InSAR building scatterer positioning method, device and equipment, storage medium and computer program product

Abstract

The invention relates to the technical field of building risk monitoring, in particular to an InSAR building scatterer positioning method, device and equipment, a storage medium and a computer program product. According to the method, urban point cloud data is segmented to obtain an urban building monomer instance point cloud set; based on the point cloud set, adopting a surface reconstruction algorithm to construct a building grid model; based on the grid model and the SAR satellite imaging parameters, adopting an SAR simulation technology to identify a radar illuminable vector frame of the building; converting the vector frame from a geographic coordinate system into an SAR image coordinate system, performing bilinear interpolation based on the geographic coordinates of the angular points of the vector frame, and determining the geographic coordinates of the scatterer in the vector frame; and if the scatterer is located in the overlapping region of the plurality of vector frames, performing overlay identification on the pixel points in the vector frames based on the geographic coordinates, determining the unique geographic coordinates, and determining the geographic positioning of the building scatterer in combination with the deformation information of the scatterer, thereby improving the positioning precision of the building scatterer.

Description

Technical Field

[0001] The present application relates to the technical field of building risk monitoring, and more particularly to an InSAR building scatterer positioning method, device, equipment, storage medium and computer program product. Background Art

[0002] With the acceleration of urbanization, the density and height of urban buildings have increased rapidly, which has put forward higher requirements for the safety and stability of building structures. Deformation monitoring of urban buildings has become an important means to ensure the safety of urban residents. Among the many deformation monitoring technologies, synthetic aperture radar interferometry (InSAR) technology has gradually become an important tool for urban building health monitoring and safety assessment due to its wide-area coverage, all-weather and all-day monitoring capabilities, and high-precision detection of small deformations. InSAR technology relies on satellite radar imaging and uses phase difference to monitor the deformation of the ground and buildings. It has been widely used in many fields such as high-rise building monitoring and deformation monitoring of buildings along underground projects. Compared with traditional geodetic and sensor monitoring methods, InSAR has the advantages of wide coverage, high monitoring accuracy and low cost. However, in practical applications, the monitoring effect of InSAR is affected by many factors such as terrain, building height, and building shape. Especially in high-density building areas, the positioning accuracy of InSAR is challenged. In InSAR monitoring, the precise positioning of building scatterers is an important factor in ensuring the accuracy of deformation monitoring, but the precise positioning of building scatterers faces the following two main problems: First, InSAR deformation measurement usually relies on a shorter spatial baseline to ensure deformation accuracy, while the height estimation of building scatterers requires a longer spatial baseline. This leads to the need to make a trade-off between deformation accuracy and height accuracy in urban building monitoring, which affects the precise positioning of building scatterers by InSAR. Secondly, due to the complex three-dimensional structure of the building, there are overlapping areas of multiple building scatterers in the InSAR image, making it difficult to accurately match each scatterer with the building features, and the positioning accuracy of InSAR is low. Therefore, how to improve the accuracy of InSAR building scatterer positioning has become a technical problem that needs to be solved urgently. Summary of the invention

[0003] The main purpose of this application is to provide an InSAR building scatterer positioning method, device, equipment, storage medium and computer program product, aiming to solve the technical problem of how to improve the accuracy of InSAR building scatterer positioning.

[0004] To achieve the above object, the present application provides an InSAR building scatterer positioning method, the method comprising the following steps:

[0005] Perform point cloud segmentation on the urban point cloud data to obtain a point cloud set of urban building monomer instances;

[0006] Based on the point cloud set of the urban building monomer instance, a point cloud surface reconstruction algorithm is used to construct a grid model of the building;

[0007] Based on the grid model and SAR satellite imaging parameters, a radar irradiable vector frame of the building is identified using SAR simulation technology;

[0008] The radar irradiable vector frame is converted from a geographic coordinate system to a SAR image coordinate system, and bilinear interpolation is performed based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame;

[0009] If there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, overlap mask recognition is performed on the pixel points located in the vector frames to determine unique geographic coordinates;

[0010] The geographic location of the building scatterer is determined based on the unique geographic coordinates and the scatterer deformation information.

[0011] In one embodiment, before the step of segmenting the urban point cloud data to obtain a point cloud set of urban building monomer instances, the method further includes:

[0012] Obtain initial city point cloud data;

[0013] Using a noise reduction algorithm based on statistical filtering to perform noise reduction processing on the initial urban point cloud data to obtain noise-reduced point cloud data;

[0014] Converting the coordinate system of the de-noised point cloud data into a preset geographic coordinate system, and performing density equalization processing on the de-noised point cloud data after the coordinate system conversion to obtain equalized point cloud data;

[0015] The height coordinates of the balanced point cloud data are normalized to obtain the city point cloud data.

[0016] In one embodiment, the step of segmenting the urban point cloud data to obtain a point cloud set of urban building monomer instances includes:

[0017] Based on a preset deep learning model, semantic segmentation is performed on the urban point cloud data;

[0018] Extract building point cloud from the semantic segmentation results to obtain building point cloud data;

[0019] Based on a preset clustering algorithm, the building point cloud data is instance segmented to obtain the urban building monomer instance point cloud set.

[0020] In one embodiment, the step of constructing a building mesh model based on the urban building monomer instance point cloud set using a point cloud surface reconstruction algorithm comprises:

[0021] Preprocessing the urban building monomer instance point cloud set to obtain a preprocessed point cloud set, wherein the preprocessing includes one or more of noise reduction, outlier removal, and downsampling;

[0022] Extracting geometric features of the preprocessed point cloud set, and based on the geometric features, generating an initial mesh model using the point cloud surface reconstruction algorithm;

[0023] Optimizing the initial mesh model to obtain an optimized model, wherein the optimization includes one or more of mesh simplification and smoothing;

[0024] The hole areas in the optimized model are filled to obtain a grid model of the building.

[0025] In one embodiment, the step of converting the radar irradiable vector frame from a geographic coordinate system to a SAR image coordinate system, and performing bilinear interpolation based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame includes:

[0026] Based on the imaging parameters and digital elevation model of the SAR image, terrain correction is performed on the radar irradiable vector frame, and a conversion relationship between geographic coordinates and SAR image coordinates is established;

[0027] Based on the conversion relationship, converting the radar irradiable vector frame from the geographic coordinate system to the SAR image coordinate system;

[0028] Taking the geographical coordinates of the corner points of the radar irradiable vector frame in the SAR image coordinate system as reference points;

[0029] Bilinear interpolation is performed on the vector frame based on the reference point to determine the geographic coordinates of the scatterer within the vector frame.

[0030] In one embodiment, if there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, the step of performing overlap recognition on the pixel points located in the vector frames to determine the unique geographic coordinates includes:

[0031] If there is a scatterer located in the overlapping area of ​​multiple vector frames, the geographic coordinates of the scatterer in the overlapping area are interpolated in different vector frames as first coordinates to obtain a first coordinate set;

[0032] The geographic coordinates determined by the InSAR algorithm are used as the original coordinates;

[0033] The coordinate in the first coordinate set with the smallest deviation from the original coordinate is used as the unique geographic coordinate.

[0034] In addition, to achieve the above-mentioned purpose, the present application also proposes an InSAR building scatterer positioning device, the InSAR building scatterer positioning device comprising:

[0035] Point cloud segmentation module, used to segment urban point cloud data and obtain point cloud sets of urban building monomer instances;

[0036] A model building module, used to build a grid model of the building based on the point cloud set of the urban building monomer instance and using a point cloud surface reconstruction algorithm;

[0037] An identification module, used to identify the radar irradiable vector frame of the building by using SAR simulation technology based on the grid model and SAR satellite imaging parameters;

[0038] A coordinate transformation module, used to transform the radar irradiable vector frame from a geographic coordinate system to a SAR image coordinate system, and to perform bilinear interpolation based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer within the vector frame;

[0039] An overlap recognition module, used for, if there is a scatterer located in the overlapping area of ​​multiple vector frames, to perform overlap recognition on the pixel points located in the vector frames based on the geographic coordinates of the scatterer to determine the unique geographic coordinates;

[0040] The target module is used to determine the geographic location of the building scatterer based on the unique geographic coordinates and the scatterer deformation information.

