ground big data is uniformly calibrated to a city standard coordinate system to obtain data with low spatiotemporal deviation; the city standard coordinate system is at least used to eliminate the spatiotemporal deviation of different source data in the integrated space-ground big data; Based on the low spatiotemporal deviation data, a first conflicting entity and a second conflicting entity are obtained; the first conflicting entity is any entity in the low spatiotemporal deviation data; the second conflicting entity is any entity in the low spatiotemporal deviation data that spatially overlaps with the first conflicting entity; the first conflicting entity and the second conflicting entity have different data sources; Correcting the spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data to obtain a conflict-free large data model; Based on the conflict-free big data model, a long short-term memory network is constructed; The LSTM network is used to identify special events in the city.

2. The air-ground integrated big data fusion method for urban measurement according to claim 1 is characterized in that: The acquisition of the city's integrated space and land big data includes: Scanning an urban area using a drone equipped with a laser radar to obtain spatial point cloud data of the city; Acquiring remote sensing image data of the city based on low-altitude remote sensing satellites; Based on fixed sensors deployed on major roads and public areas in the city, sensor network data of the city is obtained.

3. The air-ground integrated big data fusion method for urban measurement according to claim 1 is characterized in that: The city standard coordinate system includes a primary benchmark network, a secondary structure network, a third-level coverage network, and a fourth-level dynamic network; The primary benchmark network is a network composed of multiple benchmark points in the urban space; the benchmark points are deeply buried municipal benchmark points or anchor points of cross-river bridges; at least three benchmark points are arranged in each administrative unit in the city to form a spatial triangle constraint; The secondary structure network is a spatial point cloud network composed of landmark buildings in the urban space; The three-level coverage network is the spatial position network of buildings in the first-level reference network and the second-level structure network extracted from the remote sensing image data based on an image feature extraction algorithm; The four-level dynamic network is a local coordinate system network constructed based on the synchronous positioning and map construction technology with each fixed sensor as an anchor point.

4. The air-ground integrated big data fusion method for urban measurement according to claim 3 is characterized in that: The method of uniformly calibrating the air-ground integrated big data to the city standard coordinate system to obtain low spatiotemporal deviation data includes: Mapping each spatial point cloud in the secondary structure network to the primary reference network to obtain a first fusion network; Based on the first fusion network, a plurality of elevation columns are obtained; the elevation columns are used to represent corresponding buildings in the city; Based on each elevation column, obtaining building projection information corresponding to each elevation column; the projection direction of the building projection is vertically downward; the building projection information at least includes the horizontal distance between each building projection; Based on the three-level coverage network, a projection offset of each building projection is obtained; the projection offset is used to at least represent the horizontal distance difference between the building projection and the corresponding building in the three-level coverage network; Based on the projection offset, adjusting the horizontal position of each elevation column in the first fusion network to obtain a second fusion network; Based on the four-level dynamic network, obtaining motion trajectory information of targets in each local coordinate system network; the targets include pedestrians and vehicles; The motion trajectory information is back-projected to the second fusion network to obtain the low spatiotemporal deviation data.

5. The air-ground integrated big data fusion method for urban measurement according to claim 4 is characterized in that: Mapping each spatial point cloud in the secondary structure network to the primary reference network to obtain a first fusion network includes: Based on each spatial point cloud, multiple outliers are obtained; Delete each outlier point from each spatial point cloud to obtain multiple corrected spatial point clouds; Based on the iterative closest point algorithm, each reference point and each corrected space point cloud are formed into a plurality of matching point pairs; in the matching point pairs, each reference point is a source point set, and each corrected space point cloud is a target point set; Based on each matching point pair, a first fusion network is obtained.

6. The air-ground integrated big data fusion method for urban measurement according to claim 4 is characterized in that: The step of obtaining a plurality of elevation columns based on the first fusion network includes: Obtaining the surface curvature of each cloud point in the first fusion network based on Delaunay triangulation technology; Based on the surface curvature of each cloud point, multiple surface fitting results are obtained; Based on each surface fitting result, multiple elevation columns are obtained.

