Method and system for surveying and mapping area of peripheral region along railway based on unmanned aerial vehicle

Abstract

The embodiment of the invention provides a method and system for surveying and mapping the area of a peripheral area along a railway based on an unmanned aerial vehicle, and belongs to the technical field of area surveying and mapping. The method comprises the following steps: performing coordinate registration processing on laser point cloud data collected by an unmanned aerial vehicle and a railway red line vector diagram to obtain standard point cloud data and a red line reference boundary under a unified coordinate system; based on the standard point cloud data and the red line reference boundary, an edge point set of an actual occupied area outside the red line is extracted, and a closed boundary contour is obtained through fitting; based on the boundary contour obtained through fitting, unified plane projection and subdivision processing is executed, and a two-dimensional measurement grid used for area determination is generated; and based on the two-dimensional measurement grid, accumulating the area of each unit, and summarizing and outputting the total area and corresponding boundary coordinates of the red line external expansion region. According to the scheme, high-precision and full-process area surveying and mapping of the non-standard occupied area along the railway are realized, and the objectivity of boundary judgment and the stability of an area extraction result are improved.

Description

Technical Field

[0001] This invention relates to the field of area surveying technology, specifically to a method and system for surveying the area of ​​a railway line perimeter based on unmanned aerial vehicles (UAVs). Background Technology

[0002] The management of land use along railway lines has long relied on the red line boundary as the spatial basis for land use rights. However, in actual operation and construction, situations often arise where land use extends beyond the red line, such as road slopes extending outwards, waste material dumping, and vegetation spreading along slopes, creating "extended areas outside the red line." Although these areas are not within the legal boundaries, they are highly coupled with railway operations in space, and the accurate measurement of their occupied areas is directly related to management activities such as land use compliance assessment, compensation determination for land acquisition and relocation, and supervision of illegal construction.

[0003] Currently, measurements of such areas mostly rely on manual on-site reconnaissance or visual interpretation based on remote sensing images. This approach suffers from problems such as difficulty in closing the survey area, reliance on subjective experience for boundary determination, and poor accuracy in area calculation. In real-world scenarios, the areas extending beyond the boundary line often exhibit irregular spatial characteristics and lack clearly defined structural boundaries. The transition zones at the survey area edges often exist in a natural state, such as grassy slopes, gravel layers, or waste piles. Traditional methods struggle to determine accurate boundaries based on image textures or GPS data points. Furthermore, when the survey area does not form a naturally closed structure, the lack of suitable closure fitting and projection measurement mechanisms often leads to boundary instability and error accumulation in area calculations.

[0004] Therefore, in order to address the difficulties in boundary identification and uncertainty in area measurement in the extended areas outside the railway red line, it is urgent to establish a spatial mapping mechanism with boundary adaptability and high-precision area extraction capabilities to improve the measurement efficiency and judgment reliability of non-standard occupied areas along the railway. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for surveying the area of ​​railway perimeter based on unmanned aerial vehicles (UAVs), so as to at least solve the problems of difficulty in identifying the boundary of the extended area outside the red line, insufficient closure of the survey area, and unstable accuracy of the area measurement results in the prior art.

[0006] To achieve the above objectives, the first aspect of the present invention provides a method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs). The method includes: performing coordinate registration processing on laser point cloud data and railway redline vector map data collected by the UAV to obtain standard point cloud data and redline reference boundary in a unified coordinate system; extracting the edge point set of the actual occupied area outside the redline based on the standard point cloud data and redline reference boundary, and fitting a closed boundary contour; performing unified plane projection and subdivision processing based on the fitted boundary contour to generate a two-dimensional measurement grid for area determination; and accumulating the area of ​​each unit based on the two-dimensional measurement grid and summarizing and outputting the total area and corresponding boundary coordinates of the extended area outside the redline.

[0007] Optionally, coordinate registration processing is performed on the laser point cloud data and railway redline vector map data collected by the UAV to obtain standard point cloud data and redline reference boundaries in a unified coordinate system, including: Acquire time-synchronized location information output by the global positioning system on the drone, and flight attitude data output by the inertial navigation system; Based on the time synchronization location information and the flight attitude data, a spatial trajectory model of the laser point cloud is constructed, and attitude compensation and position calculation are performed on the original coordinates of each measurement point in the laser point cloud based on the spatial trajectory model. The compensated laser point cloud coordinates are projected and aligned with the geographic reference coordinate system used in the railway redline vector map to generate standard point cloud data and redline reference boundary under a unified coordinate system.

[0008] Optionally, based on standard point cloud data and the red line reference boundary, extract the edge point set of the actual occupied area outside the red line, including: Based on standard point cloud data, the elevation difference of each measurement point relative to its neighboring points is determined, and a point cloud elevation difference gradient map is constructed. Based on the point cloud elevation gradient map, the locations where the elevation difference is greater than a preset difference threshold are selected as the edge candidate point set. The spatial relationship between the candidate edge point set and the red line reference boundary is determined, and measurement points located outside the red line and whose elevation difference relative to their neighboring points is greater than a preset difference threshold are selected to form the edge point set.

[0009] Optionally, the fitting rule for the closed boundary profile is: The edge point set is sorted by curvature perception according to two-dimensional spatial coordinates, and edge path fitting is performed based on a smooth closed interpolation algorithm with minimum curvature change to form a closed boundary contour; wherein, If the sorted set of edge points does not form a closed boundary contour, the shortest connecting edge segment at the beginning and end is introduced during the fitting process to complete the edge path and form a closed boundary contour.

