A building three-dimensional model rapid construction method and system for light-weight profile line expression

Abstract

This invention discloses a method and system for rapid construction of 3D building models using lightweight contour line representation. The method includes: acquiring and geometrically optimizing the contour line structural surface data of lightweight representation; performing 2D mapping and 2D spatial decomposition on the regular 3D polygonal structural surface data to obtain multiple candidate surface primitives; constructing a graph cut optimization model based on the top and facade structural surfaces, and using the graph cut optimization model to select effective surface primitives from the multiple candidate surface primitives; mapping the effective surface primitives back to 3D space, and determining the corner elevation values ​​of the effective surface primitives according to their type; and performing 2D-to-3D extrusion based on the corner elevation values ​​of the effective surface primitives to obtain the 3D building model. This invention effectively solves the problems of insufficient modeling accuracy and structural logic errors caused by sparse and missing point clouds, improving model construction efficiency while preserving the structural details of the building, and achieving complete 3D building model construction.

Description

Technical Field

[0001] This invention relates to the field of 3D modeling technology, specifically to a method and system for rapid construction of 3D building models oriented towards lightweight contour line expression. Background Technology

[0002] Building structured 3D models is one of the core applications of lightweight point cloud data representation. Its core is to use lightweight contour data to build high-precision, fine-grained 3D models. The constructed models can provide a highly available 3D spatial foundation for application scenarios such as smart cities and digital twins, and are a key technology to support the construction of large-scale 3D models at the city level.

[0003] Current mainstream methods for 3D building reconstruction mostly employ a zenith-direction structural surface extension strategy. This method involves: first, extracting the roof planar structure using a planar segmentation algorithm; then, projecting the 2D planar boundary onto a horizontal plane to generate a polygonal outline; finally, stretching along the elevation direction to generate a 3D structural frame with height, and generating a side mesh through triangulation. While this method offers some efficiency advantages in reconstructing simple, regular buildings, its application in complex urban architectural scenarios reveals two core limitations: insufficient model accuracy and the presence of structural logic errors, making it difficult to meet the demands for high-precision and high-completeness modeling.

[0004] In scenarios where airborne laser scanning is used to acquire point cloud data of buildings, the point cloud data of the building's side walls is often incomplete due to factors such as laser incident angle and occlusion. Therefore, traditional methods can only rely on relatively complete top surface point cloud data for 3D reconstruction, completely ignoring the constraint effect of the side structural surfaces. However, with the development and improvement of laser point cloud acquisition equipment, devices such as UAV longitudinal oblique photogrammetry and mobile handheld scanners can now accurately scan the point cloud of the building's external contour, and the point cloud data of the building's facade can also be accurately acquired. How to integrate the facade point cloud structural frame data into the constraint system of the building's structured modeling has become an urgent problem to be solved to improve the accuracy of building 3D modeling.

[0005] Meanwhile, traditional methods generate a 3D structural frame by directly extruding the roof structure. The accuracy of the building structure reconstruction depends entirely on the completeness and correctness of the structural surface segmentation. In practical applications, various structural logic errors are prone to occur due to structural surface extraction problems: due to structural surface extraction errors, there may be cases where incorrect facades cross another structure; when multiple structural surfaces intersect vertically, there may be a phenomenon of structures being embedded into each other; and incomplete structural surface extraction may also lead to gaps inside the 3D building model. These problems seriously reduce the practicality and rationality of the model.

[0006] Furthermore, while existing 3D spatial decomposition modeling methods, such as Polyfit, are widely used in the construction of structured building models, these methods generate 3D tangents and planes based on pairwise cutting of structural surfaces in 3D space. This results in a much higher geometric computational load than 2D spatial decomposition methods, leading to lower modeling efficiency. Moreover, they heavily rely on the attachment of the original point cloud of the building during structural optimization, limiting their effectiveness in scenarios with sparse or missing point clouds. Additionally, during structural surface extraction or vectorization data acquisition, irregular and distorted polygonal structural surfaces are easily generated due to measurement errors or connections. If directly used for modeling, these surfaces can further generate intersecting structures, reducing model quality.

[0007] In summary, traditional 3D building modeling methods suffer from problems such as limited structural constraints, low modeling efficiency, insufficient model accuracy and completeness, and susceptibility to structural logic errors. There is an urgent need for a method that can take into account multiple types of structural constraints, does not rely on original point clouds, and can efficiently construct complete 3D models to adapt to the development of current laser point cloud acquisition technology and meet the practical application needs of smart cities, digital twins, and other scenarios. Summary of the Invention

[0008] This invention provides a method and system for rapid construction of 3D building models with lightweight contour lines, which effectively solves problems such as insufficient modeling accuracy and structural logic errors caused by sparse and missing point clouds. It improves model construction efficiency while preserving the structural details of buildings, realizes the construction of complete 3D building models, and provides technical support for the construction of large-scale 3D models at the city level.

[0009] In a first aspect, the present invention provides a method for rapidly constructing a 3D building model for lightweight contour line expression, the method comprising: Obtain lightweight contour line structure surface data and perform geometric optimization on the contour line structure surface data to obtain regular three-dimensional polygon structure surface data; The regular three-dimensional polygonal structure surface data is mapped in two dimensions to obtain a set of two-dimensional line elements; and the set of two-dimensional line elements is decomposed in two dimensions to obtain multiple candidate surface primitives. Based on the top and elevation structural surfaces in the regular 3D polygon structural surface data, a graphical cut optimization model is constructed to characterize the surface area of ​​the building, and the effective surface primitives are selected from multiple candidate surface primitives using the graphical cut optimization model. The effective surface primitives are mapped back to three-dimensional space, and the corner elevation values ​​of the effective surface primitives are determined according to their type. Based on the corner elevation values ​​of the effective surface primitives, two-dimensional to three-dimensional extrusion is performed to obtain the three-dimensional model of the building.