7. The air-ground integrated big data fusion method for urban measurement according to claim 4 is characterized in that: The obtaining of the projection offset of each building projection based on the three-level coverage network includes: Based on the projections of each building, a target building projection is obtained; the target building projection is any building projection for which a projection offset has not been obtained among the projections of each building; Based on the projection of the target building, a fixed offset is obtained; the fixed offset is the offset of the projection of the target building in the three-level coverage network; Based on the fixed offset, a projection offset of the target building projection is obtained.

8. The air-ground integrated big data fusion method for urban measurement according to any one of claims 1 to 7, characterized in that: The acquiring the first conflicting entity and the second conflicting entity based on the low spatiotemporal deviation data includes: Based on the low spatiotemporal deviation data, a first entity and a second entity are obtained; the first entity is any entity in the low spatiotemporal deviation data; the second entity is any entity in the low spatiotemporal deviation data that spatially overlaps with the first entity; and the first entity and the second entity have different data sources; Slicing the first entity and the second entity according to a spatial grid and a time window, and obtaining a spatial overlap area and a spatial overlap duration; the spatial overlap area is an area where the first entity and the second entity spatially overlap; the spatial overlap duration is a duration where the first entity and the second entity spatially overlap; Based on the spatial overlap area and the spatial overlap duration, obtaining a spatiotemporal overlap value; the spatiotemporal overlap value is used to at least characterize the size of the area and duration of the spatial overlap between the first entity and the second entity; If the spatiotemporal overlap value satisfies a preset condition, the first entity and the second entity are respectively used as a first candidate entity and a second candidate entity; the preset condition includes that the spatiotemporal overlap value is less than a first preset value; A first conflicting entity and a second conflicting entity are acquired based on the first candidate entity and the second candidate entity.

9. The air-ground integrated big data fusion method for urban measurement according to claim 8 is characterized in that: The acquiring, based on the first candidate entity and the second candidate entity, a first conflicting entity and a second conflicting entity, includes: Based on different data sources in the low spatiotemporal deviation data, a plurality of confidence probabilities are obtained; the confidence probabilities are probabilities that the first candidate entity and the second candidate entity are two conflicting entities in the corresponding data sources; Based on the trust function theory, the conflict probability and the non-conflict probability are obtained from each confidence probability; the sum of the conflict probability and the non-conflict probability is equal to 100%; Based on the conflict probability and the non-conflict probability, obtaining a probability difference; the probability difference is equal to the conflict probability minus the non-conflict probability; If the probability difference is greater than a second preset value, the first candidate entity is determined to be a first conflicting entity, and the second candidate entity is determined to be a second conflicting entity.

10. The air-ground integrated big data fusion method for urban measurement according to claim 9 is characterized in that: Correcting the spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data to obtain a conflict-free large data model includes: Using the position parameters of the first conflicting entity and the second conflicting entity as particle dimensions, randomly generating a particle swarm within an allowed adjustment range, each particle representing a potential correction position; Calculate the three-dimensional Euclidean distance between the corrected position corresponding to each particle and the positions of the first conflicting entity and the second conflicting entity in each data source, and perform weighted averaging based on the credibility of each data source obtained in the evidence fusion stage to obtain the distance error; Obtain the nearest triangle edge or triangle face in the Delaunay triangulation network for the corrected position of each particle, calculate the vertical distance from the corrected position to the triangle edge or triangle face, and if the corrected position belongs to a preset category entity and the vertical distance exceeds the allowable offset threshold of the category entity, then use the vertical distance as the topology violation degree; Constructing a fitness function based on the distance error and the topology violation; the fitness function is the sum of the distance error and the topology violation; Based on the fitness function, outputting the optimal position as a correction result; The spatial positions of the first conflicting entity and the second conflicting entity in the low spatiotemporal deviation data are corrected based on the correction result to obtain a conflict-free large data model.

Images