Building surveying method, device and equipment based on beidou communication, and storage medium

By using BeiDou satellite communication technology to perform continuous imaging and feature point recognition of buildings, the problems of building mapping being easily affected by occlusion and unstable communication in existing technologies have been solved, and high-precision three-dimensional mapping model generation has been achieved.

CN122283777APending Publication Date: 2026-06-26GUANGZHOU JIA HE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610388680.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-27
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

Existing building surveying technologies are susceptible to the effects of terrain, obstruction, climate, and unstable communication, making it difficult to achieve high-precision and consistent surveying. In particular, traditional communication methods cannot guarantee the stability and synchronization of data transmission when transmitting data across regions.

Method used

Using a BeiDou communication-based method, the target building area is continuously imaged by BeiDou satellites, an optimized image sequence is constructed, building height is inverted and feature point is identified, the outer contour is modeled by combining three-dimensional spatial coordinates, height deviation is corrected, and a three-dimensional mapping model is output.

Benefits of technology

It improves the accuracy and consistency of building surveying, reduces information loss caused by occlusion, shadows or viewing angle limitations, ensures the stability and synchronization of data transmission, and generates 3D models with high-precision geometric structure and geographic coordinate annotations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122283777A_ABST
    Figure CN122283777A_ABST
Patent Text Reader

Abstract

This invention relates to the field of satellite mapping, and more particularly to a method, apparatus, equipment, and storage medium for building mapping based on BeiDou communication. The method includes the following steps: continuously imaging a target building area using BeiDou satellites to construct an optimized image sequence; performing building height inversion on the optimized image sequence to obtain height distribution information; identifying building structural feature points based on the optimized image sequence to obtain the three-dimensional spatial coordinates of multiple feature points; modeling the outer contour based on the three-dimensional spatial coordinates, and correcting for height deviations based on the height distribution information to output a three-dimensional outer contour model; calculating the true latitude and longitude coordinates of the three-dimensional outer contour model, performing geographic coordinate annotation, and outputting a three-dimensional mapping model. This invention achieves accurate positioning and high-fidelity mapping modeling of buildings.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of satellite surveying and mapping, and in particular to a method, apparatus, equipment and storage medium for surveying and mapping buildings based on BeiDou communication. Background Technology

[0002] In the actual implementation of building surveying, the accuracy of the results is easily affected by various factors such as terrain undulations, building obstruction, weather conditions, and human error. Furthermore, with the proliferation of high-rise buildings and complex architectural structures, surveying methods based on a single perspective or data source cannot comprehensively and accurately reflect the true spatial form of buildings, easily leading to problems such as height deviation and contour distortion. In addition, in some areas, the acquisition and transmission of surveying data face limitations such as unstable communication or difficulties in data synchronization, further affecting the continuity and reliability of surveying work. Existing building surveying technologies mainly rely on ground surveying instruments, aerial photogrammetry, or satellite remote sensing. While these methods can acquire spatial information about buildings to some extent, they generally suffer from strong dependence on external communication conditions, complex data transmission links, and insufficient real-time performance. Especially when surveying data needs to be transmitted and processed centrally across regions, traditional communication methods cannot guarantee the stability and synchronization of data transmission, thus affecting the accuracy and consistency of the surveying results. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention proposes a building surveying method, apparatus, equipment, and storage medium based on BeiDou communication, thereby resolving at least one of the aforementioned technical problems.

[0004] To achieve the above objectives, this invention provides a building surveying method based on BeiDou communication, comprising the following steps: Step S1: Perform continuous imaging of the target building area based on BeiDou satellites and construct an optimized image sequence; Step S2: Perform building height inversion on the optimized image sequence to obtain height distribution information; Step S3: Based on the optimized image sequence, identify the building structure feature points to obtain the three-dimensional spatial coordinates of multiple feature points; Step S4: Model the outer contour based on the three-dimensional spatial coordinates, and correct the height deviation based on the height distribution information to output the three-dimensional outer contour model; Step S5: Calculate the true latitude and longitude coordinates of the three-dimensional outer contour model, perform geographic coordinate annotation, and output the three-dimensional mapping model.

[0005] This specification provides a building surveying device based on BeiDou communication, used to execute the building surveying method based on BeiDou communication as described above, including: The image acquisition unit is used to continuously image the target building area based on the BeiDou satellite and construct an optimized image sequence. The height inversion unit is used to invert building heights in optimized image sequences to obtain height distribution information. The feature point recognition unit is used to identify building structure feature points based on an optimized image sequence, and obtain the three-dimensional spatial coordinates of multiple feature points. The modeling unit is used to model the outer contour based on the three-dimensional spatial coordinates, correct the height deviation based on the height distribution information, and output a three-dimensional outer contour model. The coordinate annotation unit is used to calculate the true latitude and longitude coordinates of the three-dimensional outer contour model, perform geographic coordinate annotation, and output the three-dimensional mapping model.

[0006] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the building surveying method based on Beidou communication described above.