[0010] In some embodiments of the present invention, a graphical cut optimization model for characterizing the surface region of a building is constructed based on the top and elevation structural surfaces in regular three-dimensional polygonal structural surface data. The graphical cut optimization model is then used to select effective surface primitives from multiple candidate surface primitives, including: Structural surfaces are classified into top and elevation surfaces using structural surface normal vectors; Based on the structural lines of the top surface structure, construct top surface shape constraints to characterize the shape of the building's outer surface; Based on the structural lines of the facade structure, facade edge constraints are constructed to characterize the actual area of ​​the building; Weighted fusion of top surface shape constraints and facade edge constraints yields fused constraints. A graph cut optimization model is constructed based on fusion constraints, and effective surface primitives are selected from multiple candidate surface primitives by minimizing the energy function of the graph cut optimization model.

[0011] In some embodiments of the present invention, a top surface shape constraint for characterizing the shape of the building's outer surface is constructed based on the structural lines of the top surface structural plane, including: The target region is discretized into grid points using a preset sampling resolution. The target region is the smallest bounding box containing the set of line elements. Starting from a grid point, draw rays along any angle direction; The top surface occlusion rate of the grid points is determined based on the number of intersections between the structural lines and rays on the top surface and the total number of rays. The top surface shape constraint is determined based on the top surface occlusion rate of each grid point; Based on the structural lines of the facade structural surfaces, facade edge constraints are constructed to characterize the actual area of ​​the building, including: The facade occlusion rate of the grid points is determined based on the number of intersections between the structural lines and rays of the facade structure and the total number of rays. The facade edge constraints are determined based on the facade occlusion rate of each grid point.

[0012] In some embodiments of the present invention, a graph cut optimization model is constructed based on fusion constraints, and the energy function of the graph cut optimization model is minimized, including: The data terms of the graph cut optimization model are constructed based on the mean of each grid point within the surface primitive, and the smoothing terms of the graph cut optimization model are constructed based on the difference between the data terms of adjacent surface primitives. Based on the data terms and the smoothing term, determine the energy function, solve for the minimum energy function, and select the effective surface primitives.

[0013] In some embodiments of the present invention, effective surface primitives are mapped back to three-dimensional space, and the corner elevation values ​​of the effective surface primitives are determined according to the type of the effective surface primitives; based on the corner elevation values ​​of the effective surface primitives, a two-dimensional to three-dimensional extrusion is performed to obtain a three-dimensional model of the building, including: The effective surface primitives are divided into intersection surface primitives and non-intersection surface primitives; For intersection surface primitives, the corner elevation values ​​of the intersection surface primitives are determined based on the three-dimensional structural surfaces corresponding to the intersection surface primitives. For non-intersecting surface primitives, determine the corner elevation values ​​of the non-intersecting surface primitives based on the surface primitives with determined elevation values ​​in their neighborhood; Based on the corner elevation values ​​of the intersection surface primitives and the corner elevation values ​​of the non-intersection surface primitives, the effective surface primitives in the two-dimensional plane are mapped to three-dimensional structural surface primitives. A three-dimensional model of a building is constructed based on the connection relationship between the three-dimensional structural surface primitives and the corresponding two-dimensional surface primitives.

[0014] In some embodiments of the present invention, a three-dimensional model of a building is constructed based on the connection relationship between three-dimensional structural surface primitives and corresponding two-dimensional surface primitives, including: Construct a minimum bounding box based on the initial structural surface, and determine the elevation of the lowest point of the minimum bounding box as the ground elevation value; By using the ground elevation value as the corner points of the two-dimensional surface element corresponding to the effective surface element, we obtain the two-dimensional surface element located at the ground height. Using a two-dimensional surface primitive at ground level as the lower base and a three-dimensional structural surface primitive as the upper base, connect the corresponding corner points of the lower base and the upper base to construct the vertical surface corresponding to the three-dimensional structural surface primitive. A three-dimensional model of the building is constructed based on the bottom, top, and vertical facade.

[0015] In some embodiments of the present invention, the set of two-dimensional line elements includes a first two-dimensional line element and a second two-dimensional line element; By performing a two-dimensional mapping on the regular three-dimensional polygonal surface data, a set of two-dimensional line elements is obtained, including: For the top structural surface, project the top structural surface onto the horizontal plane to obtain the first two-dimensional line element; For a facade structure perpendicular to the horizontal plane, the facade structure is fitted as a two-dimensional line segment located in the horizontal plane to obtain a second two-dimensional line element.

[0016] In some embodiments of the present invention, a two-dimensional spatial decomposition is performed on the set of two-dimensional line elements to obtain multiple candidate surface primitives, including: The minimum bounding box of the building structure line is determined based on the set of two-dimensional line elements, and the minimum bounding box is used as the initial surface primitive. The initial surface primitive is continuously cut using the two-dimensional line elements in the set of two-dimensional line elements until all two-dimensional line elements have been traversed, so as to obtain multiple candidate surface primitives.

[0017] In some embodiments of the present invention, geometric optimization is performed on the contour line structure surface data to obtain regular three-dimensional polygon structure surface data, including: Based on the perimeter and area of ​​the three-dimensional polygonal structure surface in the contour line structure surface data, the shape coefficient of the three-dimensional polygonal structure surface is determined, and based on the area of ​​the three-dimensional polygonal structure surface and the area of ​​the minimum bounding box corresponding to the three-dimensional polygonal structure surface, the regular polygon determination coefficient of the three-dimensional polygonal structure surface is determined. Based on the comparison between the regular polygon determination coefficient and the preset regular threshold, deformed polygons are identified from the three-dimensional polygon structure surface; Shape optimization is performed on deformed polygons to obtain regular 3D polygon structure surface data; Shape optimization includes at least one of the following: structural point reduction, segmentation point set determination, polygon segmentation, and elongated polygon removal.

[0018] Secondly, the present invention also provides a rapid construction system for building 3D models oriented towards lightweight contour line expression, the system comprising: The data acquisition and optimization module is used to acquire lightweight contour line structure surface data and perform geometric optimization on the contour line structure surface data to obtain regular three-dimensional polygon structure surface data. The 2D mapping and decomposition module is used to perform 2D mapping on regular 3D polygonal structure surface data to obtain a set of 2D line elements; and to perform 2D spatial decomposition on the set of 2D line elements to obtain multiple candidate surface primitives. The filtering module is used to construct a graphical cut optimization model to characterize the surface area of ​​a building based on the top and elevation structural surfaces in the regular 3D polygon structural surface data, and to use the graphical cut optimization model to filter effective surface primitives from multiple candidate surface primitives. The 3D construction module is used to map effective surface primitives back to 3D space and determine the corner elevation values ​​of the effective surface primitives according to their type; based on the corner elevation values ​​of the effective surface primitives, 2D to 3D extrusion is performed to obtain the 3D model of the building.