[0041] In addition, to achieve the above-mentioned purpose, the present application also proposes an InSAR building scatterer positioning device, which includes: a memory, a processor, and an InSAR building scatterer positioning program stored in the memory and executable on the processor, and the InSAR building scatterer positioning program is configured to implement the steps of the InSAR building scatterer positioning method described above.

[0042] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, on which an InSAR building scatterer positioning program is stored. When the InSAR building scatterer positioning program is executed by a processor, the steps of the InSAR building scatterer positioning method described above are implemented.

[0043] In addition, to achieve the above-mentioned purpose, the present application also proposes a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the InSAR building scatterer positioning method as described above are implemented.

[0044] This application obtains a point cloud set of urban building monomer instances by performing point cloud segmentation on urban point cloud data; based on the point cloud set of urban building monomer instances, a point cloud surface reconstruction algorithm is used to construct a grid model of the building; based on the grid model and SAR satellite imaging parameters, SAR simulation technology is used to identify the radar irradiable vector frame of the building; the radar irradiable vector frame is converted from the geographic coordinate system to the SAR image coordinate system, and bilinear interpolation is performed based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame; if there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, the pixel points located in the vector frame are overlapped and identified to determine the unique geographic coordinates; based on the unique geographic coordinates and scatterer deformation information, the geographic location of the building scatterer is determined. This application accurately maps the geometric features of the building structure into the SAR image through building point cloud segmentation and grid model reconstruction, combined with SAR simulation and coordinate conversion, to achieve high-precision positioning of the building scatterer. Bilinear interpolation is used to ensure the consistency of geographic coordinates and image coordinates, and multi-value interference in overlapping areas is eliminated through overlap identification to determine the unique geographic coordinates of the scatterer. Finally, the unique geographic coordinates and InSAR deformation information are used to provide accurate positioning and dynamic deformation data of building scatterers, thus improving the positioning accuracy of InSAR building scatterers. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of the first embodiment of the InSAR building scatterer positioning method of the present application;

[0046] Figure 2 This is a schematic diagram of a sub-process in the second embodiment of the InSAR building scatterer positioning method of the present application;

[0047] Figure 3 This is a schematic diagram of a sub-process in the third embodiment of the InSAR building scatterer positioning method of the present application;

[0048] Figure 4 This is a schematic diagram of the module structure of the InSAR building scatterer positioning device according to an embodiment of the present application;

[0049] Figure 5 Schematic diagram of the equipment structure of the hardware operating environment involved in the InSAR building scatterer positioning method in the embodiment of the present application.

[0050] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0051] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0052] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0053] It should be noted that with the acceleration of urbanization, the density and height of urban buildings have increased rapidly, which has put forward higher requirements for the safety and stability of building structures. Deformation monitoring of urban buildings has become an important means to ensure the safety of urban residents. Among the many deformation monitoring technologies, synthetic aperture radar interferometry (InSAR) technology has gradually become an important tool for urban building health monitoring and safety assessment due to its wide-area coverage, all-weather and all-day monitoring capabilities, and high-precision detection of small deformations. InSAR technology relies on satellite radar imaging and uses phase difference to monitor the deformation of the ground and buildings. It has been widely used in many fields such as high-rise building monitoring and deformation monitoring of buildings along underground projects. Compared with traditional geodetic and sensor monitoring methods, InSAR has the advantages of wide coverage, high monitoring accuracy and low cost. However, in practical applications, the monitoring effect of InSAR is affected by many factors such as terrain, building height, and building shape. Especially in high-density building areas, the positioning accuracy of InSAR is challenged. In InSAR monitoring, the precise positioning of building scatterers is an important factor in ensuring the accuracy of deformation monitoring, but the precise positioning of building scatterers faces the following two main problems: First, InSAR deformation measurement usually relies on a shorter spatial baseline to ensure deformation accuracy, while the height estimation of building scatterers requires a longer spatial baseline. This leads to the need to make a trade-off between deformation accuracy and height accuracy in urban building monitoring, which affects the precise positioning of building scatterers by InSAR. Secondly, due to the complex three-dimensional structure of the building, there are overlapping areas of multiple building scatterers in the InSAR image, making it difficult to accurately match each scatterer with the building features, and the positioning accuracy of InSAR is low. Therefore, how to improve the accuracy of InSAR building scatterer positioning has become a technical problem that needs to be solved urgently.

[0054] The main solution of this application is: by performing point cloud segmentation on urban point cloud data, a point cloud set of urban building monomer instance instances is obtained; based on the point cloud set of urban building monomer instance instances, a point cloud surface reconstruction algorithm is used to construct a grid model of the building; based on the grid model and SAR satellite imaging parameters, SAR simulation technology is used to identify the radar irradiable vector frame of the building; the radar irradiable vector frame is converted from the geographic coordinate system to the SAR image coordinate system, and bilinear interpolation is performed based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame; if there are scatterers located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterers, the pixel points located in the vector frame are overlapped and identified to determine the unique geographic coordinates; based on the unique geographic coordinates and the deformation information of the scatterers, the geographic positioning of the building scatterers is determined.

[0055] This application uses building point cloud segmentation and grid model reconstruction, combined with SAR simulation and coordinate conversion, to accurately map the geometric features of the building structure into the SAR image, thereby achieving high-precision positioning of the building scatterer. Bilinear interpolation is used to ensure the consistency of geographic coordinates and image coordinates, and mask recognition is used to eliminate multi-value interference in overlapping areas to determine the unique geographic coordinates of the scatterer. Ultimately, through unique geographic coordinates and InSAR deformation information, accurate positioning and dynamic deformation data of building scatterers are provided, improving the positioning accuracy of InSAR building scatterers.

[0056] It should be noted that the execution subject of the method of this embodiment can be a computing service device with data processing, network communication and program running functions, or it can be the above-mentioned InSAR building scatterer positioning device with the same or similar functions. This embodiment and the following embodiments will be described by taking the InSAR building scatterer positioning device as an example.

[0057] Based on this, the first embodiment of the InSAR building scatterer positioning method of the present application is proposed, please refer to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the InSAR building scatterer positioning method of the present application.

[0058] In this embodiment, the InSAR building scatterer positioning method includes the following steps:

[0059] S1: Segment the urban point cloud data to obtain a point cloud set of urban building monomer instances;

[0060] It should be noted that urban point cloud data is a set of three-dimensional data of urban areas obtained through technologies such as laser radar (LiDAR) or photogrammetry, which contains the spatial position (X, Y, Z coordinates) and reflection intensity information of urban objects (such as buildings, trees, ground, etc.). Point cloud data records the three-dimensional features of the surface of objects in the form of discrete points and is widely used in geographic information systems (GIS), building modeling and urban planning. Point cloud segmentation is the process of dividing point cloud data into several independent areas or categories according to specific standards or algorithms. Through segmentation, different objects (such as buildings, trees, ground, etc.) or different parts of the same object can be distinguished for subsequent analysis and processing. The urban building monomer instance point cloud set refers to a point cloud data set that is segmented and extracted from the urban point cloud data and contains only a single building or a single building monomer. These point cloud sets are used to independently represent the structural characteristics of each building, providing a data basis for subsequent building modeling, structural analysis, etc.

[0061] Specifically, in the process of point cloud segmentation of urban point cloud data, the original urban point cloud data needs to be preprocessed first. Preprocessing includes removing noise points, removing outliers and density equalization to ensure the clarity and consistency of point cloud data. After that, the point cloud data is semantically segmented using a deep learning algorithm or a rule-based point cloud segmentation method. Through semantic segmentation technology, different types of objects in the city (such as buildings, ground, vegetation, etc.) can be labeled and classified separately, so as to extract point cloud data belonging to buildings.