[0010] Optionally, the rules for implementing the uniform planar projection are as follows: The principal axis direction of the closed boundary profile obtained by fitting is calculated, and a local measurement reference plane is constructed based on the principal axis direction; wherein, The measurement reference plane establishes a two-dimensional Cartesian coordinate system with the geometric center point of the boundary contour as the origin and the principal axis direction as the X-axis direction. Transform the coordinates of all boundary points in the closed boundary profile to the measurement reference plane to obtain the projected profile of the boundary profile in the corresponding two-dimensional coordinate system. The smallest bounding rectangle region containing all projected points is calculated based on the set of boundary point coordinates in the projected profile, and this bounding rectangle region is used as the spatial boundary constraint for the measurement grid subdivision to limit the spatial range of the measurement area.

[0011] Optionally, the rules for generating the two-dimensional measurement grid are as follows: Within the minimum bounding rectangle, using the projected contour boundary as the forced constraint boundary, the Delaunay triangulation algorithm is executed to generate two-dimensional measurement mesh cells covering the interior of the closed boundary; where, A mesh quality factor is introduced for each meshing cell in the Delaunay triangulation algorithm to control quality and eliminate degenerate cells that do not meet the preset minimum angle requirement.

[0012] Optionally, based on the two-dimensional measurement grid, the area of ​​each unit is accumulated and the total area and corresponding boundary coordinates of the extended region outside the red line are summarized and output, including: For each triangular cell in the two-dimensional measurement grid, obtain the corresponding vertex coordinates and calculate the cell area according to the plane geometry rules to obtain the cell area of ​​the corresponding measurement grid; The total area is obtained by sequentially summing the cell areas of all measured grids; where, Before sequentially summing the cell areas of all measurement grids to obtain the total area, the cell area calculation results of each measurement grid are spatially marked and correlated with the closed boundary contour to ensure that all area results fall within the boundary range.

[0013] Optionally, the method further includes outputting measurement results, including: Combine the measured total area with the corresponding boundary coordinates into a planar region object, and export it in GeoJSON or Shapefile format. Generate metadata files related to this measurement, including acquisition time, point cloud density, attitude compensation parameters, and projection plane definition; wherein, The area object and the metadata file together constitute the structured mapping output result.

[0014] A second aspect of the present invention provides a system for mapping the area of ​​a railway perimeter based on an unmanned aerial vehicle (UAV). The system includes: a data acquisition unit for performing coordinate registration processing on laser point cloud data and railway redline vector map data acquired by the UAV to obtain standard point cloud data and redline reference boundary in a unified coordinate system; a fitting unit for extracting the edge point set of the actual occupied area outside the redline based on the standard point cloud data and the redline reference boundary, and fitting a closed boundary contour; a processing unit for performing unified plane projection and subdivision processing based on the fitted boundary contour to generate a two-dimensional measurement grid for area determination; and an output unit for accumulating the area of ​​each unit based on the two-dimensional measurement grid and summarizing and outputting the total area and corresponding boundary coordinates of the extended area outside the redline.

[0015] On the other hand, the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described method for surveying the area of ​​the railway perimeter based on unmanned aerial vehicles.

[0016] Through the above technical solution, this invention establishes a spatial reference system under a unified coordinate system by registering laser point cloud data with railway redline vector map data, effectively eliminating the spatial offset problem caused by inconsistent data sources; based on the unified spatial benchmark, it extracts the edge point set of the extended area outside the redline and fits it to generate a closed boundary contour, realizing the automatic identification and closure control of the boundary of the natural transition area; further, it projects and divides the closed contour into a structured two-dimensional measurement grid, ensuring the integrity and integrability of the area calculation range; finally, by accumulating the area of ​​each grid unit, it outputs the total area and boundary information of the extended area outside the redline, realizing high-precision, full-process area mapping of non-standard occupied areas along the railway, improving the objectivity of boundary judgment and the stability of area extraction results.

[0017] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of the steps of a method for surveying the area of ​​a railway perimeter based on an unmanned aerial vehicle (UAV) according to one embodiment of the present invention. Figure 2 This is a schematic diagram of the process of forming a closed path by fitting an edge point set according to one embodiment of the present invention; Figure 3This is a system structure diagram of a UAV-based system for surveying the area of ​​railway perimeter areas, provided in one embodiment of the present invention. Detailed Implementation

[0019] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0020] Figure 1 This is a flowchart illustrating the steps of a method for surveying the area of ​​a railway perimeter based on an unmanned aerial vehicle (UAV) according to one embodiment of the present invention. Figure 1 As shown, this invention provides a method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs), the method comprising: Step S10: Perform coordinate registration processing on the laser point cloud data and railway red line vector map collected by the UAV to obtain standard point cloud data and red line reference boundary under a unified coordinate system.

[0021] Specifically, the system acquires time-synchronized position information from the GPS system onboard the UAV and flight attitude data from the inertial navigation system; it constructs a spatial trajectory model of the laser point cloud based on the time-synchronized position information and the flight attitude data, and performs attitude compensation and position calculation on the original coordinates of each measurement point in the laser point cloud based on the spatial trajectory model; it then performs projection transformation and coordinate alignment between the compensated laser point cloud coordinates and the geographic reference coordinate system used by the railway redline vector map, generating standard point cloud data and redline reference boundary under a unified coordinate system.

[0022] In this embodiment of the invention, during the area surveying of the extended area outside the railway red line, to ensure the consistency of spatial data and measurement accuracy, it is necessary to perform coordinate registration processing on the laser point cloud data collected by the UAV and the railway red line vector map to construct the basic surveying data under a unified spatial reference system. This processing mainly includes two parts: first, performing attitude compensation and position calculation on the laser point cloud data to form standard point cloud data; second, aligning the standard point cloud data with the geographic reference coordinate system on which the railway red line vector map depends, ultimately achieving spatial benchmark unification and laying a high-precision data foundation for subsequent boundary identification and area calculation.