[0007] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the building surveying method based on BeiDou communication described above.

[0008] The specific benefits of this invention are as follows: Continuous imaging of the target building area using BeiDou satellites allows for the acquisition of multiple frames of image data at different times and viewing angles, effectively reducing information loss caused by occlusion, shadows, or viewing angle limitations in single images. Screening, registration, and optimization of the continuous imaging results to construct an optimized image sequence helps suppress imaging distortion caused by noise interference and attitude changes, improving the quality of input data for subsequent inversion and modeling processes. Joint analysis and inversion of multi-view images effectively obtains the height distribution information of buildings in different areas, avoiding height distortion problems caused by single-view estimation. Height inversion based on the optimized image sequence integrates multi-temporal and multi-angle information, reducing height calculation errors caused by lighting changes or local occlusion. Identifying key structural feature points such as building corners, roof edges, and facade intersections accurately depicts the geometric features of buildings, improving the structural integrity of the 3D model. Spatial calculation of feature points using multi-view images yields their precise coordinates in 3D space, reducing spatial errors caused by planar projection. By modeling the outer contour of feature points using 3D coordinates, the overall shape and structure of a building can be effectively restored, avoiding morphological distortions caused by simple geometric assumptions. Introducing height distribution information into the outer contour modeling process and correcting local height deviations helps eliminate height inconsistencies caused by inversion errors and feature point discreteness. The 3D outer contour model, after height deviation correction, more closely resembles the real building in spatial structure and scale, improving the usability of surveying results in engineering applications. Calculating and labeling the latitude and longitude of the 3D outer contour model based on BeiDou satellite positioning information ensures accurate positioning in global or regional coordinate systems. 3D surveying models with accurate geographic coordinate annotations can be directly used in various application scenarios such as urban planning, Geographic Information System (GIS) analysis, and emergency command, reducing secondary registration costs. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating the steps of a building surveying method based on BeiDou communication according to the present invention. Figure 2 This is a detailed flowchart illustrating the implementation steps of step S1. Figure 3 This is a flowchart illustrating the detailed implementation steps of step S2. Detailed Implementation

[0010] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0011] This application provides a building surveying method, apparatus, equipment, and storage medium based on BeiDou communication. The execution entities of the method, apparatus, equipment, and storage medium include, but are not limited to, mechanical equipment, data processing platforms, cloud server nodes, and network upload devices mounted on the system, which can be considered as general computing nodes in this application. The data processing platform includes, but is not limited to, at least one of an audio-visual management system, an information management system, and a cloud-based data management system.

[0012] Please see Figures 1 to 3 This invention provides a building surveying method based on BeiDou communication, comprising the following steps: Step S1: Perform continuous imaging of the target building area based on BeiDou satellites and construct an optimized image sequence; Step S2: Perform building height inversion on the optimized image sequence to obtain height distribution information; Step S3: Based on the optimized image sequence, identify the building structure feature points to obtain the three-dimensional spatial coordinates of multiple feature points; Step S4: Model the outer contour based on the three-dimensional spatial coordinates, and correct the height deviation based on the height distribution information to output the three-dimensional outer contour model; Step S5: Calculate the true latitude and longitude coordinates of the three-dimensional outer contour model, perform geographic coordinate annotation, and output the three-dimensional mapping model.

[0013] In the embodiments of the present invention, see Figure 1 The diagram below illustrates the steps of a building surveying method based on BeiDou communication according to the present invention. In this example, the steps of the building surveying method based on BeiDou communication include: Step S1: Perform continuous imaging of the target building area based on BeiDou satellites and construct an optimized image sequence; In this embodiment, the image acquisition sequence is planned according to the satellite orbit parameters and shooting window of the geographical area where the building is located, ensuring that the same building is covered in different orbits or at different time periods, thus ensuring spatial and temporal continuity of the images. Imaging parameters include shooting angle, elevation angle, azimuth angle, exposure time, and satellite attitude data, which are calibrated and parsed by reading satellite image metadata. The acquired raw images are preprocessed, including radiometric correction, noise suppression, and geometric correction, to eliminate the effects of sensor noise, illumination differences, and satellite attitude disturbances. To construct an optimized image sequence, the preprocessed images are sorted according to shooting time and spatial coverage continuity, and blurry or severely occluded image frames are removed, forming a high-quality, continuous, and geometrically consistent optimized image sequence.

[0014] Step S2: Perform building height inversion on the optimized image sequence to obtain height distribution information; In this embodiment, the shadow area and projection direction of buildings in each frame of the image are determined by illumination tracing. Then, by combining the elevation angle, azimuth angle, and shooting time of the satellite image, the correspondence between the shadow length and height of each part of the building is calculated using a geometric optics model. Multi-temporal data are fused, and the error in single-frame height measurement is reduced by least-squares fitting or weighted averaging methods to obtain continuous height distribution information of buildings in different regions. The height resolution is usually set according to the image resolution and shadow detection accuracy; for example, when the image resolution is 0.5 m / pixel, the height accuracy can reach 1~2 meters.