[0062] Furthermore, after obtaining the building point cloud data, the building point cloud is further instance segmented. This step usually uses a clustering algorithm (such as DBSCAN or K-Means) to divide the building point cloud data according to the building monomer instance to obtain an independent point cloud data set for each building monomer. Through instance segmentation, it can be ensured that each building appears in the form of independent point cloud data in subsequent processing, thereby facilitating the structural feature analysis of the monomer building and subsequent 3D reconstruction operations.

[0063] Point cloud segmentation effectively removes other objects not related to the building (such as vegetation, roads, vehicles, etc.), so that the point cloud data set reflects the spatial structure of the building. This segmentation result provides a high-quality and accurate building point cloud data set for the subsequent construction of the building grid model, reduces noise interference, and improves model accuracy. In addition, the building monomer instance point cloud set obtained by segmentation can accurately represent the structural characteristics of each building, making the positioning and analysis of each building in subsequent processing more accurate. This process lays the foundation for improving the positioning accuracy of InSAR scatterers and helps to achieve refined monitoring and structural health assessment of buildings.

[0064] S2: Based on the point cloud set of the urban building monomer instance, a point cloud surface reconstruction algorithm is used to construct a grid model of the building;

[0065] It should be noted that the point cloud surface reconstruction algorithm is an algorithm that generates a continuous surface model from discrete point cloud data. It connects the points in the point cloud data into meshes or patches to create a model that represents the geometric shape of the object surface. Common surface reconstruction algorithms include Delaunay triangulation, Poisson reconstruction, Laplace smoothing, etc. The mesh model is a three-dimensional model composed of polygonal (usually triangular or quadrilateral) meshes, which is used to represent the surface structure of the object. The mesh model generated by the point cloud surface reconstruction algorithm can accurately describe the geometric shape of the building, which is convenient for subsequent three-dimensional analysis and visualization.

[0066] Specifically, based on the point cloud set of urban building monomer instances obtained by segmentation, the point cloud data is first preprocessed to ensure data quality. For example, removing noise points, smoothing and downsampling can all optimize the density and uniformity of point cloud data. Next, the point cloud surface reconstruction algorithm is used to reconstruct the three-dimensional surface of the processed point cloud data. Commonly used surface reconstruction methods include Delaunay triangulation and Poisson reconstruction. Delaunay triangulation can quickly generate a preliminary triangular mesh model of a building, while Poisson reconstruction can generate a smooth and coherent building surface, which better reflects the detailed features of the building.

[0067] Furthermore, based on the initially generated mesh model, mesh optimization processing is performed, including eliminating redundant faces, smoothing mesh surfaces, and filling holes, to improve the integrity and detail accuracy of the model. Laplace smoothing and other techniques are usually used in the mesh optimization process to ensure the smoothness and geometric continuity of the mesh surface. The final generated building mesh model can accurately reflect the three-dimensional form of the building, providing a high-precision model foundation for subsequent SAR simulation and image projection.

[0068] The grid model of the building is constructed through the point cloud surface reconstruction algorithm, so that the originally discrete point cloud data is converted into a continuous and complete three-dimensional building model, which intuitively presents the geometric shape and spatial structure of the building. The grid model accurately describes the external contours and detailed features of the building, making the subsequent SAR simulation and image projection more consistent with the real form of the building. This step provides high-quality geometric model input for the precise positioning of InSAR scatterers, thereby further improving the positioning accuracy and facilitating the accurate monitoring and dynamic deformation analysis of buildings.

[0069] S3: Based on the grid model and SAR satellite imaging parameters, a radar irradiable vector frame of the building is identified using SAR simulation technology;

[0070] It should be noted that the SAR (Synthetic Aperture Radar) satellite imaging parameters include the imaging angle, viewing direction, angle of incidence, satellite orbit information, etc. of the SAR satellite. These parameters determine the geometric relationship between the SAR satellite's imaging of the surface and buildings, and affect the radar's irradiation area and image clarity. SAR simulation technology is based on SAR satellite imaging parameters. By calculating the projection of buildings in SAR images and the reflection of radar waves, it generates simulated images or projection frames that conform to real imaging characteristics. This technology can help identify which parts of the building surface can be effectively illuminated by the radar. The radar irradiable vector frame refers to the contour area of ​​the building that can be effectively illuminated in the radar image in the SAR image coordinate system after SAR imaging simulation. The vector frame represents the geometric range of the building in the SAR image, which is convenient for locating and analyzing building scatterers.

[0071] Specifically, based on the constructed building grid model and SAR satellite imaging parameters, the three-dimensional geometric structure information of the building and SAR imaging parameters (such as satellite incidence angle, azimuth, viewing angle, etc.) are first input into the SAR simulation system. The system uses the geometric characteristics of SAR imaging to simulate the projection of the building's grid model according to the satellite's observation angle and imaging angle, and generates the projection outline of the building in the SAR image. During the simulation process, the area on the building surface that can reflect radar waves is accurately identified by comparing the radar irradiation angle and the building surface normal vector, that is, the irradiable area of ​​the building in the SAR image.

[0072] Furthermore, after obtaining the projection area of ​​the building, the contour points of the area are further defined as the vector frame boundary, and the vertex coordinates and boundary information of the radar-irradiable vector frame are marked. This vector frame corresponds to the observable area of ​​the building in the SAR image, and contains all the building surface points that can be illuminated and imaged by the radar. Finally, the radar-irradiable vector frame provides an accurate spatial range and coordinate reference for subsequent coordinate conversion and scatterer positioning.

[0073] By using SAR simulation technology to identify the radar-irradiable vector frame of the building, the observable area of ​​the building can be accurately defined in the SAR image coordinate system, providing a clear geometric range for locating scatterers. This process effectively solves the mapping problem of the three-dimensional structure of the building in the radar image, making the scatterer positioning more consistent with the actual outline of the building and avoiding interference from redundant areas or non-building areas. At the same time, the vector frame obtained by simulation provides a reliable boundary reference for the coordinate transformation and precise positioning of the scatterer, further improving the accuracy and reliability of InSAR in building deformation monitoring.

[0074] S4: converting the radar irradiable vector frame from the geographic coordinate system to the SAR image coordinate system, and performing bilinear interpolation based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame;

[0075] It should be noted that the geographic coordinate system is a coordinate system based on the position on the surface of the earth, and longitude, latitude and elevation are usually used to represent the absolute position of an object. The geographic coordinate system is widely used in GIS (Geographic Information System) to describe the position on the earth. The SAR image coordinate system is a two-dimensional coordinate system in SAR images, and pixel row and column numbers are usually used to represent the position in the image. The SAR image coordinate system is based on satellite imaging characteristics and is used to analyze the position of scatterers in radar images. Bilinear interpolation is an interpolation method based on two-dimensional data points, which is used to calculate the value of any point in the area surrounded by four known points. This method calculates the value of the intermediate position by weighted average, and is usually used to improve the accuracy and smoothness of coordinate transformation.

[0076] Specifically, after the identification of the radar irradiable vector frame is completed, the vector frame is first converted from the geographic coordinate system to the SAR image coordinate system. This step is based on the imaging parameters of the SAR image and the digital elevation model (DEM). By establishing a conversion relationship from geographic coordinates to image coordinates, the vertex geographic coordinates of the vector frame are converted to the pixel coordinates of the SAR image. This conversion process ensures the accurate mapping of the building in the geographic coordinate system and the image coordinate system, so that the location in the SAR image can correspond to the real geographic location.

[0077] Furthermore, after the coordinate conversion is completed, the geographic coordinates of each corner point in the vector frame are bilinearly interpolated. Through the interpolation technology, the precise geographic coordinates of each scatterer in the vector frame area are calculated. Bilinear interpolation uses the geographic coordinates of the corner points of the vector frame to interpolate the geographic location of each pixel in the vector frame according to the distance weight, ensuring that the mapping relationship between the geographic coordinates of each pixel and the image coordinates corresponds accurately. Finally, this step generates the geographic coordinates of all scatterers in the vector frame, providing high-precision geographic location data for subsequent scatterer positioning.