[0023] Specifically, in the laser point cloud acquisition phase, the system first acquires time-synchronized position information from the Global Positioning System (GPS) on the UAV platform, as well as flight attitude data from the Inertial Navigation System (INS), including roll, pitch, and yaw angles. GPS provides discrete spatiotemporal position points, while INS provides high-frequency attitude change information. By synchronizing the two in time, the complete spatial attitude trajectory of the aircraft during laser scanning can be constructed.

[0024] After acquiring the above data, the original relative coordinates of each laser point in three-dimensional space are constructed by combining the ranging angle, emission direction, and echo intensity information recorded internally by the lidar. Based on the position provided by GPS and the attitude angle provided by INS, a three-dimensional rigid body transformation is performed on these relative coordinates to complete the attitude compensation and geographic location calculation for each measurement point in the point cloud. This process essentially converts the relative ranging results in the airborne coordinate system to the absolute position representation in the geospatial reference system, and the output is a set of laser points containing X, Y, and Z coordinates, which is the attitude-corrected point cloud data.

[0025] Subsequently, the compensated point cloud data undergoes a geographic projection transformation to align its coordinate system with that used in the railway redline vector map. Railway land use data typically employs the Gauss-Kruger projection or the WGS84 coordinate system. The appropriate projection parameters must be selected based on the specific definition of the redline data to perform the coordinate transformation, avoiding spatial registration errors caused by data misalignment across different coordinate systems. During this process, key elements such as the projection zone number, central meridian, and ellipsoid parameters must be rigorously verified to ensure that the point cloud data and vector boundaries are aligned under the same geographic reference.

[0026] After the projection transformation is completed, the point cloud data is overlaid with the railway redline vector map, and spatial coordinate alignment verification is performed to confirm that the positional error of the two types of data in the boundary junction area is within the preset tolerance range. If there are minor deviations, fine-tuning can be performed based on common control points or boundary coincidence, ultimately generating standard point cloud data and redline reference boundaries in a unified coordinate system. This result serves as the basic input for subsequent boundary extraction, area subdivision, and projection calculation, offering technical advantages such as unified spatial reference, accurate attitude compensation, and traceable coordinate transformation.

[0027] This coordinate registration process not only solves the problem of inconsistency in coordinate systems between laser point clouds and railway vector maps, but also effectively eliminates spatial distortion of point clouds caused by changes in flight attitude. It provides a clear and accurate basic data input for the extraction of boundary contours and area measurement of the extended area outside the red line, significantly improving the reliability of subsequent boundary recognition and the spatial accuracy of measurement results.

[0028] In another possible implementation, to address the issues of GPS signal obstruction and attitude compensation inaccuracies in complex terrain areas, a registration method based on joint back-calculation of laser point clouds and ground control points is adopted to replace the traditional method of relying entirely on navigation units to construct spatial trajectory models. Specifically, several ground control points are pre-deployed in the survey area and recorded in a reference coordinate system in the form of known three-dimensional coordinates. During the laser scanning mission, the system marks the echo recognition results of these control points in the laser point cloud in real time and extracts a subset of the point cloud containing the features of the ground control points.

[0029] Subsequently, based on the spatial distribution characteristics of the control points and the positioning results in the actual point cloud, a rigid transformation model is used to perform coordinate inverse calculation on the entire point cloud dataset, automatically adjusting the original point cloud to the target geographic reference system, thus achieving spatial calibration under attitude disturbances. Compared to methods based on forward calculation of flight trajectories, this embodiment's inverse calibration method using known control points is more suitable for environmental conditions with strong interference, unstable attitude, or short-term interruptions in GNSS signals, improving the robustness of point cloud positioning. After the transformation, the point cloud data and the railway redline vector map are aligned in the same coordinate system to obtain standard point cloud data and redline reference boundaries under a unified spatial reference, which are used for subsequent boundary contour extraction and area calculation.

[0030] Step S20: Based on standard point cloud data and red line reference boundary, extract the edge point set of the actual occupied area outside the red line, and fit the closed boundary contour.

[0031] Specifically, based on standard point cloud data and the red line reference boundary, the edge point set of the actual occupied area outside the red line is extracted, including: determining the elevation difference of each measurement point relative to its neighboring points based on the standard point cloud data, and constructing a point cloud elevation difference gradient map; based on the point cloud elevation difference gradient map, selecting locations where the elevation difference is greater than a preset difference threshold as edge candidate point sets; and judging the spatial relationship between the edge candidate point set and the red line reference boundary, filtering out measurement points located outside the red line and whose elevation difference relative to their neighboring points is greater than the preset difference threshold, thus forming the edge point set.

[0032] Furthermore, the fitting rule for the closed boundary contour is as follows: the edge point set is sorted according to the curvature perception of the two-dimensional spatial coordinates, and the edge path is fitted based on the smooth closed interpolation algorithm with minimum curvature change to form the closed boundary contour; if the sorted edge point set does not form a closed boundary contour, the shortest connecting edge segment is introduced during the fitting process to complete the fitted edge path in order to form a closed boundary contour.

[0033] In this embodiment of the invention, after completing the unified coordinate registration of the standard point cloud data and the red line reference boundary, in order to further identify the actual occupied area of ​​the extended region outside the railway red line, it is necessary to extract the edge point set of this region from the standard point cloud data, and fit a closed boundary profile based on the edge point set for subsequent measurement grid subdivision and area calculation. This process needs to consider both terrain variation features and red line constraints to ensure that the identified boundary truly reflects the actual occupied profile and has high geometric stability and closure integrity.