[0019] In the rapid construction method and system for building 3D models oriented towards lightweight contour line expression provided by this invention, the contour line structural surface data of lightweight expression is optimized to obtain regular 3D polygonal structural surface data. The regular 3D polygonal structural surface data is then subjected to 2D mapping and 2D spatial decomposition to obtain multiple candidate surface primitives. In this process, no original point cloud support is required; the 3D spatial decomposition is converted into 2D planar geometric operations, significantly reducing the computational load of spatial modeling and greatly improving model construction efficiency, enabling the rapid construction of large-scale 3D models at the city level. Furthermore, based on the top and elevation structural surfaces in the regular 3D polygonal structural surface data, a graph cut optimization model is constructed to characterize the surface area of ​​the building. This graph cut optimization model is then used to select effective surface primitives from multiple candidate surface primitives, improving the integrity and standardization of the 3D model. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is one of the flowcharts illustrating the rapid construction method for building 3D models oriented towards lightweight contour line expression provided in this embodiment of the invention; Figure 2 This is the second flowchart of the method for rapid construction of 3D building models for lightweight contour line expression provided in this embodiment of the invention; Figure 3 This is a schematic diagram of the polygon optimization process provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of determining the concave point of a polygon's interior corner provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the two-dimensional planar spatial decomposition process provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the sampling grid point distribution provided in an embodiment of the present invention; Figure 7 The top surface structure surround view occlusion rate r provided in the embodiment of the present invention p Distribution heat map; Figure 8 This is a two-dimensional projection schematic diagram of the facade structure line provided in an embodiment of the present invention; Figure 9 The facade structure line constraint surround view occlusion rate r provided in the embodiment of the present invention q Distribution heat map. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0024] "A and / or B" includes the following three combinations: A only, B only, and a combination of A and B.

[0025] The use of "applies to" or "configured to" in this invention implies an open and inclusive language, which does not exclude the applicability to or configuration to devices performing additional tasks or steps. Additionally, the use of "based on" implies openness and inclusivity, because processes, steps, calculations, or other actions "based on" one or more conditions or values ​​may in practice be based on additional conditions or values ​​beyond those conditions.

[0026] In this invention, the term "exemplary" is used to mean "serving as an example, illustration, or description." Any embodiment described as "exemplary" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0027] The following describes, with reference to the accompanying drawings, a method and system for rapid construction of 3D building models for lightweight contour line expression provided by embodiments of the present invention.

[0028] like Figure 1 as well as Figure 2 As shown in the figure, this invention provides a method for rapidly constructing a 3D building model for lightweight contour line expression. The method includes the following steps: S101: Obtain lightweight contour line structure surface data and perform geometric optimization on the contour line structure surface data to obtain regular three-dimensional polygon structure surface data.

[0029] S102 performs two-dimensional mapping on the regular three-dimensional polygonal structure surface data to obtain a set of two-dimensional line elements; and performs two-dimensional spatial decomposition on the set of two-dimensional line elements to obtain multiple candidate surface primitives.

[0030] S103. Based on the top and elevation structural surfaces in the regular three-dimensional polygonal structural surface data, construct a graphical cut optimization model to characterize the surface area of ​​the building, and use the graphical cut optimization model to select effective surface primitives from multiple candidate surface primitives.

[0031] S104 maps the effective surface primitives back to three-dimensional space and determines the corner elevation values ​​of the effective surface primitives according to their type; based on the corner elevation values ​​of the effective surface primitives, a two-dimensional to three-dimensional extrusion is performed to obtain the three-dimensional model of the building.

[0032] The present invention provides a method for rapid construction of 3D building models based on lightweight contour line representation. This method optimizes the contour line structural surface data of the lightweight representation to obtain regular 3D polygonal structural surface data. It then performs 2D mapping and 2D spatial decomposition on the regular 3D polygonal structural surface data to obtain multiple candidate surface primitives. This process does not require original point cloud support, converting 3D spatial decomposition into 2D planar geometric operations, significantly reducing the computational load of spatial modeling and greatly improving model construction efficiency. This method can support the rapid construction of large-scale 3D models at the city level. Furthermore, based on the top and elevation structural surfaces in the regular 3D polygonal structural surface data, a graph cut optimization model is constructed to characterize the building's surface area. This graph cut optimization model is then used to select effective surface primitives from the multiple candidate surface primitives, improving the integrity and standardization of the 3D model.

[0033] In some embodiments of the present invention, step S103 includes the following sub-steps: S1031, using the structural surface normal vector to classify the top surface and the elevation surface of the structural surface.

[0034] It is understandable that in the process of selecting effective surface primitives, two key factors need to be considered: (1) how to fill the structural gaps caused by the absence or discrepancy of structural surfaces and construct a watertight structured three-dimensional model. For example Figure 8As shown, the discontinuity of the structural surface fitting leads to gaps between structures. (2) How to simultaneously consider the structural constraints of the top surface and the facade in the basic element optimization process to improve the accuracy of model reconstruction. Moreover, the structural characteristics of the constraints on the top surface and the facade are different. The top surface structure is mainly used to constrain the shape of the building's outer surface, while the facade structure is mainly used to constrain the actual area of ​​the building model. Therefore, this invention will calculate the influence of the two structures separately.

[0035] Classify structural surfaces into top and elevation surfaces based on their normal vectors: ; in, For structural planes Angle representation of the normal vector.

[0036] Stronger constraints are applied to the facade structure, classifying only structural surfaces that are nearly perpendicular to the horizontal plane as facades to constrain the building's extent. Other structural surfaces are classified as top surfaces to constrain the shape of the building's exterior surface.

[0037] S1032, based on the structural lines of the top surface structure, constructs a top surface shape constraint to characterize the shape of the building's outer surface.