[0078] By converting the radar irradiable vector frame from the geographic coordinate system to the SAR image coordinate system and using bilinear interpolation to calculate the geographic coordinates of the scatterer, accurate matching of geographic coordinates and image coordinates is achieved. This process effectively solves the spatial positioning problem of building scatterers in SAR images, allowing image data to be accurately mapped to the real geographic location. Bilinear interpolation ensures a smooth transition of scatterer coordinates within the vector frame, avoiding possible accuracy loss during coordinate conversion, and providing high-precision geographic coordinate support for subsequent InSAR monitoring and scatterer deformation analysis, thereby further improving positioning accuracy and consistency.

[0079] S5: If there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, overlap-mask identification is performed on the pixel points located in the vector frames to determine unique geographic coordinates;

[0080] It should be noted that the overlapping area refers to the part that intersects or overlaps between multiple vector frames. In the process of InSAR scatterer positioning, multiple candidate geographic coordinate values ​​of the same scatterer may appear in the overlapping area of ​​multiple vector frames, resulting in a multi-value phenomenon. Overlap recognition is a technology that screens multiple coordinates of the same object in the overlapping area. By comparing the accuracy or distance of each coordinate, an optimal value is determined to avoid multi-value interference and achieve unique positioning. A unique geographic coordinate refers to a single geographic coordinate determined for a scatterer after overlap recognition, ensuring that the scatterer has only one accurate geographic location in the overlapping area of ​​multiple vector frames.

[0081] Specifically, in the area where multiple vector frames overlap, the same scatterer may have multiple different candidate geographic coordinate values. To solve this problem, these candidate geographic coordinates are first identified by overlapping, and the scatterer geographic coordinate values ​​interpolated from different vector frames are collected into a candidate coordinate set, and the association between each candidate coordinate and its corresponding vector frame is recorded. Next, by calculating the distance between each candidate coordinate and the scatterer geographic coordinates originally calculated by the InSAR algorithm, the candidate coordinate with the smallest distance is selected. The coordinate with the smallest distance is considered to be closest to the true position and is thus used as the unique geographic coordinate of the scatterer.

[0082] Furthermore, after the unique geographic coordinates are selected, they are used as the final location data of the scatterer in the overlapping area, and other candidate coordinates are excluded. This unique positioning result eliminates the multi-value interference caused by the overlap of multiple vector frames, ensuring the unique corresponding geographic location of each scatterer in the SAR image. The unique geographic coordinates finally obtained provide stable and reliable scatterer location data for subsequent monitoring and analysis.

[0083] By performing mask recognition in the overlapping areas of multiple vector frames and determining unique geographic coordinates, the multi-value interference problem is solved, ensuring the unique positioning of each scatterer in the SAR image and geographic coordinate system. This process reduces the positioning error caused by inconsistent coordinates and significantly improves the accuracy and reliability of building scatterer positioning. Mask recognition ensures the accuracy of the scatterer's geographic coordinates, provides a solid data foundation for subsequent deformation analysis and building health assessment, and helps to more accurately identify the deformation trend of the building structure.

[0084] S6: Determine the geographic location of the building scatterer based on the unique geographic coordinates and the scatterer deformation information.

[0085] It should be noted that the scatterer deformation information is the displacement or deformation data of the scatterer over time obtained by InSAR technology, which is usually used to monitor the dynamic changes of building structures. The deformation information includes the deformation amount and deformation direction of the scatterer in the time series, which is an important basis for judging the health status of the building structure. The geographic positioning of the building scatterer is the position data of the scatterer in the geographic coordinate system determined by the unique geographic coordinates and deformation information. It provides the precise location of the scatterer and its deformation information, providing data support for subsequent structural monitoring and risk assessment.

[0086] Specifically, after confirming the unique geographic coordinates of each building scatterer, the geographic positioning of the building scatterer is completed by combining the scatterer deformation information calculated by InSAR technology. First, the unique geographic coordinates are used as the basic position of the scatterer in geographic space to ensure its accuracy in the geographic coordinate system. Next, the deformation information (such as displacement and direction) calculated by InSAR is superimposed on the unique geographic coordinates to generate data points containing spatial position and dynamic changes. This process can record the deformation of the scatterer over time, so that the geographic positioning not only includes the static position of the scatterer, but also its deformation characteristics.

[0087] Furthermore, geolocation data based on unique geographic coordinates and scatterer deformation information will provide key input for building monitoring systems. The location data of each building scatterer will include its geographic location, current status, and deformation information over time, facilitating dynamic monitoring of the overall health of the building. Ultimately, this location result will be used to generate thematic maps of the geographic distribution and deformation of building scatterers to facilitate the identification of building parts that may be at risk.

[0088] By combining the unique geographic coordinates with the deformation information of the scatterer, the geographic positioning of the building scatterer realizes the unification of spatial position and deformation data. This method not only provides the precise location of the scatterer, but also reveals its dynamic deformation characteristics, effectively improving the monitoring accuracy of the building structure status. Based on this geographic positioning information, building health assessment and deformation monitoring are more comprehensive and reliable, and can timely identify potential problems and trends in the building structure, providing important data information support for risk management and early warning of urban buildings.

[0089] This embodiment obtains a point cloud set of urban building monomer instances by performing point cloud segmentation on urban point cloud data; based on the point cloud set of urban building monomer instances, a point cloud surface reconstruction algorithm is used to construct a grid model of the building; based on the grid model and SAR satellite imaging parameters, a SAR simulation technology is used to identify the radar irradiable vector frame of the building; the radar irradiable vector frame is converted from the geographic coordinate system to the SAR image coordinate system, and bilinear interpolation is performed based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame; if there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, the pixel points located in the vector frame are overlapped and identified to determine the unique geographic coordinates; based on the unique geographic coordinates and the deformation information of the scatterer, the geographic location of the building scatterer is determined. This embodiment accurately maps the geometric features of the building structure into the SAR image through building point cloud segmentation and grid model reconstruction, combined with SAR simulation and coordinate conversion, to achieve high-precision positioning of the building scatterer. Bilinear interpolation is used to ensure the consistency of the geographic coordinates and the image coordinates, and the multi-value interference of the overlapping area is eliminated through overlap identification to determine the unique geographic coordinates of the scatterer. Finally, the unique geographic coordinates and InSAR deformation information are used to provide accurate positioning and dynamic deformation data of building scatterers, thus improving the positioning accuracy of InSAR building scatterers.

[0090] Based on the above first embodiment, a second embodiment of the InSAR building scatterer positioning method of the present application is proposed. Figure 2 , Figure 2 This is a schematic diagram of a sub-process in the second embodiment of the InSAR building scatterer positioning method of the present application.

[0091] like Figure 2 As shown, in this embodiment, before step S1, the following steps are also included:

[0092] S1a: Obtain initial city point cloud data;

[0093] S1b: using a noise reduction algorithm based on statistical filtering to perform noise reduction processing on the initial urban point cloud data to obtain noise-reduced point cloud data;

[0094] S1c: converting the coordinate system of the noise-reduced point cloud data into a preset geographic coordinate system, and performing density equalization processing on the noise-reduced point cloud data after the coordinate system conversion to obtain equalized point cloud data;

[0095] S1d: Normalizing the height coordinates of the balanced point cloud data to obtain the city point cloud data.

