Rapid construction method of building facade model based on BIM technology

Abstract

The application relates to the technical field of three-dimensional modeling, in particular to a building outer surface model rapid construction method based on BIM technology, which is used to solve the technical problem of insufficient model precision in BIM modeling in the prior art. The method comprises the following steps: constructing a first triangular mesh model based on building outer surface data; calculating a first folding cost of each edge of each triangle in the first triangular mesh model based on a quadratic error measurement method; correcting the first folding cost of each edge based on a first inhibition factor of each edge and / or a second inhibition factor of each edge to obtain a second folding cost of each edge; the first inhibition factor is used to inhibit the folding cost of a high-detail-density area, and the second inhibition factor is used to inhibit the folding cost of a high-shape-change area; and performing simplification processing on the edges in the first triangular mesh model based on the second folding cost of each edge to obtain a second triangular mesh model.

Description

Technical Field

[0001] This invention relates to the field of 3D modeling technology, and more specifically to a method for rapidly constructing building exterior surface models based on BIM technology. Background Technology

[0002] Currently, industries such as architectural design and scheme creation, construction and project management, urban planning and digital twins, and historical building restoration all require high-precision images of building exterior surfaces. Building Information Modeling (BIM) technology can be used to construct building exterior surfaces, offering advantages such as integrating various data sources, achieving 3D visualization, prefabrication, and precision control. However, the sheer volume of data involved in BIM technology leads to low modeling efficiency and hinders the rapid acquisition of high-precision models of building exterior surfaces. To improve modeling efficiency, BIM models need to be simplified. For example, edge folding algorithms based on quadratic error metrics can be used to simplify BIM models, reducing their data volume. These algorithms calculate the geometric distance error caused by edge folding as the folding cost and fold the edges with the lowest cost sequentially to achieve model simplification.

[0003] However, the edge folding algorithm with secondary error measurement only uses a single geometric distance error as the criterion for measuring the folding cost during the simplification process. The criterion is relatively simple and cannot accurately reflect the details of the building's outer surface, resulting in insufficient accuracy of the final BIM model. Summary of the Invention

[0004] To address the technical problem of insufficient model accuracy in existing BIM modeling technologies, the present invention aims to provide a method for rapidly constructing building exterior surface models based on BIM technology. The specific technical solution adopted is as follows:

[0005] A first triangular mesh model is constructed based on the building's external surface data. The first triangular mesh model is a model composed of multiple triangles spliced ​​together to represent the shape of the building's external surface.

[0006] The first folding cost of each edge of each triangle in the first triangular mesh model is calculated based on the quadratic error metric method.

[0007] Based on the first suppression factor and / or the second suppression factor of each edge, the first folding cost of each edge is modified to obtain the second folding cost of each edge; wherein, the first suppression factor is used to suppress the folding cost of high detail density regions, and the second suppression factor is used to suppress the folding cost of high shape change regions.

[0008] The edges in the first triangular mesh model are simplified based on the second folding cost of each edge to obtain the second triangular mesh model.

[0009] As one possible implementation, the method further includes: calculating the average area of ​​the adjacent triangles at each of the two endpoints of the target edge; the adjacent triangle refers to a triangle with one of the two endpoints of the target edge as its vertex; the target edge is any side in the first triangular mesh model; calculating the average of the average areas of the adjacent triangles at each endpoint; and calculating a first suppression factor of the target edge based on the average of the average areas, wherein the first suppression factor of the target edge is negatively correlated with the average area of ​​the adjacent triangles.

[0010] As one possible implementation, the method further includes: determining a first inhibition factor adjustment hyperparameter; calculating the ratio of the first inhibition factor adjustment hyperparameter to the average value of the average area; and using the sum of the ratio and a first preset value as the first inhibition factor.

[0011] As one possible implementation, the method further includes: calculating the curvature feature values ​​of the two endpoints of the target edge, the curvature feature values ​​being used to characterize the average degree of change of the surface at the location of the endpoints; the target edge being any edge in the first triangular mesh model; and calculating a second suppression factor of the target edge based on the maximum value of the curvature feature values ​​of the two endpoints, the second suppression factor being positively correlated with the maximum value.

[0012] As one possible implementation, the method further includes: calculating the unit normal vector of each adjacent triangle at the endpoint; multiplying the unit normal vector of each adjacent triangle by its area to obtain the weighted normal vector of each adjacent triangle; summing all the weighted normal vectors and normalizing the summation result to obtain the average weighted normal vector of the endpoint; calculating the spatial angle between the unit normal vector and the average weighted normal vector of each adjacent triangle; and calculating the average value of the spatial angle as the curvature feature value.