[0034] Specifically, the extraction of edge point sets is based on standard point cloud data as input. First, the neighborhood relationship of each point cloud measurement point is constructed, often using a fixed search radius method or the k-nearest neighbor method to determine the local neighborhood point set. For each measurement point, the difference between its elevation in the Z direction and all its neighboring points is calculated, and the maximum elevation difference of that point is taken as the local elevation difference response value of that measurement point. Further, the elevation difference response values ​​of all measurement points are summarized across the entire point cloud to construct a point cloud elevation gradient map, forming a spatial elevation difference distribution dataset related to abrupt terrain changes. This elevation gradient map serves as the basis for boundary determination, revealing areas of undulation and abrupt change in local terrain, facilitating the identification of natural or man-made occupancy boundaries such as landform boundaries, deposit outlines, and vegetation boundaries.

[0035] After obtaining the elevation gradient map, preliminary screening is performed based on a set difference threshold to identify measurement points with elevation differences greater than the threshold, and these are used as the initial set of edge candidate points. The difference threshold should be adaptively adjusted according to the type of terrain features, point cloud density, and scanning accuracy. Generally, it is set to 3 to 5 times the standard deviation of point cloud noise, which can effectively eliminate misjudgment interference caused by local micro-undulations. To avoid misjudging natural undulation areas inside the red line as boundary areas, further spatial location filtering is required in conjunction with the red line reference boundary. That is, for each point in the preliminary edge candidate point set, it is determined whether it is located outside the red line reference boundary. If the spatial relationship constraint is satisfied and the maximum elevation difference of the point is still higher than the preset threshold, it is ultimately retained as an edge point and added to the edge point set.

[0036] It is important to note that the phrase "located outside the red line" does not simply mean determining that a point is in a certain direction outside the boundary line. Instead, it requires a rigorous geometric determination based on the vector polygon structure of the red line boundary, using a point-in-polygon determination method or constructing a spatial buffer zone. This avoids misjudgments of spatial relationships caused by the complex shape of the boundary line or measurement errors. The final set of edge points should possess the following characteristics: their spatial location is concentrated in the area outside the red line; their local terrain features exhibit significant abrupt changes in elevation; and their spatial distribution shows a continuous, banded characteristic along the edge of the actual occupied area.

[0037] After obtaining the edge point set, to ensure the validity and integrability of subsequent area measurements, the edge point set needs to be fitted into a continuous closed boundary profile. The fitting rules for the boundary profile are as follows: First, the edge point set is sorted according to its two-dimensional spatial coordinates. This sorting process takes into account the overall geometric distribution of the point set and typically uses a curvature-direction-based perceptual sorting method. Specifically, the outermost edge point in the spatial distribution is selected as the starting point, the direction of the line connecting it to its adjacent points is calculated, and this direction is used as the starting point to guide the entire point set to be sorted point by point along the path of minimum curvature change, thus restoring the actual boundary orientation as closely as possible.

[0038] After sorting the point set, a smooth closure interpolation algorithm is performed based on the sequence to construct a continuous boundary path. Commonly used interpolation methods include spline curve interpolation, B-spline fitting, or flexible fitting methods based on local weighted regression. To avoid sharp angles or contour breaks in sparse point regions, a minimum curvature change constraint is introduced during the fitting process, and the tangent angle between adjacent segments in the interpolation path is smoothly adjusted to ensure that the overall boundary contour presents a smooth closure. In cases of uneven edge point density or local discontinuities in the contour, this method can automatically mitigate the transition between paths, improving the integrity of the fitted contour.

[0039] When the fitted path fails to form a naturally closed structure, for example, if there is a spatial distance between the starting and ending points that exceeds the set closure tolerance, a contour patching mechanism is introduced. In this mechanism, the Euclidean distance between the first and last points is calculated, and this distance is used as a boundary patch, appended to the boundary contour as part of the fitted path. This patch can be formed using linear connections, circular arc connections, or interpolation based on local curvature trends, with the specific method chosen based on the difference in the direction vectors between the first and last points and the overall curvature trend of the point set. The introduction of this patch ensures the closure of the fitted path's topology and avoids mesh fragmentation or area loss problems caused by non-closed contours in subsequent meshing processes.

[0040] The resulting closed boundary profile is not only geometrically continuous and closed, but also spatially located strictly outside the red-line reference boundary. Furthermore, the corresponding edge points all exhibit abrupt elevation changes, accurately describing the outward boundary morphology of the actual occupied area outside the red line. This closed profile will serve as the spatial input boundary for subsequent projection subdivision and measurement mesh generation, playing a crucial role in defining the survey area and controlling the integration region during area calculation, effectively improving the boundary consistency of the area results and the geometric controllability of the survey area.

[0041] In one possible implementation, such as Figure 2 , Figure 2(a) shows the boundary contour fitting result under ideal conditions, where the edge point set is sorted by curvature perception and a continuous path is formed by fitting a smooth curve. The fitting path itself is naturally closed, and the first and last points are precisely aligned in space, ultimately generating a closed boundary contour.