[0096] It should be noted that the initial urban point cloud data is the original three-dimensional data collected by LiDAR or photogrammetry technology, which contains the spatial location information of various objects in the city. The data is usually represented in the form of discrete points, recording the spatial coordinates and reflection intensity of different buildings, roads, vegetation, etc. Statistical filtering is a point cloud data processing method based on statistical characteristics, which is used to remove noise points. By analyzing the neighborhood distribution characteristics of each point in the point cloud, abnormal points or noise points are identified and filtered out, so as to obtain more accurate and clear point cloud data. Density equalization processing is a method of adjusting the density of point cloud data, aiming to make the point cloud data evenly distributed in space, avoid excessive concentration of high-density or low-density areas, so as to improve the accuracy of the model and the uniformity of the data. Height coordinate normalization is to convert the height value of the point cloud data to the same benchmark to eliminate the influence of terrain undulations or different height benchmarks, so that the point cloud data is consistent in height, which is convenient for subsequent analysis and processing.

[0097] Specifically, the initial urban point cloud data is obtained from the data source. Point cloud data often contains a lot of noise after collection, so a noise reduction algorithm based on statistical filtering is used to process the data. In the statistical filtering process, the neighborhood characteristics of each point, such as the distance to the adjacent points, are analyzed, and the threshold is set to remove outliers and outliers to obtain more accurate noise-reduced point cloud data. This process can effectively remove noise points that do not conform to the data distribution law and ensure the clarity and accuracy of the point cloud data.

[0098] Furthermore, the denoised point cloud data is converted into a preset geographic coordinate system to ensure alignment of the point cloud data with other geographic data. Subsequently, density equalization is performed to make the spatial distribution of the point cloud data more uniform and avoid the phenomenon of inconsistent density of data sets in different areas. After density equalization, the height coordinates of the point cloud data are normalized and the data height is adjusted to the same reference height to eliminate errors caused by terrain undulations. Finally, the urban point cloud data that has been processed by denoising, coordinate conversion, density equalization and height normalization is obtained, providing high-quality basic data for subsequent segmentation and model reconstruction.

[0099] The quality and consistency of the point cloud data were significantly improved by denoising, coordinate transformation, density equalization and height normalization of the initial urban point cloud data. Noise reduction removed outliers, making the data purer and ensuring the accuracy of the model. Density equalization and height normalization reduced the spatial inconsistency of the data and ensured the reliability of subsequent analysis results. These processing steps laid a solid data foundation for subsequent point cloud segmentation and building mesh model reconstruction, making model construction and positioning results more accurate, and providing high-quality input data for the positioning and deformation monitoring of InSAR building scatterers.

[0100] Based on the above first embodiment, in this embodiment, step S1 includes:

[0101] S11: performing semantic segmentation on the city point cloud data based on a preset deep learning model;

[0102] S12: extracting building point cloud from the semantic segmentation result to obtain building point cloud data;

[0103] S13: Based on a preset clustering algorithm, instance segmentation is performed on the building point cloud data to obtain an instance point cloud set of the urban building monomers.

[0104] It should be noted that the deep learning model is a multi-layer neural network structure used to automatically learn and extract features from data. By training a large amount of data, the deep learning model can recognize complex patterns and structures. It is often used for tasks such as classification and segmentation of images and point clouds. For example, models such as U-Net and PointNet are suitable for semantic segmentation of point clouds. Semantic segmentation is a technique for pixel-level or point-level classification of input data, aiming to assign each data point to a specific category. For point cloud data, semantic segmentation can separate different objects in the city (such as buildings, roads, vegetation, etc.) for subsequent analysis. Building point cloud extraction is to filter out point cloud data belonging to the building category from the semantic segmentation results, exclude points of other categories such as ground and vegetation, and provide input for subsequent building model reconstruction and analysis. Instance segmentation is a technique for separately dividing objects in the data to distinguish different independent objects. For building point cloud data, instance segmentation can divide different building monomers into their own independent point cloud data sets.

[0105] Specifically, a preset deep learning model is used to perform semantic segmentation on urban point cloud data. The point cloud data is input into a deep learning model that has been trained to recognize different types of objects (such as buildings, ground, trees, etc.). The semantic segmentation model classifies each point into a corresponding category based on its characteristics, thereby generating a segmentation result containing all categories. This result marks the urban point cloud data into multiple categories of areas, laying the foundation for further screening of building point cloud data.

[0106] Furthermore, after obtaining the semantic segmentation results, point cloud data belonging to the "building" category is extracted from it, which is called building point cloud data. Then, the preset clustering algorithm (such as DBSCAN or K-Means) is used to perform instance segmentation on the building point cloud data. Instance segmentation divides the building point cloud data into multiple independent building monomer point cloud sets, and clusters the point cloud data into multiple building monomer instances by identifying density distribution or neighborhood features. The final urban building monomer instance point cloud set can clearly reflect the structure of each independent building, providing point cloud data of the monomer building for subsequent modeling and analysis.

[0107] Through semantic segmentation using a deep learning model, buildings in urban point cloud data can be effectively separated from other objects, providing more accurate data for subsequent building modeling. Instance segmentation further divides the building point cloud into individual instances, allowing each building to be independently analyzed and located in subsequent processing. This process provides clear input for the three-dimensional reconstruction and high-precision positioning of buildings, reduces noise interference, ensures that the structural characteristics of buildings are preserved, and helps InSAR technology accurately locate building scatterers and monitor deformation, providing higher-quality basic data for health assessment and risk management of urban buildings.

[0108] This embodiment obtains a point cloud set of urban building monomer instances by performing point cloud segmentation on urban point cloud data; based on the point cloud set of urban building monomer instances, a point cloud surface reconstruction algorithm is used to construct a grid model of the building; based on the grid model and SAR satellite imaging parameters, a SAR simulation technology is used to identify the radar irradiable vector frame of the building; the radar irradiable vector frame is converted from the geographic coordinate system to the SAR image coordinate system, and bilinear interpolation is performed based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame; if there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, the pixel points located in the vector frame are overlapped and identified to determine the unique geographic coordinates; based on the unique geographic coordinates and the deformation information of the scatterer, the geographic location of the building scatterer is determined. This embodiment accurately maps the geometric features of the building structure into the SAR image through building point cloud segmentation and grid model reconstruction, combined with SAR simulation and coordinate conversion, to achieve high-precision positioning of the building scatterer. Bilinear interpolation is used to ensure the consistency of the geographic coordinates and the image coordinates, and the multi-value interference of the overlapping area is eliminated through overlap identification to determine the unique geographic coordinates of the scatterer. Finally, the unique geographic coordinates and InSAR deformation information are used to provide accurate positioning and dynamic deformation data of building scatterers, thus improving the positioning accuracy of InSAR building scatterers.

[0109] Based on the above second embodiment, a third embodiment of the InSAR building scatterer positioning method of the present application is proposed. Figure 3 , Figure 3 This is a schematic diagram of a sub-process in the third embodiment of the InSAR building scatterer positioning method of the present application.

[0110] In this embodiment, step S2 includes:

[0111] S21: preprocessing the urban building monomer instance point cloud set to obtain a preprocessed point cloud set, wherein the preprocessing includes one or more of noise reduction, outlier removal, and downsampling;

[0112] S22: extracting geometric features of the preprocessed point cloud set, and based on the geometric features, using the point cloud surface reconstruction algorithm to generate an initial mesh model;

[0113] S23: Optimizing the initial mesh model to obtain an optimized model, wherein the optimization includes one or more of mesh simplification and smoothing;

[0114] S24: Filling the hole areas in the optimized model to obtain a grid model of the building.

[0115] It should be noted that preprocessing is a step to clean and optimize the original data, aiming to improve the data quality and facilitate subsequent analysis. For point cloud data, preprocessing usually includes steps such as noise reduction, outlier removal, and downsampling to ensure the accuracy and uniformity of the point cloud. Noise reduction is the process of removing error points or noise points in point cloud data. Data points with large differences from neighboring points are removed using methods such as statistical filtering to make the data clearer and more reliable. Outlier removal is to identify and delete isolated points that deviate from the main group in point cloud data to improve the accuracy of the point cloud. Outliers are usually invalid points caused by acquisition errors. Removing them can improve the accuracy of the model. Downsampling is a method to reduce the amount of point cloud data. While maintaining the main features of the data, it reduces data redundancy and optimizes computational efficiency. Geometric features are the basic shape and structural information of point cloud data, including surface normal vectors, edges, corners, etc., which reflect the spatial structural characteristics of buildings in point clouds. Point cloud surface reconstruction algorithms are algorithms that generate three-dimensional surfaces from point cloud data. They usually use triangulation or Poisson reconstruction technology to generate mesh models, and represent the surface of objects with continuous patches. Mesh optimization refers to smoothing and simplifying the initial mesh model, removing redundant faces and filling holes to make the model smoother and more in line with the actual shape.