[0038] Understandably, the main problem in selecting top-level primitives lies in... Figure 5 Selection of intermediate segment structures. For segment primitives lacking structural surface support, in modeling methods based on 3D point cloud data, the number of points falling within the segment primitive can be calculated as the support degree of that primitive. However, the computational complexity increases exponentially with the number of point clouds and candidate primitives. The goal of this invention is to construct the most complete building exterior surface shape possible using simple structural surface data. Therefore, a thermal probabilistic map model based on surround view occlusion is proposed to solve the problem of lacking point cloud support.

[0039] Specifically, the target region is discretized into grid points using a preset sampling resolution. The target region is the smallest bounding box containing the set of line elements. Starting from the grid points, rays are drawn along arbitrary angles. The top surface occlusion rate of the grid points is determined based on the number of intersections between the structural lines and rays on the top surface and the total number of rays. The top surface shape constraint is determined based on the top surface occlusion rate of each grid point.

[0040] Indicatively, based on sampling resolution [ , Equal-interval sampling discretizes the initial surface primitive region into grid points. , These are the sampling resolutions in the x and y directions, respectively.

[0041] Set the line element collection Minimum bounding box Box ( Let ) be the target region, and let the length of the target region in the x-direction be . Lx The length in the y direction is Ly .

[0042] The coordinates of the sampling point in the x-direction are: ; The coordinates of the sampling point in the y-direction are: ; in, for Box ( The minimum value on the x-axis. for Box ( The minimum value on the y-axis. For example... Figure 6 As shown, the target area is sampled into 100×100 grid points P=[ ].

[0043] Depend on Figure 6 It can be observed that for grid points located inside the structural plane, their spatial distribution exhibits a topological feature of being completely surrounded by structural lines; for grid points in the interval region, structural lines create a large-scale visual obstruction during 360° panoramic observation; while for observation points in the outer region of the building, only local structural line obstruction exists within their field of view.

[0044] Based on this, this embodiment of the invention designs a probability index for the internal and external of grid points based on occlusion of the surrounding view. Specifically, taking the grid point as the starting point, around the 360° surrounding view ray, the proportion of the ray that is occluded by any structural line is calculated as the probability value of the grid point belonging to the internal point.

[0045] From grid points ( , Draw rays at any angle. The resulting ray equation can be expressed as:

[0046] ; in, >0 indicates that the ray originates from the starting point ( , )along Extending infinitely in direction For any angle, The set total number of rays, This indicates a ray that traverses from 0° to 360° at equal intervals, where kt is the proportional interval, used to determine a specific emission angle. The set of structure lines for traversing the top structure surface , among which, line segment Represented by the coordinates of its endpoints. The equation of the line is expressed as: ; Calculate the structural lines sequentially With current rays Does an intersection exist? , And satisfy the following formula: ; If the above formula is satisfied, then it is determined that the ray is occluded by a structure line, the traversal of the structure line set is stopped, and the occlusion count of the current grid point is increased. n = n +1, proceed to the next ray traversal; if the current ray has no intersection with any structure line, increment the occlusion count. n Maintain the status quo. Until all directions are unchanged. After traversing the entire grid, calculate the top surface occlusion rate of each grid point based on the number of intersections between the structure lines and rays on the top surface and the total number of rays. : ; In the formula, It is the number of intersections between the structural lines and rays of the top structural surface, and N is the total number of rays.

[0047] To accelerate the calculation process, this embodiment of the invention assigns values ​​to grid points falling inside the structural surface. Since all the surrounding rays of grid points falling inside the structural surface will inevitably be occluded by the structural lines, values ​​are directly assigned to grid points inside the structural surface. Secondly, the remaining grid points are sequentially connected to the structure lines. Perform an intersection check. If the currently traversed ray encounters an intersecting structure line, immediately determine that the ray is an intersecting ray and terminate the intersection check with subsequent structure lines, then directly traverse the next grid point ray.

[0048] like Figure 7 As shown, its colors range from blue to red. From 0 to 1, red represents the mesh points inside the structural surface. =1; For spaced structures, since their grid points are located between structures, most of the panoramic view is obstructed. A larger value indicates a higher probability that the point belongs to the building structure; for points outside the structural area, A smaller value indicates a higher probability that it does not belong to the surface area of ​​the building model.

[0049] S1033, based on the structural lines of the facade structure, construct facade edge constraints to characterize the actual area of ​​the building.

[0050] It is understandable that the facade structure corresponds to the wall structure that restricts the scope of the building; therefore, the facade structure is used as the boundary constraint of the building's 3D model. For example... Figure 8 As shown, in this embodiment of the invention, structural lines are constructed based on the projection of the facade structural surface onto a two-dimensional plane, and these line segment information are used to apply strong constraints to them.

[0051] Specifically, consistent with step S1032, the facade occlusion rate of the grid points is determined based on the number of intersections between the structural lines and rays of the facade structural surface and the total number of rays. = , It is the number of intersections between structural lines and rays on the facade; based on the facade occlusion rate of each grid point, the facade edge constraints are determined.

[0052] Indicatively, for Figure 8 The occlusion rate of the structural lines shown The distribution heat map is as follows Figure 9 As shown, the closer the grid point color is to red, the stronger the color. The larger the value, the closer the color is to blue, and the smaller the value. It can be observed that the facade structure line only constrains the edges of the grid points within the building's minimum bounding box. Furthermore, for the complete and defined portions of the structure line, the occlusion rate of the grid points is larger, while for uncertain areas with missing structures, the value tends to decrease as it moves away from the structure line. This is used to measure the edge certainty of the building's location.

[0053] S1034, weighted fusion of top surface shape constraints and facade edge constraints to obtain fused constraints.

[0054] Schematic representation: for each grid point P=[ Top surface occlusion rate With facade occupancy rate Construct grid point weighted probability values (i.e., fusion constraints) , is a weighting coefficient used to weigh the weights of shape and edge features applied to grid points.

[0055] S1035 constructs a graph cut optimization model based on fusion constraints, and selects effective surface primitives from multiple candidate surface primitives by minimizing the energy function of the graph cut optimization model.

[0056] Specifically, the data terms of the graph cut optimization model are constructed based on the mean of each grid point within the surface primitive, and the smoothing terms of the graph cut optimization model are constructed based on the difference between the data terms of adjacent surface primitives; based on the data terms and the smoothing terms, the energy function is determined, the minimum energy function is calculated, and effective surface primitives are selected.