[0015] Step S3: Based on the optimized image sequence, identify the building structure feature points to obtain the three-dimensional spatial coordinates of multiple feature points; In this embodiment, edge detection and contour extraction are performed on each frame of the image to identify key points such as roof corners, facade edge intersections, and contour polyline nodes. Combining the imaging parameters of each frame, the two-dimensional pixel coordinates are projected onto a ground reference coordinate system to obtain two-dimensional spatial coordinates. Through multi-temporal feature point matching and spatial geometric constraints (including the length of the line connecting adjacent corners, angle consistency, and surface normal continuity), abnormal matches are eliminated, and stable multi-frame matching information is obtained. Finally, multi-view geometric solution or the ray intersection method is used to calculate the coordinates of each feature point in three-dimensional space, and the projection error is optimized using BundleAdjustment.

[0016] Step S4: Model the outer contour based on the three-dimensional spatial coordinates, and correct the height deviation based on the height distribution information to output the three-dimensional outer contour model; In this embodiment, feature points are connected in a topological order to form polylines and closed polygons, constructing the building's roof and facade boundaries, thus forming a preliminary 3D outer contour model. Combining the height distribution information obtained in step S2, deviation correction is performed on the model node heights. Deviation identification is achieved by calculating the difference between the model height and the height distribution data. Nodes with differences exceeding a threshold (e.g., 0.5~1 meter) are marked as local deviation areas, and the deviation nodes are smoothed using Laplacian smoothing or surface fitting to ensure the overall model height is continuous and consistent with the actual building height. The final output 3D outer contour model reflects both the building's geometry and closely matches the height data.

[0017] Step S5: Calculate the true latitude and longitude coordinates of the three-dimensional outer contour model, perform geographic coordinate annotation, and output the three-dimensional mapping model.

[0018] In this embodiment, the latitude and longitude of the building's center point and key nodes are obtained using BeiDou satellite positioning information, and the vertices of the local model coordinate system are transformed to WGS84 or a projected coordinate system. During the transformation, model rotation, translation, and scaling are considered to ensure the geometry matches the geographical location while preserving height information. Each vertex records its latitude, longitude, and elevation, achieving a correspondence between 3D space and geographic coordinates. The final output 3D mapping model not only possesses complete geometric shape and height information but also includes real geographic location annotations, which can be directly used for urban digital modeling, precise surveying, visualization rendering, and navigation analysis, achieving high-precision 3D mapping of buildings based on BeiDou satellite positioning imagery.

[0019] In this embodiment, see Figure 2 The diagram below illustrates the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: The imaging parameters of the multi-temporal satellite images are analyzed to extract the imaging parameter set, which includes the imaging angle, azimuth angle, elevation angle, and shooting time. Based on the imaging parameter set, multi-temporal satellite images are time-series labeled, and the first image sequence is output. Frame-by-frame sharpness calculation is performed on the first image sequence to obtain the sharpness index of each frame. The sharpness index is compared based on a preset sharpness threshold, and image frames below the sharpness threshold are removed to obtain a second image sequence; The second image sequence is subjected to brightness equalization and contrast normalization to construct an optimized image sequence.

[0020] In this embodiment, the geometric information of the imaging is obtained using satellite image metadata or satellite attitude control records. By reading the row and column numbers of the images, the satellite orbital position, and attitude data, the observation direction and spatial orientation corresponding to each image are calculated. The imaging angle describes the angle between the central optical axis of the satellite optical sensor and the vertical direction of the ground; the azimuth angle represents the deviation angle of the satellite's heading from the ground reference direction; the pitch angle describes the elevation and deflection angle of the satellite around its horizontal axis; and the capture time precisely records the moment information at the instant the image is acquired, typically with millisecond-level accuracy. By analyzing these parameters for each frame of multi-temporal images, a time-series-based imaging parameter set can be formed, providing a geometric reference and temporal information basis for subsequent time-series marking, image registration, and 3D modeling. Using the capture time as the main sequence index, each image is arranged in chronological order, and the imaging angle, azimuth angle, and pitch angle are combined to assist in sorting the images, ensuring that adjacent images in the sequence are continuous and reasonable in terms of spatial observation angle and time. Through this method, a first image sequence can be obtained, in which each frame of the image has a timestamp and spatial attitude information, facilitating subsequent sharpness assessment and image registration operations. To avoid confusion in the sequence caused by satellite orbital disturbances or slight attitude deviations, imaging parameters are usually normalized during the sequencing process. For example, the azimuth and elevation angles are mapped to the range of 0° to 360°, and the continuity of attitude between adjacent frames is verified by the minimum Euclidean distance method.

[0021] After converting the image to grayscale, the local gradient or second derivative of each pixel is calculated, and the gradient magnitudes of the entire image or segmented regions are statistically summed to obtain a sharpness quantification. This sharpness index reflects the detail and texture sharpness of the image; a higher value indicates better edge sharpness, making it suitable for building outline recognition and 3D reconstruction. To ensure comparability of sharpness between different images, the image size is typically scaled uniformly during the calculation process, for example, satellite imagery is standardized to 1024×1024 or 2048×2048 pixels.