[0042] Figure 2 (b) illustrates a case where the fitted path does not close naturally, meaning that after fitting the edge point set, the resulting path fails to close at both ends, leaving a certain spatial gap. To form a closed boundary structure, the shortest connecting edge segment is introduced during the fitting process to connect the first and last points of the path, completing the closed region. In the illustration, this connecting edge segment is represented by a dashed line to distinguish it from the original fitted curve. It should be noted that this connecting edge segment is not limited to an absolute straight line segment; its shape can be adjusted according to the actual spatial distribution between the first and last points, the overall trend of the boundary, and the fitting strategy. For example, in the case of curvature constraints, the introduced connecting edge segment can be a smooth curve. However, in all cases, the principle of "completing the shortest feasible path between the first and last points" should be satisfied to ensure that the contour structure is geometrically and topologically closed, avoiding boundary breaks or closure failures during subsequent measurement region subdivision.

[0043] Through the above-mentioned edge point extraction and contour fitting process, the problems of difficulty in identifying boundary transition zones and blurred boundaries of natural occupied areas in traditional image interpretation methods have been solved. An automatic boundary identification mechanism based on elevation difference features and spatial constraints has been constructed, providing a strong geometric support foundation for the accurate surveying and mapping of railway red-line extension areas. This effectively improves the objectivity, reproducibility and applicability of boundary extraction in non-standard land use surveying tasks outside the red line.

[0044] Step S30: Based on the fitted boundary contour, perform unified plane projection and subdivision processing to generate a two-dimensional measurement mesh for area determination.

[0045] Specifically, based on the fitted boundary contour, a unified planar projection and meshing process is performed, including: calculating the principal axis direction of the fitted closed boundary contour, and constructing a local measurement reference plane based on the principal axis direction; wherein, the measurement reference plane establishes a two-dimensional Cartesian coordinate system with the geometric center point of the boundary contour as the origin and the principal axis direction as the X-axis direction; transforming the coordinates of all boundary points in the closed boundary contour to the measurement reference plane to obtain the projected contour of the boundary contour in the corresponding two-dimensional coordinate system; calculating the minimum bounding rectangle region containing all projected points based on the set of boundary point coordinates in the projected contour, and using the bounding rectangle region as the spatial boundary constraint for the measurement mesh meshing to limit the spatial range of the measurement area.

[0046] Furthermore, the generation rule for the two-dimensional measurement mesh is as follows: within the minimum bounding rectangle, the projected contour boundary is used as the forced constraint boundary, and the Delaunay triangulation algorithm is executed to generate two-dimensional measurement mesh cells covering the closed boundary; wherein, a mesh quality factor is introduced for each triangulation cell in the Delaunay triangulation algorithm for quality control, so as to eliminate degenerate cells that do not meet the preset minimum angle requirement.

[0047] In this embodiment of the invention, after identifying the edge point set and smoothly fitting the boundary contour, to achieve accurate area measurement of the extended region outside the red line, it is necessary to further project the boundary contour in three-dimensional space onto a unified two-dimensional measurement plane, and perform mesh generation within this plane to generate a well-structured and geometrically closed measurement mesh to support quantitative integration calculation of the area. This process involves not only coordinate system transformation and normalization, but also precise closure constraints on the boundary of the survey area, and the generation of high-quality mesh cells through a suitable meshing algorithm to ensure the stability of subsequent area calculations and the repeatability of measurement results.

[0048] Specifically, the principal axis direction is first determined based on the fitted closed boundary profile. The principal axis direction is the main reference axis describing the extension direction of the boundary profile in the two-dimensional plane, used to construct a local coordinate reference system to improve the clarity of the projected boundary shape. The principal axis direction can be calculated by using the least squares fitting ellipse principal axis method of the boundary point set, or by obtaining the first principal component direction as the principal axis direction vector through principal component analysis (PCA) of the covariance matrix of the boundary point set. Taking the ellipse fitting method as an example, let the two-dimensional coordinates of the boundary point set under spatial projection be (xi, yi). By minimizing the sum of squared residuals from the boundary points to the edge of the ellipse through the fitting function, the major axis direction of the fitted ellipse is obtained, which is the principal axis direction of the boundary profile.

[0049] After obtaining the principal axis direction, a local measurement reference plane is constructed. This reference plane uses the geometric center point of the boundary contour as the origin and the principal axis direction as the X-axis, constructing a two-dimensional Cartesian coordinate system. The geometric center point can be obtained by averaging the X and Y coordinates of all boundary points, representatively representing the center position of the boundary region. To ensure good coordinate orthogonality of this coordinate system, a rotation transformation matrix is ​​further constructed by calculating the angle between the principal axis direction and the global reference coordinate axis, mapping the coordinates of all boundary points from the original global coordinate system to this local two-dimensional coordinate system, thus achieving planar projection of the boundary points.

[0050] During the projection process, the coordinates of each boundary point undergo the following transformation: First, the point's position vector is translated to a local coordinate system with the geometric center as the origin. Then, a linear transformation is performed on this vector based on the rotation matrix constructed according to the principal axis directions to obtain the two-dimensional projected coordinates in the local coordinate system. The mathematical expression for this transformation is: ; Where (x,y) are the original coordinates, (x... c ,y c Let R be the coordinates of the geometric center point, and R be a two-dimensional rotation matrix constructed based on the principal axis direction. Through this coordinate transformation, the coordinate expressions of all boundary points in the two-dimensional measurement reference plane can be obtained, thus forming a "projected profile".

[0051] After obtaining the projected profile, to clarify the boundary control range of the mesh subdivision, it is necessary to further construct the minimum bounding rectangle covering the profile region. The minimum bounding rectangle is the smallest area rectangle containing all projected boundary points, with its boundary direction aligned with the coordinate axes of the measurement plane, used to define the spatial encapsulation range of subsequent subdivision operations. The minimum bounding rectangle can be calculated based on the set of boundary points using a rotating caliper algorithm or a convex hull envelope method, and its corner coordinates are used to define the spatial bounding box of the measurement region. This bounding box, on the one hand, restricts the boundary region of the generated mesh, and on the other hand, serves as the basic unit for measurement graphic visualization and data structure organization.