[0057] Schematic, using fusion constraints Obtain the probability distribution of the building's structural surfaces within the minimum bounding box region, and then fuse the constraints. Transform into surface-based metadata item D(B) i ): ; in, Represents grid points For falling on the surface primitive The point in the middle; Let be a counting function, for each primitive falling into the surface. The points in the middle are cumulatively increased by 1.

[0058] The smoothing term in the graph cut optimization model measures the smoothing cost between nodes. By penalizing adjacent nodes with different labels, it smooths the labels between adjacent face primitives, ensuring that face primitives belonging to the same structure have the same category label. Therefore, the average value of grid points in two adjacent face primitives is used as the smoothing term. The difference is used as the smoothing term S(B) i B j ): .

[0059] Based on the smoothing term, the optimal constraint formula of this invention is constructed as the criterion for selecting the optimal primitive: ; in, Node configuration representation ,when When, it indicates that the surface element is an effective surface structure inside the building. This indicates that the surface primitive is outside the structural region of the building and is therefore an invalid primitive. Solving the minimum energy function and selecting the optimal primitives are the valid surface primitives that appear as patches of the three-dimensional model of the building in the two-dimensional plane.

[0060] The method for rapid construction of 3D building models for lightweight contour line expression provided by this invention incorporates both the top and facade structures into the modeling constraint system, effectively solving the problem of low modeling accuracy caused by insufficient structural surface constraints in traditional methods. At the same time, through spatial decomposition and graph cut algorithm optimization, it solves the problem of model structural logic errors caused by incomplete structural surface extraction, and can construct a 3D model that conforms to the actual structure of the building.

[0061] In some embodiments of the present invention, step S104 includes the following sub-steps: S1041 divides the effective surface primitives into intersection surface primitives and non-intersection surface primitives.

[0062] S1042, For the intersection surface primitive, determine the corner elevation value of the intersection surface primitive based on the three-dimensional structural surface corresponding to the intersection surface primitive.

[0063] Schematic, since the intersection surface primitives have corresponding structural surfaces in three-dimensional space, according to the principle of spatial cutting methods, they are completely contained within the projection range of the structural surfaces. Therefore, the three-dimensional structural surface corresponding to the current surface primitive is taken as its mapping target, and its corresponding three-dimensional structural surface equation is assumed to be: .

[0064] Using surface primitives respectively Each corner point Draw a straight line along the perpendicular direction to find the intersection point. First, expand the two-dimensional corner point into a three-dimensional axial coordinate system. The parametric equation for the perpendicular line is:

[0065] Substituting these values ​​into the three-dimensional structural surface equations above, we can solve for the unknown parameters. : ; The final intersection point coordinates are: ; The surface elements are obtained sequentially through the above process. The three-dimensional intersection points corresponding to the corner points are then used to construct the three-dimensional structural surface primitives in the original order. ,in .

[0066] Since buildings have multi-layered structures, a surface element may be contained by multiple structural surfaces. In this embodiment of the invention, the outermost structural surface is used as the supporting surface of the building's outer surface model, that is, the highest-level structural surface is used as the mapping target of the surface element.

[0067] S1043, For non-intersecting surface primitives, determine the corner elevation values ​​of the non-intersecting surface primitives based on the surface primitives with determined elevation values ​​in their neighborhood.

[0068] Schematic, for non-intersecting surface primitives, their corresponding regions in three-dimensional space are the gap regions between structural surfaces. Since the exact mapping surface cannot be determined, this embodiment of the invention uses the structural surface with a determined height within its neighborhood as its corresponding height.

[0069] The surface primitives generated by spatial decomposition are constructed in a compact manner, with each internal corner (non-edge corner) contained by at least two surface primitives. Therefore, for the same corner, there may be more than two elevations. Value, for non-intersecting surface primitives Iterate through all corner points and determine whether these corner points are contained within a given height surface primitive. If a corner point already has an assigned value, use that value as the elevation value of the current corner point. If no corner point is assigned a value, it is recorded as 0. After traversal, the set of elevation values ​​for non-intersecting surface primitives is: .

[0070] Since the floor height of the gap region cannot be accurately obtained, in order to avoid generating redundant structures, this embodiment of the invention uses the minimum elevation value of the non-intersecting surface primitives with determined heights in their neighborhood as the height value of all corner points of the non-intersecting surface primitive, i.e., the non-intersecting surface primitive. The elevation values ​​of all corner points are represented as follows: ; The optimal elevation value of the non-intersecting surface primitive is determined through multiple iterations.

[0071] S1044: Based on the corner elevation values ​​of the intersection surface primitives and the corner elevation values ​​of the non-intersection surface primitives, the effective surface primitives in the two-dimensional plane are mapped to three-dimensional structural surface primitives.

[0072] Schematic, the polygonal structural surface is positioned and stretched according to the height of the corner points of the structural surface primitive to form a three-dimensional structural frame.

[0073] S1045, based on the connection relationship between the three-dimensional structural surface primitives and the corresponding two-dimensional surface primitives, a three-dimensional model of the building is constructed.

[0074] It is understandable that since the three-dimensional structural surface primitives are vertically mapped from the two-dimensional surface primitives, the corner points of the three-dimensional structural surface primitives are connected to the corner points of their corresponding two-dimensional surface primitives to construct a vertical facade. That is, a three-dimensional structural frame is constructed with the two-dimensional surface primitives as the bottom base and the corresponding three-dimensional structural surface primitives as the top base, thereby obtaining the three-dimensional model of the building.

[0075] In some embodiments of the present invention, a three-dimensional model of a building is constructed based on the connection relationship between three-dimensional structural surface primitives and corresponding two-dimensional surface primitives, including: Construct a minimum bounding box based on the initial structural surface, and determine the elevation of the lowest point of the minimum bounding box as the ground elevation value; By using the ground elevation value as the corner points of the two-dimensional surface element corresponding to the effective surface element, we obtain the two-dimensional surface element located at the ground height. Using a two-dimensional surface primitive at ground level as the lower base and a three-dimensional structural surface primitive as the upper base, connect the corresponding corner points of the lower base and the upper base to construct the vertical surface corresponding to the three-dimensional structural surface primitive. A three-dimensional model of the building is constructed based on the bottom, top, and vertical facade.