[0022] After obtaining the sharpness index of each image frame, it is compared with a preset sharpness threshold. Image frames below the threshold are removed to ensure the reliability of the image sequence used for 3D mapping. The threshold can be set according to the satellite image resolution, target building size, and reconstruction accuracy requirements. For example, for satellite imagery with a resolution of 0.5 m / pixel, the sharpness threshold can be set to a normalized gradient variance of 0.25~0.3. The removal process is performed sequentially according to the time series, retaining images with acceptable sharpness and consecutive adjacent frames. The image sequence after removal is the second image sequence. This step avoids image interference caused by blur, cloud occlusion, or motion blur, improving the accuracy of 3D reconstruction and texture matching. The removal strategy can be combined with the average sharpness of adjacent frames. If multiple adjacent frames are all below the threshold, they can be removed in batches to avoid local noise affecting the final sequence quality. The image grayscale histogram is equalized or adaptively histogram equalized to improve the visibility of local details. Image contrast is normalized to map the grayscale dynamic range of each image to a uniform range, such as 0~255 or 0~1, ensuring that images within a sequence can be directly compared and superimposed in terms of brightness and contrast. Brightness equalization and contrast normalization eliminate the effects of uneven lighting, differences in image brightness, or sensor exposure deviations, ensuring consistency in texture and geometric features across multiple temporal images.

[0023] In this embodiment, see Figure 3 The diagram below illustrates the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Dynamic illumination tracking was performed on the optimized image sequence to obtain the facade illumination distribution information at different time periods; Based on the optimized image sequence, shadow regions are identified, and building shadow outlines are marked; Based on the facade lighting distribution information and building shadow outline, the shadow length, projection direction and offset over time are calculated to obtain the light time migration information; Building height is inverted based on the temporal migration information of light rays using the imaging parameter set, and height distribution information is obtained.

[0024] In this embodiment, each frame of satellite imagery is converted into a grayscale or luminance map. Pixel-level illumination intensity is quantified through illumination gradient calculation or an adaptive thresholding method based on local luminance ratios. Combining the shooting time, satellite viewpoint, azimuth, and elevation angle information from the imaging parameter set, the angle between the incident light direction and the normal to the building surface can be calculated, thereby deriving the illumination intensity distribution received by the building facade. To ensure illumination tracking accuracy, the image is typically divided into small grid blocks, such as 32×32 pixel units, and the average luminance and local gradient direction are statistically analyzed within the grid to reduce the impact of local noise on illumination calculations. By comparing illumination changes in multi-temporal image sequences, the illumination gradient of different parts of the building surface over time can be analyzed, providing basic data for shadow identification and subsequent height inversion. Using image luminance and color information as input, shadow extraction is performed using threshold-based segmentation, edge detection, or machine vision methods. An adaptive threshold is set based on the difference in illumination intensity between the light intensity and the brightness of surrounding non-building areas, identifying areas in the image significantly below the threshold as potential shadows. The shadow region boundaries are extracted using edge detection operators (such as the Canny operator), and small spots or noisy areas are eliminated through morphological processing to make the shadow contour smooth and continuous. To improve temporal consistency, the shadow recognition results are compared with multi-temporal illumination distribution information to remove abnormal regions that are inconsistent with the illumination.

[0025] Each frame of shadow contour is projected onto a ground reference coordinate system, and the ground projection scale and orientation are calculated based on imaging parameters and satellite position. By analyzing the continuous positional changes of the same building's shadow on the ground coordinates in multi-temporal images, the change in shadow length over time and the offset angle of the projection direction are calculated. Using time-series data, the continuous changes in shadow length and direction are processed by difference or curve fitting to obtain the migration speed and trajectory of light along the building's projection plane at different time periods. This time-series light migration information can accurately reflect the shadow change pattern caused by the change in the direction of solar or satellite illumination over time, providing dynamic geometric constraints for building height calculation. Imaging parameters such as the incident angle of light, satellite viewpoint, azimuth angle, elevation angle, and shooting time are combined with shadow length and projection direction, and height is calculated using a geometric optics model. For each shadow contour projection, the building height can be calculated using simple trigonometric relationships, i.e., the vertical height is estimated based on the satellite light incident angle and the ground shadow length. To increase the accuracy of height estimation, a least-squares fitting method for multi-frame shadow data is usually used to comprehensively optimize the height results obtained from different time periods, reducing the impact of single-frame measurement errors. It can calculate the height of different facades of a building in segments to obtain information on the spatial height distribution of the building.

[0026] In this embodiment, step S3 includes the following steps: Building structural feature points are identified and geometric feature points are marked based on optimized image sequences; the geometric feature points include roof corner points, facade edge intersections, and contour polyline nodes. Calculate the two-dimensional spatial coordinates of the geometric feature points at different time periods; Spatial geometric constraint matching is performed on the two-dimensional spatial coordinates to obtain feature point matching information; Multi-visual geometric calculations are performed based on feature point matching information to obtain the three-dimensional spatial coordinates of multiple feature points.