[0052] After completing the construction of the measurement plane and the projection contour envelope, the next step is to mesh the 2D measurement grid. The meshing operation must be performed based on the principles of high conformality, integrability, and boundary constraints, with the Delaunay triangulation algorithm being the preferred choice. This algorithm generates geometrically high-quality triangular meshes while ensuring no mesh overlap and maximizing the minimum angle, which is beneficial for subsequent integration operations and graphical analysis. Before performing Delaunay meshing, the projection contour must be forcibly input as a boundary constraint, ensuring that the meshing process is only performed within the closed boundary, and that all generated triangular mesh elements are constrained within this closed boundary. To achieve this, a constrained Delaunay triangulation (CDT) strategy can be adopted, introducing boundary vertex chain constraints into the standard meshing algorithm to ensure that the boundary is accurately propagated into the meshed structure.

[0053] During the meshing process, mesh quality control constraints are applied to each triangular element. Typically, the minimum interior angle of each triangle is set to be no less than 20° or the maximum aspect ratio to not exceed a set threshold, to avoid integration distortion or computational instability caused by sharp corners or elongated shapes. Quality control can be achieved by performing local optimization operations (such as edge flipping and node smoothing) after the initial meshing, or by preprocessing before meshing, such as adjusting the meshing density through density control point placement. If degenerate elements (such as triangles with extremely small areas and extremely sharp angles) are generated during the meshing process, element filtering and removal operations should be performed after meshing to remove mesh elements that do not meet the quality factor, and the mesh should be re-meshed and completed based on the adjacent structure.

[0054] Furthermore, to ensure the integrity of the area calculation closure, connectivity analysis and boundary closure verification must be performed on the final generated two-dimensional measurement mesh structure. This can be achieved by traversing the spatial relationships between mesh boundary cells and the projected contour, confirming that all mesh boundaries strictly coincide with the closed contour, thus avoiding edge crossings, missing edges, or disconnected regions during the meshing process. If isolated mesh segments or internal hole regions are found to be uncovered, a meshing completion mechanism or local mesh expansion operation can be introduced to close and complete the mesh, ensuring the integrity of the topological structure of the measurement area.

[0055] The resulting two-dimensional measurement grid consists of a large number of high-quality triangular units with confined boundaries, and has the following technical advantages: First, all grid units completely cover the closed boundary of the extended area outside the red line, ensuring the consistency of the measurement area for area calculation; Second, all grid units meet geometric quality constraints, avoiding the accumulation of calculation errors; Third, the boundary contours are completely embedded in the grid structure in a structured manner, providing high traceability and user-friendly graphical visualization.

[0056] This processing not only establishes a complete geometric mapping mechanism from 3D boundary contours to 2D integral regions, but also provides a precise, closed, and structurally stable grid carrier for area measurement. It exhibits high adaptability and repeatability in automated mapping of large-scale railway land use areas outside the red line. Especially in scenarios with complex terrain, irregular land use areas, or discontinuous boundaries, the measurement grid constructed through unified projection and high-quality subdivision effectively avoids problems such as boundary distortion, projection skew, or area omissions inherent in traditional image-based or manual estimation methods. This enhances the technical reliability and automation level of identifying and extracting areas of non-planned regions in railway land use supervision.

[0057] Step S40: Based on the two-dimensional measurement grid, accumulate the area of ​​each unit and summarize and output the total area and corresponding boundary coordinates of the extended area outside the red line.

[0058] Specifically, for each triangular cell in the two-dimensional measurement grid, the corresponding vertex coordinates are obtained and the cell area is calculated according to the plane geometry rules to obtain the cell area of ​​the corresponding measurement grid; the cell areas of all measurement grids are sequentially summed to obtain the total area; before the process of sequentially summing the cell areas of all measurement grids to obtain the total area, the cell area calculation results of each measurement grid are spatially marked and correlated with the closed boundary contour to ensure that all area results fall within the boundary range.

[0059] In this embodiment of the invention, after the projection and subdivision of the two-dimensional measurement grid are completed, it is necessary to further calculate the actual occupied area of ​​the extended region outside the red line based on the generated measurement grid structure, and output the corresponding boundary information for visualization and data archiving. To achieve this goal, the area of ​​each triangular unit constituting the two-dimensional measurement grid must first be calculated, and then the area results of all units are accumulated one by one to finally obtain the complete total area value. At the same time, a spatial consistency verification mechanism must be introduced during the area calculation process to ensure that each grid unit included in the total area falls strictly within the previously fitted closed boundary contour, avoiding area deviations caused by boundary errors or subdivision redundancy.

[0060] Specifically, for each triangular element in the two-dimensional measurement mesh, the coordinates of its three corresponding vertices are extracted sequentially. The vertex coordinates originate from the set of nodes within the projected boundary during the meshing process. The coordinate values ​​reside in the local two-dimensional Cartesian coordinate system established by the measurement reference plane and are typically expressed as floating-point numbers. For any triangular element, its area can be calculated using the classic plane geometric area formula, i.e.: ; Where (x1, y1), (x2, y2), and (x3, y3) are the coordinates of the three vertices of the triangle in the measurement reference plane, respectively. The above calculation method has advantages such as strong versatility, numerical stability, and ease of batch processing, making it suitable for automated execution in large-scale mesh cell structures.