[0076] Schematic representation, due to the corner points of the two-dimensional surface primitives on the bottom surface Since the elevation cannot be determined, this embodiment of the invention uses the initial structural plane. Construct the minimum bounding box and obtain the elevation of the lowest point of the minimum bounding box. All elevation values ​​that serve as the basic elements of the lower two-dimensional surface, i.e. .

[0077] By assigning elevation values ​​to the two-dimensional surface primitives based on the ground height (or using the lowest point in the structural surface as the ground height), the upper and lower bases of the polygonal structural frame are determined. By connecting the corresponding corner points of the three-dimensional structural surface primitives and the two-dimensional surface primitives, a three-dimensional facade is constructed, thus realizing the construction of a 2D-3D three-dimensional model.

[0078] In some embodiments of the present invention, the set of two-dimensional line elements includes a first two-dimensional line element and a second two-dimensional line element; the two-dimensional line element set is obtained by performing two-dimensional mapping on regular three-dimensional polygonal structure surface data, including: For the top structural surface, project the top structural surface onto the horizontal plane to obtain the first two-dimensional line element; For a facade structure perpendicular to the horizontal plane, the facade structure is fitted as a two-dimensional line segment located in the horizontal plane to obtain a second two-dimensional line element.

[0079] It is understandable that in existing methods for reconstructing 3D building models based on roof surface data, the building's roof structure is directly used for 2D reconstruction. This approach ignores the constraints of the building's facade structure. In real building scenarios, the roof structure usually extends outward beyond the facade structure by a certain distance, and the actual area occupied by the building is based on the facade walls. Therefore, in order to reconstruct the 3D building model more accurately, this embodiment of the invention also uses the building facade structure as a constraint basis, and the facade structure lines are also used as spatial decomposition elements.

[0080] Specifically, for the top structural surface, it is directly projected onto the horizontal plane to obtain two-dimensional line elements; for the facade structure perpendicular to the horizontal plane, it is fitted as a whole into a two-dimensional line segment, with its endpoints being a pair of farthest points in the set of corner points of the structural surface. This process transforms the three-dimensional structural surface data... Mapped to two-dimensional line data, the first two-dimensional line element. Second two-dimensional line elements .

[0081] In some embodiments of the present invention, a two-dimensional spatial decomposition is performed on the set of two-dimensional line elements to obtain multiple candidate surface primitives, including: The minimum bounding box of the building structure line is determined based on the set of two-dimensional line elements, and the minimum bounding box is used as the initial surface primitive. The initial surface primitive is continuously cut using the two-dimensional line elements in the set of two-dimensional line elements until all two-dimensional line elements have been traversed, so as to obtain multiple candidate surface primitives.

[0082] Indicatively, such as Figure 5 As shown, using the two-dimensional line elements in the two-dimensional line element set First, the minimum bounding box B1 is cut to generate two polygonal face primitives B2 and B3, and at the same time, new corner points v3 and v4 and line elements are generated. The generated B2 and B3 are candidate surface primitives that constitute the structural surfaces of the building. Next, new two-dimensional line elements are added to the set of two-dimensional line elements. The cutting process continues, dividing the polygonal region B2 into face primitives B4 and B5, generating new corner points v7 and v8, and line primitives. Repeat this segmentation operation until all two-dimensional line elements in the set of two-dimensional line elements have been spatially segmented. Understandably, as new segmentation line elements are added, the boundary constraints applied between each structural surface increase, and the segmented surface primitives become increasingly closer to the distribution of the actual building surface.

[0083] During spatial segmentation, the minimum bounding box of the building's structural lines is used as the initial surface primitive. The creation of this initial surface primitive limits the maximum range of the reconstructed building model and provides the initial segmentation surface. Following the above cutting operation, polygons are sequentially cut using 2D line elements to generate a mesh model, where the polygons are surface primitives obtained by cutting with line elements. After segmenting the initial surface primitive using all 2D line elements, a large number of polygonal surface primitives are generated as candidate surface primitives, represented as follows: ; ; in, Represents the candidate face primitive. This represents the corner points that make up the basic elements of the candidate surface.

[0084] Understandably, in constructing a 3D model using point clouds as input data, the quantity or area ratio of point clouds contained in the candidate surface primitives needs to be used as a metric for data items and smoothing items, and its computational complexity will be... level, n For the number of point clouds, mThe number of candidate primitives. Compared to point cloud data, the data used in this invention is lightweight building structural surface outline data.

[0085] In some embodiments of the present invention, step S101 includes the following sub-steps: S1011. Based on the perimeter and area of ​​the three-dimensional polygonal structure surface in the contour line structure surface data, determine the shape coefficient of the three-dimensional polygonal structure surface, and based on the area of ​​the three-dimensional polygonal structure surface and the area of ​​the minimum bounding box corresponding to the three-dimensional polygonal structure surface, determine the regular polygon determination coefficient of the three-dimensional polygonal structure surface.

[0086] Understandably, for structural surfaces that adhere to the Manhattan principle, their shapes primarily conform to convex polygons, and their structures are all compact polygons. That is, if a standard circular object is considered neither compact nor bulging, then its shape factor can be defined. r : ; In the formula, Let be the perimeter of the three-dimensional polygonal structure surface. S Let be the area of ​​the surface of the three-dimensional polygon structure. When When <1, the polygon is compact; when When = 1, the polygon is a standard circle; when When the value is greater than 1, the polygon is an bloated type.

[0087] Based on the shape factor r, a regular polygon determination factor k is designed, and its area ratio is calculated using the minimum bounding box of the structural polygon as the benchmark. ; In the formula, This represents the minimum bounding box area corresponding to the three-dimensional polygon structure surface.

[0088] Regular polygon determination coefficient Used to measure the volume ratio of the area occupied by a polygon, compared to the shape factor. r Decision coefficients for regular polygons It can more directly measure whether there are unusual shapes causing the minimum bounding box of a polygon to occupy an abnormally large area. This can be achieved by setting rule thresholds. To determine whether the current polygon is a regular polygon, when... Represented as a regular polygon. It is represented as a deformed polygon.

[0089] S1012, based on the comparison result between the regular polygon determination coefficient and the preset regular threshold, identify deformed polygons from the 3D polygon structure surface. It is understandable that deformed polygons are caused by the presence of redundant points and faces.