[0116] Specifically, the point cloud set of urban building monomer instances is preprocessed to improve data quality. The preprocessing process includes noise reduction, outlier removal, and downsampling. Noise points in the data are removed by noise reduction, and outliers are identified and removed to make the point cloud data purer. Downsampling then reduces the density of points to a moderate amount of data, providing an efficient and accurate data set for subsequent calculations. The final preprocessed point cloud set has higher accuracy and uniformity, and is suitable for further geometric reconstruction.

[0117] Furthermore, geometric features are extracted from the preprocessed point cloud set, and the initial mesh model is generated using the geometric features and the point cloud surface reconstruction algorithm. The initial mesh model is generated by triangulation or Poisson reconstruction technology to outline the three-dimensional shape of the building. Then, the initial mesh model is optimized, including mesh simplification and surface smoothing steps to reduce redundant facets and improve mesh smoothness. Finally, the hole areas in the optimized model are filled to ensure the integrity of the model, and the final mesh model of the building is generated to provide an accurate three-dimensional model foundation for subsequent SAR simulation and image projection.

[0118] Through preprocessing, geometric feature extraction, mesh optimization and hole filling, the generated building mesh model can more accurately reflect the structural characteristics and three-dimensional form of the building. Preprocessing removes noise and outliers to ensure the quality of point cloud data, while mesh optimization and hole filling improve the integrity and smoothness of the model. The final building mesh model provides a reliable geometric benchmark for subsequent InSAR scatterer positioning, making scatterer positioning more accurate and providing high-quality three-dimensional model support for building health monitoring and risk analysis.

[0119] Based on the above second embodiment, in this embodiment, step S4 includes:

[0120] S41: Based on the imaging parameters and digital elevation model of the SAR image, terrain correction is performed on the radar irradiable vector frame, and a conversion relationship between geographic coordinates and SAR image coordinates is established;

[0121] S42: Based on the conversion relationship, convert the radar irradiable vector frame from the geographic coordinate system to the SAR image coordinate system;

[0122] S43: taking the geographical coordinates of the corner points of the radar irradiable vector frame in the SAR image coordinate system as reference points;

[0123] S44: Perform bilinear interpolation on the vector frame based on the reference point to determine the geographic coordinates of the scatterer in the vector frame.

[0124] It should be noted that imaging parameters refer to the geometric and physical parameters required by SAR satellites during the imaging process, including the angle of incidence, azimuth, satellite orbit altitude, radar wavelength, etc. These parameters determine the imaging geometry and resolution of radar images and are an important basis for analyzing SAR image data. Digital elevation model is a digital representation of the surface height distribution, usually providing the altitude of each pixel in the form of a raster. DEM is used in SAR image analysis to compensate for the geometric deviation caused by terrain undulations and achieve accurate geographic correction. Terrain correction is the process of adjusting SAR images using DEM data to compensate for the influence of terrain undulations on the geometric position of the image and make the position of objects in the image more accurate. The coordinate transformation relationship is the mathematical mapping relationship between the geographic coordinate system and the SAR image coordinate system. Through this transformation, the real geographic location can be matched with the SAR image pixel coordinates to achieve mutual positioning between the two coordinate systems. Bilinear interpolation is an interpolation method that calculates the value of any intermediate point by weighting the values ​​of four known points. This method can be used to calculate the geographic coordinates of internal points based on the reference point in coordinate transformation to improve positioning accuracy.

[0125] Specifically, terrain correction is performed on the radar irradiable vector frame based on the imaging parameters and digital elevation model (DEM) of the SAR image. The geometric deformation caused by the undulating terrain in the image is adjusted through the DEM data, so that the position of the object in the image is more consistent with the real geographical location. Next, the coordinate conversion relationship between the geographic coordinate system and the SAR image coordinate system is established according to the imaging parameters. This conversion relationship maps the geographic location coordinates to the pixel coordinates in the SAR image through a mathematical mapping model to ensure the accurate correspondence between the image and the geographical location.

[0126] Furthermore, after the conversion relationship is established, the radar irradiable vector frame is converted from the geographic coordinate system to the SAR image coordinate system, and the geographic coordinates of the corner points of the vector frame are used as reference points. Then, based on these reference points, bilinear interpolation is used to interpolate the positions of the scatterers in the vector frame. The interpolation calculation determines the geographic coordinates of all scatterers in the vector frame through the known coordinates of the four corner points. This method ensures the smooth transition and precise positioning of the coordinates of the scatterers in the vector frame, and finally obtains the geographic location of each scatterer in the vector frame.

[0127] Through the terrain correction of imaging parameters and DEM and the conversion relationship between geographic coordinates and image coordinates, the precise matching of geographic location and SAR image is achieved. Bilinear interpolation interpolates the position of scatterers in the vector frame to ensure the smooth transition and positioning accuracy of geographic coordinates. This process greatly improves the accuracy of building scatterer positioning in SAR images, so that the geographic coordinates of each scatterer can be clearly mapped to the image, providing a reliable spatial reference frame for InSAR technology, thereby further improving the accuracy and effectiveness of building monitoring and deformation analysis.

[0128] Based on the above second embodiment, in this embodiment, step S5 includes:

[0129] S51: if there is a scatterer located in the overlapping area of ​​multiple vector frames, use the geographical coordinates of the scatterer in the overlapping area determined by interpolation in different vector frames as first coordinates to obtain a first coordinate set;

[0130] S52: Using the geographic coordinates determined by the InSAR algorithm as the original coordinates;

[0131] S53: Taking the coordinate in the first coordinate set with the smallest deviation from the original coordinate as the unique geographic coordinate.

[0132] It should be noted that the vector frame refers to the outline of the irradiable area of ​​the building calibrated in the SAR image coordinate system, so as to accurately locate the position and range of the building in the image. Multiple vector frames may partially overlap in the image, thus affecting the geographic positioning of the scatterer. The first coordinate set refers to the set of geographic coordinates of the scatterer calculated using the bilinear interpolation method in multiple overlapping vector frames. For scatterers located in the overlapping area of ​​multiple vector frames, different vector frames will generate multiple geographic coordinates to form a coordinate set. The original coordinates refer to the geographic coordinates of the scatterer directly calculated by the InSAR algorithm, which are used as the positioning reference to compare the accuracy and consistency of each candidate coordinate. The deviation value refers to the distance difference between the candidate geographic coordinates in the first coordinate set and the original coordinates, which is used to measure the closeness of each candidate coordinate to the original result of the InSAR algorithm. The unique geographic coordinate is the candidate coordinate closest to the original coordinate selected by comparing the deviation value to ensure the unique positioning of the scatterer in the overlapping area.

[0133] Specifically, when processing a scatterer in an overlapping area of ​​multiple vector frames, since different vector frames will generate the geographic coordinates of the scatterer in the overlapping area, it is first necessary to collect the coordinates obtained by interpolation of these different vector frames into a first coordinate set. Each coordinate in the first coordinate set represents a possible position of the scatterer calculated in different vector frames, providing a candidate value for finally determining its unique geographic coordinate.

[0134] Furthermore, the geographic coordinates directly calculated by the InSAR algorithm are used as the original coordinates and compared with the candidate coordinates in the first coordinate set. The deviation between each candidate coordinate and the original coordinate is calculated, and the candidate coordinate with the smallest deviation is selected as the unique geographic coordinate of the scatterer. This process ensures that the selected unique geographic coordinate is the coordinate closest to the original coordinate, minimizes multi-value interference in the overlapping area, and ensures the uniqueness and accuracy of the scatterer in the image and geographic coordinate system.