[0013] As one possible implementation, the method further includes: subtracting a second preset value from the sum of the first inhibition factor and the second inhibition factor of the target edge to obtain a combined inhibition factor; and using the product of the combined inhibition factor and the first folding cost as the second folding cost of the target edge.

[0014] As one possible implementation, the method further includes: Step 1, sorting the edges in ascending order of their second folding costs; Step 2, selecting the edge with the smallest second folding cost in the current sort and folding it, and updating the connection relationship between the remaining edges and vertices to obtain a new triangular mesh model; Step 3, recalculating the second folding cost of each edge in the new triangular mesh model; iteratively executing steps 1 to 3 until a preset number of iterations is reached, or the simplification error of the new triangular mesh model is less than a first threshold, and outputting the current triangular mesh model as the second triangular mesh model.

[0015] As one possible implementation, the method further includes: deleting the currently required folding edge, the two endpoints of the folding edge, and the two triangles with the folding edge as a common edge; calculating the positions of the two vertices of the folding edge based on a preset vertex generation algorithm to generate new vertices; and updating the topological connection relationship of the triangular mesh model based on the new vertices and the other two vertices of the two triangles to form new triangles.

[0016] As one possible implementation, the method further includes: acquiring initial three-dimensional point cloud data of the building's exterior surface obtained by scanning the building's exterior surface using a lidar device; performing outlier filtering on the initial three-dimensional point cloud data to remove noise, thereby obtaining processed three-dimensional point cloud data; converting the processed three-dimensional point cloud data into a triangular mesh based on a Poisson surface reconstruction algorithm; and performing geometric processing, semantic mapping, and parametric modeling on the triangular mesh to obtain a first triangular mesh model.

[0017] As one possible implementation, the method further includes: setting a distance threshold based on the neighborhood distance between each point cloud data in the initial three-dimensional point cloud data; deleting target points in the initial three-dimensional point cloud data to obtain processed three-dimensional point cloud data, wherein the target point is a point whose number of point cloud data within a radius of the distance threshold centered on the target point is less than a second threshold.

[0018] This invention offers the following advantages: By introducing a first inhibition factor and a second inhibition factor to collaboratively correct edge folding costs, this application constructs a hierarchical model simplification mechanism. Compared to the limitations of traditional simplification methods that rely solely on a single geometric error, this application achieves differentiated processing of different feature regions of the building's exterior surface model. By simultaneously incorporating detail density and shape curvature into the folding cost evaluation system, the edge folding priority in high-detail and high-curvature regions is effectively reduced. This significantly reduces the amount of model data while substantially improving the fidelity of the simplified model to the complex features of the building facade, thereby enhancing the accuracy of BIM modeling. Attached Figure Description

[0019] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0020] Figure 1 A flowchart illustrating a method for rapid construction of building exterior surface models based on BIM technology, provided as an embodiment of the present invention. Figure 1 ;

[0021] Figure 2 A flowchart illustrating a method for rapid construction of building exterior surface models based on BIM technology, provided as an embodiment of the present invention. Figure 2 ;

[0022] Figure 3 A flowchart illustrating a method for rapid construction of building exterior surface models based on BIM technology, provided as an embodiment of the present invention. Figure 3 ;

[0023] Figure 4 A flowchart illustrating a method for rapid construction of building exterior surface models based on BIM technology, provided as an embodiment of the present invention. Figure 4 ;

[0024] Figure 5 A flowchart illustrating a method for rapid construction of building exterior surface models based on BIM technology, provided as an embodiment of the present invention. Figure 5 ;

[0025] Figure 6 A flowchart illustrating a method for rapid construction of building exterior surface models based on BIM technology, provided as an embodiment of the present invention. Figure 6 . Detailed Implementation

[0026] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method for rapid construction of building exterior surface models based on BIM technology proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0028] The following description, in conjunction with the accompanying drawings, details the specific scheme of a method for rapidly constructing a building exterior surface model based on BIM technology provided by this invention.

[0029] Please see Figure 1 The diagram illustrates a method flowchart for rapidly constructing a building exterior surface model based on BIM technology, according to an embodiment of the present invention. The method includes the following steps:

[0030] Step 101: Construct the first triangular mesh model based on the building's external surface data.

[0031] The first triangular mesh model is a model composed of multiple triangles spliced ​​together to represent the shape of the building's outer surface.

[0032] In one possible implementation, this application can use a drone equipped with a lidar to scan the building's exterior surface to obtain three-dimensional point cloud data of the building's exterior surface. Since obstacles such as birds and dust particles may affect the three-dimensional point cloud data of the building's exterior surface during the acquisition process, this application can perform noise reduction processing on this data to remove noisy data, and perform BIM modeling based on the noise-reduced three-dimensional point cloud data to obtain a first triangular mesh model.