[0090] S1013 optimizes the shape of the deformed polygon to obtain regular three-dimensional polygon structure surface data.

[0091] Shape optimization includes at least one of the following: structural point reduction, segmentation point set determination, polygon segmentation, and elongated polygon removal.

[0092] In some examples, shape optimization is performed on malformed polygons, such as... Figure 3 As shown, structural point reduction for deformed polygons includes: traversing all corner points, deleting the current corner point v from polygon S, and determining whether the current point is a valid point by the change in the regular polygon determination coefficient k. If k increases after deleting the point, it means that the point is a valid structural support point; if k decreases, it means that the point is an invalid support point. The existence of the point will increase the proportion of the polygon shape space, but the actual effective area of ​​the polygon will not increase significantly. Therefore, it is determined to be an invalid support point and is removed.

[0093] Determining the structural point set for malformed polygons includes: after traversing all structural points, if the determination coefficient k of the regular polygon is still greater than the threshold... If the polygon has redundant structures, it indicates that the polygon is concave. Such polygons are typically concave and can be divided into regular polygons through polygon cutting. The key is to find suitable cutting points to optimize the concave polygon into a convex polygon. For a concave polygon, the key point is the "concave" point. Cutting along the edge direction traversed by the "concave" point can divide a concave polygon into two convex polygons. Therefore, the problem becomes how to locate the "concave" point of the concave polygon. This invention determines the "concave" point of the polygon based on the values ​​of its interior angles, such as... Figure 4 As shown, the normal interior angles of a polygon are all less than 180°, while the interior angles of potential "concave" points are all greater than 180°. Therefore, it is necessary to traverse all interior angles of the polygon. ,Pick Corner points v with an angle greater than 180° are considered as candidate "concave" points and added to the cutting point set Q.

[0094] Polygon segmentation for deformed polygons includes: traversing all cut point sets Q, segmenting the polygon along the edges where the cut points are located, generating two polygons Poly1 and Poly2, calculating the regular polygon determination coefficient k for each polygon, and if both are within the threshold T. k Within this range, it indicates that the two polygons are regular polygons. If... > This indicates that the point is not a "concave" point, and it should be removed from Q. Finally, if multiple cutting points satisfy... < The cutting point v corresponding to the minimum volume ratio k of each group of segmented polygons is selected as the "concave" point to segment the polygons.

[0095] Finally, to prevent the formation of long, narrow polygons, the two segmented polygons are judged. If the shape coefficient data r is greater than a set threshold... Remove long, narrow polygons, among which Threshold for line-shaped polygons. It is usually taken as 1.3.

[0096] The method for rapid construction of 3D building models for lightweight contour line expression provided in this invention ensures the regularity of basic modeling data and improves the standardization of 3D models by identifying and optimizing the shape of deformed polygons.

[0097] Secondly, embodiments of the present invention also provide a rapid construction system for building 3D models oriented towards lightweight contour line expression, the system comprising: The data acquisition and optimization module is used to acquire lightweight contour line structure surface data and perform geometric optimization on the contour line structure surface data to obtain regular three-dimensional polygon structure surface data. The 2D mapping and decomposition module is used to perform 2D mapping on regular 3D polygonal structure surface data to obtain a set of 2D line elements; and to perform 2D spatial decomposition on the set of 2D line elements to obtain multiple candidate surface primitives. The filtering module is used to construct a graphical cut optimization model to characterize the surface area of ​​a building based on the top and elevation structural surfaces in the regular 3D polygon structural surface data, and to use the graphical cut optimization model to filter effective surface primitives from multiple candidate surface primitives. The 3D construction module is used to map effective surface primitives back to 3D space and determine the corner elevation values ​​of the effective surface primitives according to their type; based on the corner elevation values ​​of the effective surface primitives, 2D to 3D extrusion is performed to obtain the 3D model of the building.

[0098] The rapid construction system for building 3D models oriented towards lightweight contour line expression provided in this embodiment of the invention corresponds to the rapid construction method for building 3D models oriented towards lightweight contour line expression provided in any of the above embodiments, and will not be described again here.

[0099] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0100] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0101] The foregoing has provided a detailed description of a method and system for rapid construction of 3D building models for lightweight contour line expression, as provided in the embodiments of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A building three-dimensional model fast construction method oriented to lightweight profile line expression, characterized in that, The method includes: Obtain lightweight contour line structure surface data, and perform geometric optimization on the contour line structure surface data to obtain regular three-dimensional polygon structure surface data; The three-dimensional polygonal structure surface data of the rule is mapped in two dimensions to obtain a set of two-dimensional line elements; and the set of two-dimensional line elements is decomposed in two dimensions to obtain multiple candidate surface primitives. Based on the top and elevation structural surfaces in the regular three-dimensional polygonal structural surface data, a graphical cut optimization model is constructed to characterize the surface area of ​​the building, and the graphical cut optimization model is used to select effective surface primitives from the multiple candidate surface primitives; The effective surface primitives are mapped back to three-dimensional space, and the corner elevation values ​​of the effective surface primitives are determined according to their type. Based on the corner elevation values ​​of the effective surface primitives, two-dimensional to three-dimensional extrusion is performed to obtain a three-dimensional model of the building.

2. The building 3D model rapid construction method according to claim 1, wherein, The step involves constructing a graphical cut optimization model to characterize the surface region of a building based on the top and elevation structural surfaces in the regular three-dimensional polygonal structural surface data, and using the graphical cut optimization model to select effective surface primitives from the plurality of candidate surface primitives, including: Structural surfaces are classified into top and elevation surfaces using structural surface normal vectors; Based on the structural lines of the top surface structure, construct top surface shape constraints to characterize the shape of the building's outer surface; Based on the structural lines of the facade structure, facade edge constraints are constructed to characterize the actual area of ​​the building; The top surface shape constraint and the facade edge constraint are weighted and fused to obtain a fused constraint. The graph cut optimization model is constructed based on the fusion constraints, and the effective surface primitives are selected from the plurality of candidate surface primitives by minimizing the energy function of the graph cut optimization model.