[0027] In this embodiment, the image undergoes edge detection and contour extraction processing. For example, the Canny or Sobel operator is used to extract the building contour edges, and morphological processing and region filtering are performed on noise and false edges to obtain continuous and smooth building edges. Polyline nodes are detected on the edge contours, and roof corners and facade edge intersections are identified through angle changes and local curvature calculations. Roof corners correspond to key turning points in the building's roof geometry, facade edge intersections mark the geometric features at the junction of the facade and roof, or different facades, while contour polyline nodes are used to describe the continuous polyline contours of the facade and roof edges. For multi-temporal images, feature point extraction for each frame is performed independently, and the image frame number, pixel coordinates, and corresponding image quality indicators are recorded. To ensure the reliability of the extracted feature points, a minimum edge length threshold is typically set, such as 20-30 pixels, to eliminate small-area stray noise points. At the same time, local gradient magnitudes are thresholded to ensure that edges and corners remain stable under illumination changes. By combining the imaging parameter set of each frame, including satellite imaging viewpoint, azimuth, elevation angle, and capture time, the pixel coordinates are transformed to a ground reference coordinate system (geographic coordinates or projected coordinate system). The pixel-to-spatial coordinate transformation can be achieved through ray projection calculations of the azimuth and elevation angles within the image; that is, mapping the pixel along the optical axis onto a hypothetical ground or building reference plane to obtain two-dimensional planar coordinates. For multi-temporal images, this calculation process is repeated to obtain the two-dimensional coordinates of each feature point at different time periods, and the time label and corresponding frame number are recorded. To reduce projection errors, interpolation or multi-point fitting methods can be used in the ray projection calculation, such as cubic interpolation using neighboring satellite orbital position data, to achieve a mapping accuracy of 0.5 to 1 pixel in the two-dimensional coordinate calculation.

[0028] Feature point descriptors are established, such as calculating feature vectors based on local grayscale textures or local edge structures, to represent each two-dimensional coordinate point. Candidate matching is performed between feature point sets from different time periods using metrics such as Euclidean distance or Hamming distance, and filtering is conducted in conjunction with spatial geometric constraints. Geometric constraints include constraints on the relative distance between feature points, limitations on angle changes, and linear or polyline consistency constraints on adjacent point sequences. In roof corner point matching, the projection spacing and angle changes of consecutive corner points should be consistent with the actual structure of the building roof, allowing a small range of error of 0.5~1 meter. Through geometric constraint matching, erroneous and duplicate points can be eliminated, obtaining feature point matching information across multiple time phases. Each matching record contains the starting frame, target frame, two-dimensional coordinates, and matching position information, providing a reliable correspondence for 3D solution. After completing feature point matching, the matching point information is used for multi-view geometric solution to calculate the 3D spatial coordinates of each feature point. Specifically, each set of matched two-dimensional coordinate points and the satellite position and attitude information in its corresponding imaging parameter set are input into the multi-view geometric model, and the 3D coordinates of the feature points are calculated using the spatial ray intersection method or the least squares 3D reconstruction method. Multi-view geometric solving considers the projection error of each feature point across multiple frames of images, minimizing the reprojection error between each projection point and the measured 2D coordinates through iterative optimization (such as Bundle Adjustment). To improve reconstruction accuracy, at least 3-5 frames of images from different perspectives can be selected for solving each feature point, achieving a 3D coordinate accuracy at a ground scale corresponding to 1-2 pixels. The final output is a set of coordinates of key geometric feature points of the building in 3D space, including the X, Y, and Z coordinates of roof corners, facade edge intersections, and contour polyline nodes, providing reliable 3D control points for accurate 3D mapping of buildings and achieving a complete mapping from 2D images to 3D structures.

[0029] In this embodiment, step S4 includes the following steps: Based on the three-dimensional spatial coordinates, spatial topological relationship analysis is performed to determine the boundary orientation and structural connection relationship of the building's outer contour; Based on the boundary orientation and structural connection relationship, the outer contour is modeled to generate the building's outer contour model; Based on height distribution information, height distribution deviations are identified in the outer contour model, and local deviation areas are marked. The local deviation areas are smoothed to output a three-dimensional outer contour model.