[0135] By comparing the deviations of the candidate coordinates obtained by interpolating different vector frames in the overlapping area and selecting the optimal value, the multi-value problem caused by the overlap of multiple vector frames is solved. The selected unique geographic coordinates determine the position of the scatterer in the overlapping area, eliminating the problem of inconsistent positioning caused by the overlap. This step effectively improves the accuracy and reliability of scatterer positioning, provides a stable geographic reference for InSAR monitoring and refined deformation analysis of buildings, and helps to more accurately monitor the health and risk assessment of building structures.

[0136] This embodiment obtains a point cloud set of urban building monomer instances by performing point cloud segmentation on urban point cloud data; based on the point cloud set of urban building monomer instances, a point cloud surface reconstruction algorithm is used to construct a grid model of the building; based on the grid model and SAR satellite imaging parameters, a SAR simulation technology is used to identify the radar irradiable vector frame of the building; the radar irradiable vector frame is converted from the geographic coordinate system to the SAR image coordinate system, and bilinear interpolation is performed based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame; if there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, the pixel points located in the vector frame are overlapped and identified to determine the unique geographic coordinates; based on the unique geographic coordinates and the deformation information of the scatterer, the geographic location of the building scatterer is determined. This embodiment accurately maps the geometric features of the building structure into the SAR image through building point cloud segmentation and grid model reconstruction, combined with SAR simulation and coordinate conversion, to achieve high-precision positioning of the building scatterer. Bilinear interpolation is used to ensure the consistency of the geographic coordinates and the image coordinates, and the multi-value interference of the overlapping area is eliminated through overlap identification to determine the unique geographic coordinates of the scatterer. Finally, the unique geographic coordinates and InSAR deformation information are used to provide accurate positioning and dynamic deformation data of building scatterers, thus improving the positioning accuracy of InSAR building scatterers.

[0137] Exemplarily, in order to help understand the technical concept or technical principle of the InSAR building scatterer positioning method of the above embodiment, the specific implementation of the InSAR building scatterer positioning method in one embodiment includes the following six steps.

[0138] Step 1: Obtain urban point cloud data, and use the point cloud denoising method based on statistical filtering to denoise the point cloud by analyzing local statistical features and applying nonlinear filtering technology. The urban point cloud set is represented as P = {p1,…,p i ,…,p N}, where P represents a point set consisting of N points, i = 1, ..., N. For each point p in the point cloud i Search for the k nearest neighbor points to the point, and the k neighbor points form point p i Neighborhood N(p i ). Neighborhood N(pi ) represents the point p i We use all point-to-point p i The average distance d i As features:

[0139]

[0140] The characteristics of all points in the point set P are D = {d1,…,d i ,…,d N}, calculate the mean μ for feature D D and standard deviation σ D :

[0141]

[0142] Calculate the outlier threshold τ based on α times the standard deviation:

[0143] τ=μ D +ασ D #(4)

[0144] The point set after being screened by the outlier threshold τ is expressed as P′={p′0,…,p′ N′}, where N ′ is the number of points after filtering.

[0145] Step 2: By building a Transformer point cloud semantic segmentation network based on the U-Net architecture, using the Transformer encoder and Transformer decoder as basic modules to learn the local and global geometric features of the point cloud from the multi-scale hierarchical super-point graph structure, so as to achieve the denoised point cloud P ′ Perform semantic segmentation of point clouds. The point cloud semantic segmentation model classifies urban point cloud data into 8 categories: ground, vegetation, cars, trucks, poles, wires, fences, and buildings.

[0146] Step 3: Perform building instance segmentation on the point cloud after semantic segmentation, as follows:

[0147]

[0148] Where P budiling Point cloud representing the building, F DBSCAN (·) represents the DBSCAN clustering algorithm, P b represents the set of all building instances, The set of points representing the nth building instance.

[0149] Step 4: Use the point cloud surface reconstruction algorithm to reconstruct the surface of the building point cloud and generate a grid model of the building. Using the building 3D outline, orientation and other information provided by the building grid model, combined with the SAR satellite's heading, line of sight, incident angle and other imaging parameters, use SAR simulation technology to identify the radar irradiable vector frame of the building.

[0150] Step 5: Based on the imaging parameters of the SAR satellite and the digital elevation model of the area, use the terrain correction projection conversion technology to establish a SAR image coordinate-geographic coordinate lookup table. According to the lookup table, the radar irradiable vector frame of the building is projected from the geographic coordinate system to the SAR image coordinate system.

[0151] Step 6: According to the building radar irradiable vector frame projected on the SAR image, the geographic coordinates of each vector frame corner point are used for bilinear interpolation to calculate the geographic coordinates of the scatterer in the vector frame. The scatterer located in the overlapping area of ​​the vector frames is an overlapping scatterer, and the scatterer needs to be identified by overlapping. For overlapping scatterers, obtain the geographic coordinates of the scatterer calculated by each vector frame containing the scatterer, compare the geographic coordinates calculated by multiple vector frames with the geographic coordinates calculated by the InSAR algorithm, and take the geographic coordinates of the vector frame closest to the geographic coordinates calculated by the InSAR algorithm as the final coordinates of the PS point, as shown in the following formula:

[0152]

[0153] PS xy is the permanent scatterer with row number x and column number y in the SAR image, (B, L, H) is the PS xy The final latitude, longitude and elevation, (B0, L0, H0) are the geographic coordinates of the scatterer calculated by the InSAR algorithm, (B i ,L i ,H i ) is the PS calculated for vector frame i xy latitude, longitude, and altitude.

[0154] The present application also provides an InSAR building scatterer positioning device, please refer to Figure 4 , Figure 4 This is a schematic diagram of the module structure of the InSAR building scatterer positioning device according to an embodiment of the present application. The InSAR building scatterer positioning device comprises:

[0155] The point cloud segmentation module 401 is used to segment the urban point cloud data to obtain a point cloud set of urban building monomer instances;

[0156] A model building module 402 is used to build a grid model of the building based on the point cloud set of the urban building monomer instance using a point cloud surface reconstruction algorithm;

[0157] An identification module 403 is used to identify the radar irradiable vector frame of the building by using SAR simulation technology based on the grid model and SAR satellite imaging parameters;

[0158] A coordinate transformation module 404 is used to transform the radar irradiable vector frame from a geographic coordinate system to a SAR image coordinate system, and to perform bilinear interpolation based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame;

[0159] The overlap recognition module 405 is used to, if there is a scatterer located in the overlapping area of ​​multiple vector frames, perform overlap recognition on the pixel points located in the vector frames based on the geographic coordinates of the scatterer to determine the unique geographic coordinates;

[0160] The target module 406 is used to determine the geographic location of the building scatterer based on the unique geographic coordinates and the scatterer deformation information.

[0161] The InSAR building scatterer positioning device provided in the embodiment of the present application adopts the InSAR building scatterer positioning method in the above embodiment, which can solve the technical problem of how to improve the accuracy of InSAR building scatterer positioning. Compared with the prior art, the beneficial effects of the InSAR building scatterer positioning device provided in the embodiment of the present application are the same as the beneficial effects of the InSAR building scatterer positioning method provided in the above embodiment, and other technical features in the InSAR building scatterer positioning device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0162] The present application provides an InSAR building scatterer positioning device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the InSAR building scatterer positioning method in the above-mentioned embodiment.