[0033] Step 102: Calculate the first folding cost of each edge of each triangle in the first triangular mesh model based on the quadratic error metric method.

[0034] It should be noted that the initially generated BIM models are usually large in size, resulting in problems such as high computer resource consumption, slow loading speed, and display crashes. They place high demands on the hardware configuration and network resources of the operating platform. Real-time interaction and rapid loading are not possible on digital twin platforms, leading to slow acquisition of building exterior surfaces. Related technologies can use edge folding algorithms based on quadratic error metrics to fold edges of the model's triangular mesh according to folding costs (higher folding costs and lower priority for more complex surfaces, and lower folding costs and higher priority for flatter surfaces). Each iteration selects the edge with the lowest folding cost for folding, and after multiple iterations, the triangular mesh is simplified, enabling multi-level simplification of the model and faster processing speed. However, existing folding cost calculations only consider distance and do not take into account factors such as the area of ​​local triangular mesh regions, edge length, and vertex curvature, resulting in a single measurement standard. The building's exterior surface includes windows, protruding balconies and eaves, brick joints, sharp edges, damaged areas, or unfinished surfaces. Its BIM model has rich details, with numerous small triangles. Furthermore, in areas such as inflection points and sharp corners in the BIM 3D mesh model, the model exhibits significant variations and curvature. If folded uniformly like flat areas, the model would lose details such as the edges and corners of the building's exterior surface, resulting in deformation or blurring of the modeled exterior surface, failing to accurately reflect the actual surface conditions. Therefore, this application modifies the first folding cost based on the regional characteristics of high-detail-density and high-shape-variation areas, making the modified folding cost more consistent with the actual conditions of the building's exterior surface.

[0035] Step 103: Based on the first inhibition factor and / or the second inhibition factor of each edge, modify the first folding cost of each edge to obtain the second folding cost of each edge.

[0036] The first inhibition factor is used to suppress the folding cost in high detail density regions, and the second inhibition factor is used to suppress the folding cost in high shape change regions.

[0037] In one possible implementation, this application combines a first suppression factor and a second suppression factor according to a preset rule to form a combined suppression coefficient, and then applies a weighted correction to the basic folding cost based on the combined suppression coefficient. It should be noted that the combined suppression coefficient determined in this application has a value close to 1 in regions with indistinct features to avoid significantly affecting the original cost; while in regions with dense details or drastic shape changes, the coefficient value should be significantly greater than 1, thereby increasing the folding cost of edges in these regions, reducing the folding of edges in regions with dense details or drastic shape changes, and improving the modeling effect.

[0038] Optionally, the combined inhibition coefficient can be calculated by adding the first inhibition factor and the second inhibition factor and then subtracting a preset value; or by directly multiplying the two inhibition factors to construct the combined inhibition coefficient; or by using the maximum value of the two inhibition factors as the combined inhibition coefficient; or by establishing a decision system based on fuzzy logic to comprehensively determine the final cost correction range based on the input values ​​of the two inhibition factors; this application does not limit this.

[0039] Step 104: Simplify the edges in the first triangular mesh model based on the second folding cost of each edge to obtain the second triangular mesh model.

[0040] In one possible implementation, this application simplifies the process by: establishing an edge selection mechanism based on cost ranking, a standardized mesh topology update process, and a dynamic cost update strategy to ensure that each folding operation selects the edge that has the least impact on the current model shape, and maintains the accuracy of the simplification decision through dynamic updates, and finally outputs the simplified model when the preset termination condition is met.

[0041] It should be noted that the consistency of the model topology needs to be maintained during the simplification process. This ensures that the adjacency relationship is updated in a timely manner after each folding operation, and the folding cost of the affected region is recalculated to maintain the accuracy of subsequent simplification decisions.

[0042] Based on the above technical solution, this application constructs a hierarchical model simplification mechanism by introducing a first inhibition factor and a second inhibition factor to synergistically correct edge folding costs. Compared with the limitations of traditional simplification methods that rely solely on a single geometric error, this application achieves differentiated processing of different feature areas of the building's exterior surface model. By simultaneously incorporating detail density and shape curvature into the folding cost evaluation system, the edge folding priority in high-detail and high-curvature areas is effectively reduced. This significantly reduces the amount of model data while substantially improving the fidelity of the simplified model to the complex features of the building facade, thereby enhancing the accuracy of BIM modeling.

[0043] like Figure 2 As shown, in one possible implementation, for any edge in the first triangular mesh model, the first suppression factor involved in step 103 above can be determined through the following process.