[0030] In this embodiment, the adjacency relationships of roof corners, facade edge intersections, and contour polygonal nodes are constructed. By calculating the Euclidean distance and spatial connection direction between each feature point, it is determined which points are spatially connected to form a continuous contour. Combined with geometric constraints, such as the roof edge being typically straight and its length not less than 2-5 meters, and the normal consistency constraint at the intersection of the facade edge and the roof, outliers and isolated points can be eliminated. The feature point set is topologically sorted to form a preliminary boundary orientation sequence, clarifying the closed path of the outer contour and the connection order between each edge. By calculating the spatial angle and height difference between adjacent nodes, the connection relationship of the building facade is identified, and the structural logic between the roof and the facade, and between different facades, is determined. The three-dimensional spatial feature points are connected according to the topological sequence to construct polylines and closed polygons, forming the preliminary outer contour geometric structure. For the connection between the roof and the facade, spatial straight line segments or polygonal segments are used for modeling to ensure that the boundary is continuous and without breaks. Through surface fitting or planar interpolation methods, the contour nodes are connected to form a polyhedral model of the building's outer surface, establishing a geometric topological relationship between the roof plane and the facade plane. To improve model accuracy, a minimum boundary spacing threshold (e.g., 1-2 meters) can be set during the modeling process to avoid model distortion caused by overlapping or excessively dense nodes. Align the model with the ground coordinates at the bottom of the building to ensure that the outer contour model is consistent with the actual geographical location in terms of spatial position and scale.

[0031] The Z-coordinates of each node or surface vertex of the outer contour model are compared point-by-point with the height distribution information to calculate the height deviation. Regions with deviations exceeding a preset threshold (e.g., 0.5-1 meter) are marked as local deviation areas. These deviation areas may be caused by inconsistent model heights due to lighting calculation errors, shadow discrimination errors, or image registration deviations. By marking deviations, regions with height anomalies or local discontinuities in the outer contour model can be quickly located, providing a basis for subsequent shape correction. To reduce the interference of single-point noise on deviation judgment, neighborhood smoothing filtering or local window mean statistical methods can be used for height deviations to ensure the spatial continuity and structural rationality of the deviation-marked areas. The vertices or polygonal surfaces of the local deviation areas are smoothed, for example, using Laplacian smoothing or Gaussian filtering methods, to ensure a continuous transition of height distribution and boundary contours, eliminating sharp protrusions or depressions. During the processing, control parameters can be set for the smoothing intensity, such as a neighborhood size of 3×3 to 5×5 vertices and an iteration count of 5 to 10, to balance local correction and global geometric fidelity. Smoothing not only optimizes local height distribution but also improves the surface smoothness and overall appearance consistency of the model. The final output 3D outer contour model takes into account the realistic structural form and height accuracy of the building, making it suitable for subsequent applications such as precise 3D architectural mapping, texture mapping, and urban digital modeling. It provides a complete and reliable spatial foundation for high-precision 3D reconstruction of BeiDou satellite positioning images.

[0032] In this embodiment, step S5 includes the following steps: Visual recognition of building surface texture is performed based on optimized image sequences, and texture layer mapping is performed on the three-dimensional outer contour model to construct a texture structure layer model; The true latitude and longitude coordinates of the area where the building is located are calculated based on the BeiDou satellite. Perform overall geometric correction on the texture structure layer model to generate a geometrically corrected model; Geographic coordinate annotation is performed on the geometric correction model based on the actual latitude and longitude coordinates, and a three-dimensional mapping model is output.

[0033] In this embodiment, each face or polygonal unit of the 3D outer contour model is matched with its corresponding 2D image projection area. Information such as grayscale, color histogram, edge density, and texture direction are extracted through image texture feature analysis. Combining multi-temporal images, temporal redundancy can be used to enhance the robustness of texture recognition. For example, feature fusion or weighted averaging can be performed on the textures of the same facade under different lighting conditions to mitigate the impact of lighting variations on texture recognition. The identified texture information is mapped onto the surface of the 3D outer contour model and processed in layers according to different levels: roof, facade, and ground, forming a layered texture structure model. Each layer records the texture type, texture intensity, and texture direction information corresponding to that building surface. The layering accuracy can be set according to the size of the building unit; for example, a roof unit is 2×2 meters, and a facade unit is 1×1 meters.

[0034] By combining satellite orbit parameters, BeiDou positioning data, and image capture time, the coordinates of the building projected onto the Earth's surface are calculated. First, the local reference coordinate system of the 3D model is aligned with the geographic coordinate system of the satellite imagery. Using the known satellite positions and attitudes, the model's origin and key points are mapped to the WGS84 coordinate system or a projected coordinate system. To ensure positional accuracy, weighted average positioning based on multi-satellite observation data is typically used. Simultaneously, the average coordinates of multiple vertices within the building's area are calculated to obtain the precise latitude and longitude coordinates of the building's overall center position. Positional accuracy can be set according to the satellite positioning accuracy; for example, under BeiDou B1 / B2 signal conditions, horizontal positioning accuracy can reach 1-2 meters, and vertical height accuracy can reach 2-3 meters.