[0163] Reference below Figure 5 , Figure 5The device structure diagram of the hardware operating environment involved in the InSAR building scatterer positioning method in the embodiment of the present application shows a schematic diagram of the structure of the InSAR building scatterer positioning device suitable for implementing the embodiment of the present application. The InSAR building scatterer positioning device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The InSAR building scatterer positioning device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0164] like Figure 5 As shown, the InSAR building scatterer positioning device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. Various programs and data required for the operation of the InSAR building scatterer positioning device are also stored in the RAM 1004. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 may allow the InSAR building scatterer locating device to communicate with other devices wirelessly or by wire to exchange data. Although the InSAR building scatterer locating device with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have instead.

[0165] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0166] The InSAR building scatterer positioning device provided by the present application adopts the InSAR building scatterer positioning method in the above embodiment, which can solve the technical problem of how to improve the accuracy of InSAR building scatterer positioning. Compared with the prior art, the beneficial effects of the InSAR building scatterer positioning device provided by the present application are the same as the beneficial effects of the InSAR building scatterer positioning method provided by the above embodiment, and other technical features in the InSAR building scatterer positioning device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0167] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0168] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0169] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the InSAR building scatterer positioning method in the above-mentioned embodiment.

[0170] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0171] The computer-readable storage medium may be included in the InSAR building scatterer positioning device; or may exist independently without being assembled into the InSAR building scatterer positioning device.

[0172] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the InSAR building scatterer positioning device, the InSAR building scatterer positioning device: performs point cloud segmentation on urban point cloud data to obtain a point cloud set of urban building monomer instances; based on the point cloud set of urban building monomer instances, uses a point cloud surface reconstruction algorithm to construct a grid model of the building; based on the grid model and SAR satellite imaging parameters, uses SAR simulation technology to identify the radar irradiable vector frame of the building; converts the radar irradiable vector frame from a geographic coordinate system to a SAR image coordinate system, and performs bilinear interpolation based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame; if there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, overlaps and identifies the pixel points located in the vector frame to determine a unique geographic coordinate; based on the unique geographic coordinates and scatterer deformation information, determines the geographic positioning of the building scatterer. Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0173] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0174] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0175] The readable storage medium provided in the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned InSAR building scatterer positioning method, and can solve the technical problem of how to improve the accuracy of InSAR building scatterer positioning. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the present application are the same as the beneficial effects of the InSAR building scatterer positioning method provided in the above-mentioned embodiment, and will not be repeated here.

[0176] An embodiment of the present application provides a computer program product, including a computer program, which implements the steps of the above-mentioned InSAR building scatterer positioning method when executed by a processor.

[0177] The computer program product provided in this application can solve the technical problem of how to improve the accuracy of InSAR building scatterer positioning. Compared with the prior art, the beneficial effects of the computer program product provided in the embodiment of this application are the same as the beneficial effects of the InSAR building scatterer positioning method provided in the above embodiment, which will not be repeated here.

[0178] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent processing scope of the present application.

Claims

1. An InSAR building scatterer positioning method, characterized in that: The method comprises: Perform point cloud segmentation on the urban point cloud data to obtain a point cloud set of urban building monomer instances; Based on the point cloud set of the urban building monomer instance, a point cloud surface reconstruction algorithm is used to construct a grid model of the building; Based on the grid model and SAR satellite imaging parameters, a radar irradiable vector frame of the building is identified using SAR simulation technology; The radar irradiable vector frame is converted from a geographic coordinate system to a SAR image coordinate system, and bilinear interpolation is performed based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame; If there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, overlap mask recognition is performed on the pixel points located in the vector frames to determine unique geographic coordinates; The geographic location of the building scatterer is determined based on the unique geographic coordinates and the scatterer deformation information.

2. The method according to claim 1, characterized in that Before the step of segmenting the urban point cloud data to obtain a point cloud set of urban building monomer instances, the method further includes: Obtain initial city point cloud data; Using a noise reduction algorithm based on statistical filtering to perform noise reduction processing on the initial urban point cloud data to obtain noise-reduced point cloud data; Converting the coordinate system of the de-noised point cloud data into a preset geographic coordinate system, and performing density equalization processing on the de-noised point cloud data after the coordinate system conversion to obtain equalized point cloud data; The height coordinates of the balanced point cloud data are normalized to obtain the city point cloud data.

3. The method according to claim 1, characterized in that The step of segmenting the urban point cloud data to obtain a point cloud set of urban building monomer instances includes: Based on a preset deep learning model, semantic segmentation is performed on the urban point cloud data; Extract building point cloud from the semantic segmentation results to obtain building point cloud data; Based on a preset clustering algorithm, the building point cloud data is instance segmented to obtain the urban building monomer instance point cloud set.

4. The method according to claim 1, characterized in that The step of constructing a building mesh model based on the urban building monomer instance point cloud set by using a point cloud surface reconstruction algorithm comprises: Preprocessing the urban building monomer instance point cloud set to obtain a preprocessed point cloud set, wherein the preprocessing includes one or more of noise reduction, outlier removal, and downsampling; Extracting geometric features of the preprocessed point cloud set, and based on the geometric features, generating an initial mesh model using the point cloud surface reconstruction algorithm; Optimizing the initial mesh model to obtain an optimized model, wherein the optimization includes one or more of mesh simplification and smoothing; The hole areas in the optimized model are filled to obtain a grid model of the building.

5. The method according to claim 1, characterized in that The step of converting the radar irradiable vector frame from a geographic coordinate system to a SAR image coordinate system and performing bilinear interpolation based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer in the vector frame includes: Based on the imaging parameters and digital elevation model of the SAR image, terrain correction is performed on the radar irradiable vector frame, and a conversion relationship between geographic coordinates and SAR image coordinates is established; Based on the conversion relationship, converting the radar irradiable vector frame from the geographic coordinate system to the SAR image coordinate system; Taking the geographical coordinates of the corner points of the radar irradiable vector frame in the SAR image coordinate system as reference points; Bilinear interpolation is performed on the vector frame based on the reference point to determine the geographic coordinates of the scatterer within the vector frame.

6. The method according to any one of claims 1 to 5, characterized in that: If there is a scatterer located in the overlapping area of ​​multiple vector frames, based on the geographic coordinates of the scatterer, the step of performing overlapping recognition on the pixel points located in the vector frames to determine the unique geographic coordinates includes: If there is a scatterer located in the overlapping area of ​​multiple vector frames, the geographic coordinates of the scatterer in the overlapping area are interpolated in different vector frames as first coordinates to obtain a first coordinate set; The geographic coordinates determined by the InSAR algorithm are used as the original coordinates; The coordinate in the first coordinate set with the smallest deviation from the original coordinate is used as the unique geographic coordinate.

7. An InSAR building scatterer positioning device, characterized in that: The device comprises: Point cloud segmentation module, used to segment urban point cloud data and obtain point cloud sets of urban building monomer instances; A model building module, used to build a grid model of the building based on the point cloud set of the urban building monomer instance and using a point cloud surface reconstruction algorithm; An identification module, used to identify the radar irradiable vector frame of the building by using SAR simulation technology based on the grid model and SAR satellite imaging parameters; A coordinate transformation module, used to transform the radar irradiable vector frame from a geographic coordinate system to a SAR image coordinate system, and to perform bilinear interpolation based on the geographic coordinates of each vector frame corner point to determine the geographic coordinates of the scatterer within the vector frame; An overlap recognition module, used for, if there is a scatterer located in the overlapping area of ​​multiple vector frames, to perform overlap recognition on the pixel points located in the vector frames based on the geographic coordinates of the scatterer to determine the unique geographic coordinates; The target module is used to determine the geographic location of the building scatterer based on the unique geographic coordinates and the scatterer deformation information.

8. A computer device, characterized in that: The device comprises: a memory, a processor, and an InSAR building scatterer positioning program stored in the memory and executable on the processor, wherein the InSAR building scatterer positioning program is configured to implement the steps of the InSAR building scatterer positioning method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium stores an InSAR building scatterer positioning program, and when the InSAR building scatterer positioning program is executed by the processor, the steps of the InSAR building scatterer positioning method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the InSAR building scatterer positioning method according to any one of claims 1 to 6 are implemented.

Images