3. The building 3D model rapid construction method according to the lightweight profile line expression of claim 2, characterized in that, The structural lines based on the top surface structure plane construct top surface shape constraints to characterize the shape of the building's outer surface, including: The target region is discretized into grid points using a preset sampling resolution, where the target region is the smallest bounding box containing the set of line elements. Starting from the grid points, rays are drawn along arbitrary angles and directions; The top surface occlusion rate of the grid points is determined based on the number of intersections between the structural lines of the top surface and the rays, as well as the total number of rays. The top surface shape constraint is determined based on the top surface occlusion rate of each grid point; The structural lines based on the facade structural surfaces construct facade edge constraints to characterize the actual area of ​​the building, including: The facade occlusion rate of the grid points is determined based on the number of intersections between the structural lines of the facade and the rays, as well as the total number of rays. The facade edge constraints are determined based on the facade occlusion rate of each grid point.

4. The method of claim 2, wherein the method is characterized by, The process of constructing the graph cut optimization model based on the fusion constraints and minimizing the energy function of the graph cut optimization model includes: The data terms of the graph cut optimization model are constructed based on the mean of each grid point within the surface primitive, and the smoothing terms of the graph cut optimization model are constructed based on the difference between the data terms of adjacent surface primitives. Based on the data items and the smoothing items, the energy function is determined, the minimum energy function is calculated, and the effective surface primitives are selected.

5. The method of claim 1, wherein, The effective surface primitive is mapped back to three-dimensional space, and the corner elevation value of the effective surface primitive is determined according to the type of the effective surface primitive; Based on the corner elevation values ​​of the effective surface primitives, a two-dimensional to three-dimensional extrusion is performed to obtain a three-dimensional model of the building, including: The effective surface primitives are divided into intersection surface primitives and non-intersection surface primitives; For the intersection surface primitive, the corner elevation values ​​of the intersection surface primitive are determined according to the three-dimensional structural surface corresponding to the intersection surface primitive; For the non-intersecting surface primitive, the corner elevation value of the non-intersecting surface primitive is determined based on the surface primitives with determined elevation values ​​in the neighborhood of the non-intersecting surface primitive. Based on the corner elevation values ​​of the intersection surface primitives and the corner elevation values ​​of the non-intersection surface primitives, the effective surface primitives in the two-dimensional plane are mapped to three-dimensional structural surface primitives. Based on the connection relationship between the three-dimensional structural surface primitives and the corresponding two-dimensional surface primitives, the three-dimensional model of the building is constructed.

6. The building 3D model rapid construction method according to claim 5, wherein, The construction of the three-dimensional building model based on the connection relationship between the three-dimensional structural surface primitives and the corresponding two-dimensional surface primitives includes: Construct a minimum bounding box based on the initial structural surface, and determine the elevation of the lowest point of the minimum bounding box as the ground elevation value; Using the ground elevation value as the corner points of the two-dimensional surface element corresponding to the effective surface element, a two-dimensional surface element located at the ground height is obtained; Using the two-dimensional surface primitive at ground level as the lower base and the three-dimensional structural surface primitive as the upper base, connect the corresponding corner points of the lower base and the upper base to construct the vertical surface corresponding to the three-dimensional structural surface primitive. Based on the lower base, the upper base, and the vertical facade, a three-dimensional model of the building is constructed.

7. The method of claim 1, wherein the method is a method of building a three-dimensional model of a building quickly by using a lightweight profile line expression. The set of two-dimensional line elements includes first two-dimensional line elements and second two-dimensional line elements; The step of performing a two-dimensional mapping on the three-dimensional polygonal structure surface data of the rule to obtain a set of two-dimensional line elements includes: For the top structural surface, project the top structural surface onto the horizontal plane to obtain the first two-dimensional line element; For a facade structure perpendicular to the horizontal plane, the facade structure is fitted into a two-dimensional line segment located in the horizontal plane to obtain the second two-dimensional line element.

8. The building 3D model rapid construction method according to the lightweight profile line expression of claim 1, characterized in that, The step of performing two-dimensional spatial decomposition on the set of two-dimensional line elements to obtain multiple candidate surface primitives includes: The minimum bounding box of the building structure line is determined based on the set of two-dimensional line elements, and the minimum bounding box is used as the initial surface primitive. The initial surface primitive is continuously cut using the two-dimensional line elements in the set of two-dimensional line elements until all two-dimensional line elements have been traversed, so as to obtain multiple candidate surface primitives.

9. The building 3D model rapid construction method according to any one of claims 1 to 8, wherein, The step of geometrically optimizing the contour line structure surface data to obtain regular three-dimensional polygon structure surface data includes: Based on the perimeter and area of ​​the three-dimensional polygonal structure in the contour line structure data, the shape coefficient of the three-dimensional polygonal structure is determined, and based on the area of ​​the three-dimensional polygonal structure and the area of ​​the minimum bounding box corresponding to the three-dimensional polygonal structure, the regular polygon determination coefficient of the three-dimensional polygonal structure is determined. Based on the comparison result between the regular polygon determination coefficient and the preset rule threshold, deformed polygons are identified from the three-dimensional polygon structure surface; The deformed polygon is shaped and optimized to obtain the regular three-dimensional polygon structure surface data; The shape optimization includes at least one of the following: structural point reduction, segmentation point set determination, polygon segmentation, and elongated polygon removal.

10. A rapid construction system for 3D building models oriented towards lightweight contour line expression, characterized in that, The system includes: The data acquisition and optimization module is used to acquire lightweight contour line structure surface data and perform geometric optimization on the contour line structure surface data to obtain regular three-dimensional polygon structure surface data. The two-dimensional mapping and decomposition module is used to perform two-dimensional mapping on the three-dimensional polygonal structure surface data of the rule to obtain a set of two-dimensional line elements; and to perform two-dimensional spatial decomposition on the set of two-dimensional line elements to obtain multiple candidate surface primitives. The filtering module is used to construct a graphical cut optimization model to characterize the surface area of ​​a building based on the top surface and the facade surface in the regular three-dimensional polygonal structural surface data, and to use the graphical cut optimization model to filter effective surface primitives from the multiple candidate surface primitives. A 3D construction module is used to map the effective surface primitives back to 3D space and determine the corner elevation values ​​of the effective surface primitives according to their type; and to perform 2D to 3D extrusion based on the corner elevation values ​​of the effective surface primitives to obtain a 3D model of the building.

Images