[0035] The model's local reference coordinate system is rotated, translated, and scaled to align the model vertices with the geographic coordinate system while maintaining the texture and geometric proportions of each unit. Rotation correction is adjusted by calculating the angle between the model's principal axis and the azimuth of the BeiDou coordinate system. Translation correction is made by adjusting the difference between the model's center point and the ground coordinates mapped from the building's actual latitude and longitude. Scale correction can be fine-tuned according to the ratio of the actual ground length to the model's length, ensuring that the roof width and facade height match the actual building dimensions. For multi-layered texture models, each layer undergoes consistent geometric correction to ensure that the texture still accurately corresponds to the geometric structure after correction. Through overall geometric correction, a geometrically corrected model is generated, making the 3D model not only geometrically accurate but also highly consistent with geographic coordinates and texture distribution. The local coordinate system vertex coordinates of the model are added with the translation offset relative to the geographic coordinate system, and the rotated and scaled vertices are mapped to WGS84 or the selected projected coordinate system. Each 3D vertex records its corresponding latitude, longitude, and height information, giving the entire model a true location positioning capability within a global geographic reference frame. By associating texture layering information with vertex coordinates, the accuracy of each building surface in terms of spatial location, texture, and height is ensured. After annotation, the output 3D mapping model contains not only the complete geometric structure and height information of the building, but also texture layering data and geospatial positioning information. It can be directly used for applications such as urban digital modeling, 3D visualization, precise surveying, and navigation analysis, achieving high-precision 3D mapping of buildings based on BeiDou satellite positioning imagery.

[0036] In this embodiment, the specific steps for performing visual recognition of building surface texture based on optimized image sequences, performing texture layer mapping on the three-dimensional outer contour model, and constructing a texture structure layer model are as follows: Visual recognition of building surface texture is performed based on optimized image sequences to extract texture visual features; the texture visual features include material reflectance distribution, texture frequency features, and color gradient changes. Based on the texture visual features, spatial constraint matching is performed on the three-dimensional outer contour model to obtain texture matching information; Based on texture matching information, a texture layer mapping is performed on the 3D outer contour model to construct a texture structure layer model.

[0037] In this embodiment, each frame of the optimized image is converted into a grayscale image and a multi-channel color image. Texture details are enhanced through image filtering and local enhancement processing, such as high-pass filtering or adaptive local contrast enhancement methods. Texture analysis algorithms are used to extract the reflectivity distribution, texture frequency features, and color gradient changes of the building surface material. The material reflectivity distribution reflects the light reflection intensity of different areas of the building surface and can be obtained by correcting local grayscale values ​​and illumination angles. Texture frequency features can be extracted through local Fourier transform or wavelet transform to characterize the repetitive structure and texture periodicity of the building surface. Color gradient changes are obtained by calculating the gradient changes in the RGB or HSV space between pixels, used to distinguish different materials or surface decoration areas. To improve the robustness of multi-temporal recognition, texture features of the same surface in images from different times can be fused, for example, by using weighted averaging or principal component analysis to eliminate the influence of illumination and image noise. The spatial positions of each face or polygonal unit of the 3D outer contour model are mapped to the projection of the texture region in the optimized image, and the 2D image features are mapped to the 3D surface through a ray projection model. Matching is performed by combining visual texture features, using material reflectance similarity, texture frequency matching degree, and color gradient similarity as constraint indicators. A weighted synthesis is then used to obtain a unit matching score. To reduce false matches, spatial geometric constraints can be added, such as the continuity of texture features between adjacent surfaces, the connectivity between roof corners and facade edges, and the consistency of surface normals. A threshold is typically set during the matching process; matching scores higher than 0.7-0.8 are considered valid matches, while matches below the threshold are excluded or recalculated.

[0038] Successfully matched texture regions are layered according to building function and surface type, such as roof layer, facade layer, and ground or decorative layer. Each layer records corresponding texture type, texture direction, frequency, and reflectivity. The mapping process projects the 2D image texture directly onto the 3D model surface, smoothing the texture boundaries to ensure continuity between the texture and the geometric structure without significant misalignment. To ensure consistency of textures across multiple time phases, textures from images of the same 3D surface taken at different times can be fused, for example, by generating a unified texture layer using texture fusion weights or principal component analysis.

[0039] In this embodiment, a building surveying device based on BeiDou communication is provided for executing the building surveying method based on BeiDou communication as described above, including: The image acquisition unit is used to continuously image the target building area based on the BeiDou satellite and construct an optimized image sequence. The height inversion unit is used to invert building heights in optimized image sequences to obtain height distribution information. The feature point recognition unit is used to identify building structure feature points based on an optimized image sequence, and obtain the three-dimensional spatial coordinates of multiple feature points. The modeling unit is used to model the outer contour based on the three-dimensional spatial coordinates, correct the height deviation based on the height distribution information, and output a three-dimensional outer contour model. The coordinate annotation unit is used to calculate the true latitude and longitude coordinates of the three-dimensional outer contour model, perform geographic coordinate annotation, and output the three-dimensional mapping model.

[0040] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the building surveying method based on Beidou communication described above.

[0041] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the building surveying method based on BeiDou communication described above.