[0061] After calculating the unit area, to ensure the spatial accuracy of the area statistics, it is necessary to mark the spatial location of each calculated unit area result and perform boundary consistency verification. Specifically, using the centroid coordinates of each triangular unit as a spatial marker, the known polygonal boundary structure in the fitted boundary contour is called to perform a point-in-polygon (point inside the polygon) determination. If the centroid is located within the closed boundary contour, the area of ​​that triangular unit is included in the valid statistical range; conversely, if the centroid falls outside the boundary or within the boundary overlap error zone, it should be discarded or further reclassified as appropriate.

[0062] The aforementioned spatial consistency verification plays a crucial role in area accuracy. Due to potential errors such as error propagation, incomplete boundary alignment, or overlapping of mesh elements at corners during the meshing process, failure to determine the spatial assignment of mesh cells can easily lead to redundancy or undercounting of areas. For example, when the boundary profile is curved or has a complex structure, some cells generated during meshing may cross boundary curves or contain locally overlapping areas. Without boundary verification, the cell areas may be mistakenly counted as "legitimate survey areas," thus affecting the accuracy of the overall results.

[0063] After spatial location determination, the area values ​​of all grid cells that passed the determination are accumulated one by one. Area accumulation is performed using a linear summation method, which can be executed in cell number order or spatial partition order. The total area value can be expressed as: ; in, This represents the area of ​​the i-th triangular mesh cell that passes boundary verification, where n is the total number of valid cells that pass the screening. This total area value serves as the final measurement result for the extended area outside the red line, possessing characteristics such as boundary closure control, consistency of projected coordinates, and traceability of mesh structure. It facilitates comparison, linkage, and result verification with information such as cadastral maps, planning lines, and land use certificates.

[0064] Preferably, the method further includes outputting measurement results, including: combining the measured total area with the corresponding boundary coordinates into a planar region object and exporting it in GeoJSON or Shapefile format; generating a metadata file related to this measurement, the metadata file including acquisition time, point cloud density, attitude compensation parameters and projection plane definition; wherein, the planar region object and the metadata file together constitute a structured mapping output result.

[0065] In this embodiment of the invention, after calculating the area of ​​the extended region outside the red line, to meet the requirements of standardized management and cross-platform use of the measurement results, the calculation results and related data need to be structured and encapsulated and output to a data format with a universal data interface, forming a set of surveying and mapping results files that can be accessed and reused long-term. This process includes not only the output of the core measurement results—total area and boundary data—but also the archiving of metadata for key control parameters and data processing constraints during the measurement task, to ensure that the source of the results is traceable, the calculation process is traceable, and the output results are analyzable.

[0066] Specifically, the total area of ​​the region extending beyond the red line is combined with the corresponding set of closed boundary contour coordinates to construct a standard areal region object. This areal region object uses the closed boundary contour as its geometric boundary, the total area value as an additional attribute field, and includes spatial order information of the contour points to ensure consistent topological representation across various geographic information platforms. This object can be output as a GeoJSON or Shapefile file, depending on the application requirements. GeoJSON format is suitable for lightweight interaction on web platforms and embedding in map browsers, while Shapefile format is suitable for loading, analysis, and spatial database management in traditional GIS software environments.

[0067] While generating the areal region object, a metadata file associated with this measurement process also needs to be constructed to record the spatial control parameters and processing status information involved in the measurement task. The metadata file includes, but is not limited to, the following fields: acquisition time (indicating the timestamp of point cloud data acquisition), point cloud density (expressing spatial coverage accuracy), attitude compensation parameters (describing the initial attitude matrix or inertial navigation correction model used for point cloud data attitude correction), and projection plane definition (clarifying the local coordinate system parameters upon which the boundary contour projection is based, such as origin coordinates, principal axis direction, and projection type). Furthermore, auxiliary information such as the measurement area number, operator identification, and data processing version number can be added to enhance the ownership and management capabilities of the surveying data.

[0068] Finally, the aforementioned areal area objects are associated and packaged with metadata files to form a structured surveying output. This result can serve as a basic survey area unit, directly linked to the railway asset management system, land use review platform, or local planning archives, enabling the archiving, verification, and visual tracking of survey area data. The structured output not only improves the data integrity and interoperability of the survey results but also lays a standardized data foundation for the system-wide use of survey area results in actual management, approval, and auditing processes, possessing strong application and promotion value.

[0069] Figure 3 This is a system structure diagram of a UAV-based area mapping system for the surrounding area of ​​a railway line, provided in one embodiment of the present invention. Figure 3As shown, this invention provides an area mapping system for railway perimeter areas based on unmanned aerial vehicles (UAVs). The system includes: a data acquisition unit, used to perform coordinate registration processing on laser point cloud data and railway redline vector map data acquired by the UAV to obtain standard point cloud data and redline reference boundary in a unified coordinate system; a fitting unit, used to extract the edge point set of the actual occupied area outside the redline based on the standard point cloud data and redline reference boundary, and fit to obtain a closed boundary contour; a processing unit, used to perform unified plane projection and subdivision processing based on the fitted boundary contour to generate a two-dimensional measurement grid for area determination; and an output unit, used to accumulate the area of ​​each unit based on the two-dimensional measurement grid and summarize and output the total area and corresponding boundary coordinates of the extended area outside the redline.

[0070] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described method for surveying the area of ​​the railway perimeter based on unmanned aerial vehicles (UAVs).

[0071] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a microcontroller, chip, or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0072] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details described above. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe the various possible combinations.

[0073] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the embodiments of the present invention, they should also be regarded as the content disclosed by the embodiments of the present invention.

Claims

1. A method for surveying the area of ​​a railway line's surrounding region based on unmanned aerial vehicles (UAVs), characterized in that, The method includes: Coordinate registration processing was performed on the laser point cloud data and railway red line vector map data collected by the UAV to obtain standard point cloud data and red line reference boundary under a unified coordinate system; Based on standard point cloud data and red line reference boundaries, the edge point set of the actual occupied area outside the red line is extracted, and the closed boundary contour is obtained by fitting. Based on the fitted boundary contour, a unified plane projection and subdivision process is performed to generate a two-dimensional measurement mesh for area determination. Based on the two-dimensional measurement grid, the area of ​​each unit is accumulated and the total area and corresponding boundary coordinates of the extended area outside the red line are summarized and output.

2. The method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, Coordinate registration processing was performed on the laser point cloud data and railway redline vector map data collected by the UAV to obtain standard point cloud data and redline reference boundaries in a unified coordinate system, including: Acquire time-synchronized location information output by the global positioning system on the drone, and flight attitude data output by the inertial navigation system; Based on the time synchronization location information and the flight attitude data, a spatial trajectory model of the laser point cloud is constructed, and attitude compensation and position calculation are performed on the original coordinates of each measurement point in the laser point cloud based on the spatial trajectory model. The compensated laser point cloud coordinates are projected and aligned with the geographic reference coordinate system used in the railway redline vector map to generate standard point cloud data and redline reference boundary under a unified coordinate system.

3. The method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, Based on standard point cloud data and the red line reference boundary, the edge point set of the actual occupied area outside the red line is extracted, including: Based on standard point cloud data, the elevation difference of each measurement point relative to its neighboring points is determined, and a point cloud elevation difference gradient map is constructed. Based on the point cloud elevation gradient map, the locations where the elevation difference is greater than a preset difference threshold are selected as the edge candidate point set. The spatial relationship between the candidate edge point set and the red line reference boundary is determined, and measurement points located outside the red line and whose elevation difference relative to their neighboring points is greater than a preset difference threshold are selected to form the edge point set.

4. The method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs) according to claim 3, characterized in that, The fitting rule for the closed boundary contour is: The edge point set is sorted by curvature perception according to two-dimensional spatial coordinates, and edge path fitting is performed based on a smooth closed interpolation algorithm with minimum curvature change to form a closed boundary contour; wherein, If the sorted set of edge points does not form a closed boundary contour, the shortest connecting edge segment at the beginning and end is introduced during the fitting process to complete the edge path and form a closed boundary contour.

5. The method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, Based on the fitted boundary contour, a unified plane projection and subdivision process is performed, including: The principal axis direction of the closed boundary profile obtained by fitting is calculated, and a local measurement reference plane is constructed based on the principal axis direction; wherein, The measurement reference plane establishes a two-dimensional Cartesian coordinate system with the geometric center point of the boundary contour as the origin and the principal axis direction as the X-axis direction. Transform the coordinates of all boundary points in the closed boundary profile to the measurement reference plane to obtain the projected profile of the boundary profile in the corresponding two-dimensional coordinate system. The smallest bounding rectangle region containing all projected points is calculated based on the set of boundary point coordinates in the projected profile, and this bounding rectangle region is used as the spatial boundary constraint for the measurement grid subdivision to limit the spatial range of the measurement area.

6. The method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs) according to claim 5, characterized in that, The rules for generating two-dimensional measurement grids are as follows: Within the minimum bounding rectangle, using the projected contour boundary as the forced constraint boundary, the Delaunay triangulation algorithm is executed to generate two-dimensional measurement mesh cells covering the interior of the closed boundary; where, A mesh quality factor is introduced for each meshing cell in the Delaunay triangulation algorithm to control quality and eliminate degenerate cells that do not meet the preset minimum angle requirement.

7. The method for surveying the area of ​​a railway perimeter based on an unmanned aerial vehicle (UAV) according to claim 6, characterized in that, Based on the two-dimensional measurement grid, the area of ​​each unit is accumulated and the total area and corresponding boundary coordinates of the extended area outside the red line are summarized and output, including: For each triangular cell in the two-dimensional measurement grid, obtain the corresponding vertex coordinates and calculate the cell area according to the plane geometry rules to obtain the cell area of ​​the corresponding measurement grid; The total area is obtained by sequentially summing the cell areas of all measured grids; where, Before sequentially summing the cell areas of all measurement grids to obtain the total area, the cell area calculation results of each measurement grid are spatially marked and correlated with the closed boundary contour to ensure that all area results fall within the boundary range.

8. The method for surveying the area of ​​a railway perimeter based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method also includes outputting measurement results, including: Combine the measured total area with the corresponding boundary coordinates into a planar region object, and export it in GeoJSON or Shapefile format. Generate metadata files related to this measurement, including acquisition time, point cloud density, attitude compensation parameters, and projection plane definition; wherein, The area object and the metadata file together constitute the structured mapping output result.

9. A system for surveying the area of ​​a railway line's surrounding region based on unmanned aerial vehicles (UAVs), characterized in that, The system includes: The acquisition unit is used to perform coordinate registration processing on the laser point cloud data and railway red line vector map data acquired by the UAV, so as to obtain standard point cloud data and red line reference boundary under a unified coordinate system; The fitting unit is used to extract the edge point set of the actual occupied area outside the red line based on standard point cloud data and red line reference boundary, and fit it to obtain the closed boundary contour. The processing unit is used to perform unified planar projection and subdivision processing based on the fitted boundary contour to generate a two-dimensional measurement mesh for area determination. The output unit is used to accumulate the area of ​​each unit based on the two-dimensional measurement grid and summarize and output the total area and corresponding boundary coordinates of the extended area outside the red line.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the UAV-based method for surveying the area of ​​a railway perimeter as described in any one of claims 1-8.

Images