[0042] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0043] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein are implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A building mapping method based on Beidou communication, characterized in that, Includes the following steps: Step S1: Perform continuous imaging of the target building area based on BeiDou satellites and construct an optimized image sequence; Step S2: Perform building height inversion on the optimized image sequence to obtain height distribution information; Step S3: Based on the optimized image sequence, identify the building structure feature points to obtain the three-dimensional spatial coordinates of multiple feature points; Step S4: Model the outer contour based on the three-dimensional spatial coordinates, and correct the height deviation based on the height distribution information to output the three-dimensional outer contour model; Step S5: Calculate the true latitude and longitude coordinates of the three-dimensional outer contour model, perform geographic coordinate annotation, and output the three-dimensional mapping model.

2. The building mapping method based on Beidou communication according to claim 1, characterized in that, The specific steps of step S1 are as follows: Multi-temporal satellite images are obtained by continuously imaging the target building area using BeiDou satellites. The imaging parameters of the multi-temporal satellite images are analyzed to extract the imaging parameter set, which includes the imaging angle, azimuth angle, elevation angle, and shooting time. Based on the imaging parameter set, multi-temporal satellite images are time-series labeled, and the first image sequence is output. Frame-by-frame sharpness calculation is performed on the first image sequence to obtain the sharpness index of each frame. The sharpness index is compared based on a preset sharpness threshold, and image frames below the sharpness threshold are removed to obtain a second image sequence; The second image sequence is subjected to brightness equalization and contrast normalization to construct an optimized image sequence.

3. The building surveying method based on BeiDou communication according to claim 1, characterized in that, The specific steps of step S2 are as follows: Dynamic illumination tracking was performed on the optimized image sequence to obtain the facade illumination distribution information at different time periods; Based on the optimized image sequence, shadow regions are identified, and building shadow outlines are marked; Based on the facade lighting distribution information and building shadow outline, the shadow length, projection direction and offset over time are calculated to obtain the light time migration information; Building height is inverted based on the temporal migration information of light rays using the imaging parameter set, and height distribution information is obtained.

4. The building surveying method based on BeiDou communication according to claim 1, characterized in that, Step S3 is as follows: Building structural feature points are identified and geometric feature points are marked based on optimized image sequences; the geometric feature points include roof corner points, facade edge intersections, and contour polyline nodes. Calculate the two-dimensional spatial coordinates of the geometric feature points at different time periods; Spatial geometric constraint matching is performed on the two-dimensional spatial coordinates to obtain feature point matching information; Multi-visual geometric calculations are performed based on feature point matching information to obtain the three-dimensional spatial coordinates of multiple feature points.

5. The building surveying method based on BeiDou communication according to claim 1, characterized in that, The specific steps of step S4 are as follows: Based on the three-dimensional spatial coordinates, spatial topological relationship analysis is performed to determine the boundary orientation and structural connection relationship of the building's outer contour; Based on the boundary orientation and structural connection relationship, the outer contour is modeled to generate the building's outer contour model; Based on height distribution information, height distribution deviations are identified in the outer contour model, and local deviation areas are marked. The local deviation areas are smoothed to output a three-dimensional outer contour model.

6. The building surveying method based on BeiDou communication according to claim 1, characterized in that, The specific steps of step S5 are as follows: Visual recognition of building surface texture is performed based on optimized image sequences, and texture layer mapping is performed on the three-dimensional outer contour model to construct a texture structure layer model; The true latitude and longitude coordinates of the area where the building is located are calculated based on the BeiDou satellite. Perform overall geometric correction on the texture structure layer model to generate a geometrically corrected model; Geographic coordinate annotation is performed on the geometric correction model based on the actual latitude and longitude coordinates, and a three-dimensional mapping model is output.

7. The building surveying method based on BeiDou communication according to claim 6, characterized in that, The specific steps for performing visual recognition of building surface texture based on optimized image sequences, and for performing texture layer mapping on the three-dimensional outer contour model to construct a texture structure layer model are as follows: Visual recognition of building surface texture is performed based on optimized image sequences to extract texture visual features; the texture visual features include material reflectance distribution, texture frequency features, and color gradient changes. Based on the texture visual features, spatial constraint matching is performed on the three-dimensional outer contour model to obtain texture matching information; Based on texture matching information, a texture layer mapping is performed on the 3D outer contour model to construct a texture structure layer model.

8. A building surveying device based on BeiDou communication, characterized in that, The method for performing building surveying based on BeiDou communication as described in claim 1 includes: The image acquisition unit is used to continuously image the target building area based on the BeiDou satellite and construct an optimized image sequence. The height inversion unit is used to invert building heights in optimized image sequences to obtain height distribution information. The feature point recognition unit is used to identify building structure feature points based on an optimized image sequence, and obtain the three-dimensional spatial coordinates of multiple feature points. The modeling unit is used to model the outer contour based on the three-dimensional spatial coordinates, correct the height deviation based on the height distribution information, and output a three-dimensional outer contour model. The coordinate annotation unit is used to calculate the true latitude and longitude coordinates of the three-dimensional outer contour model, perform geographic coordinate annotation, and output the three-dimensional mapping model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the building surveying method based on Beidou communication as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the building surveying method based on Beidou communication as described in any one of claims 1 to 7.