[0044] Step 201: Calculate the average area of ​​the adjacent triangles at each of the two endpoints of the target side.

[0045] Here, an adjacent triangle refers to a triangle whose vertex is one of the two endpoints of the target edge; the target edge is any side in the first triangular mesh model.

[0046] Optionally, the area of ​​the triangle in this application can be determined based on Heron's formula, which will not be elaborated upon here.

[0047] Step 202: Calculate the average area of ​​the adjacent triangles at each of the two endpoints.

[0048] Step 203: Calculate the first inhibition factor of the target edge based on the average of the average area.

[0049] Among them, the first inhibition factor of the target side is negatively correlated with the average area of ​​the adjacent triangles.

[0050] It should be noted that in areas with rich details on the building's exterior surface, triangles are small and numerous, resulting in a dense mesh distribution; while in flat areas, triangles are large and fewer, leading to a sparse mesh distribution. If triangular meshes within a region share a common vertex v, they are considered to be triangular meshes adjacent to vertex v. Therefore, the average area can be calculated by summing the number and area of ​​triangles adjacent to vertex v. Constructing inhibitors This results in a larger inhibition factor in areas with many detailed features on the building's exterior surface, leading to higher reconstruction edge folding costs. Therefore, the priority of edge folding should be reduced to avoid loss of detailed features.

[0051] As one possible implementation, step 203 can be implemented as follows: determine the first inhibition factor adjustment hyperparameter; calculate the ratio of the first inhibition factor adjustment hyperparameter to the average value of the average area; and use the sum of the ratio and the first preset value as the first inhibition factor.

[0052] For example, the first inhibitory factor Satisfy the following formula:

[0053]

[0054] This represents the average area of ​​the triangular mesh adjacent to the two endpoints a and b of the edge, respectively. and The average value, This represents the hyperparameter regulating the first inhibitory factor, typically ranging from 0.0001 to 0.01. The larger the area, the higher the detail protection. The area of ​​a triangular mesh in a flat area is generally greater than 0.005, while in areas with complex textures it is generally less than 0.005. In this application... Set experience value Ensure that areas with more detailed surface textures The value is relatively large. The larger the value, the stronger the suppression, the greater the cost of edge folding, and the lower the folding priority; conversely, the smaller the value, the lower the folding priority in flat regions. The value is basically 0. A value close to 1 indicates minimal suppression, and the cost of edge folding remains essentially unchanged. It should be understood that the length unit of the BIM model needs to be adjusted in conjunction with the first suppression factor to control the hyperparameter. The units must be compatible; for example, the length unit of the BIM model can be meters. Therefore, the above-mentioned first suppression factor adjusts the hyperparameter. The unit is the square of meters.

[0055] In actual calculations, if If the value is 0 or less than the preset non-zero minimum value, then... Set it to the preset non-zero minimum value (such as 10^-6) to avoid division by zero errors.

[0056] Optionally, this application may be based on a first inhibitory factor. The first folding cost is modified, and in this case, for example, the modified second folding cost satisfies... The following formula:

[0057]

[0058] in, As the first inhibitory factor, This is the cost of the first fold.

[0059] Based on this formula, A larger value indicates that the triangular mesh area belongs to a building exterior surface area with more detailed textures; the greater the cost of edge collapsing, the lower the priority of edge collapsing. Conversely, A smaller value indicates that the triangular mesh area belongs to a relatively flat building surface. The smaller the cost of edge folding, the higher the priority of edge folding.

[0060] Based on the above Figure 2 The technical solution described in this application constructs a quantitative perception capability for local detail density by calculating a first suppression factor based on the average area of ​​adjacent triangles at a vertex. This capability can automatically identify dense triangular mesh regions in the model formed by complex textures or fine structures and transform the features of these regions into calculable suppression parameters. By making the suppression factor negatively correlated with the average area, a higher folding cost is ensured in densely detailed regions, thereby effectively preventing these regions from being prematurely flattened in the early stages of simplification and ensuring the integrity of fine structures such as brick seams and carvings.

[0061] like Figure 3 As shown, in one possible implementation, for any edge in the first triangular mesh model, the second suppression factor involved in step 103 above can be determined through the following process.

[0062] Step 301: Calculate the curvature eigenvalues ​​of the two endpoints of the target edge.

[0063] Among them, the curvature feature value is used to characterize the average degree of change of the surface at the location of the endpoint; the target edge is any side in the first triangular mesh model.

[0064] It should be noted that if the building's exterior surface has sharp edges, corners, creases, etc., the triangular mesh of its BIM model will vary greatly, with a large curvature characteristic value, and this area will have rich details. When simplifying the mesh, its folding cost should be increased, the folding priority should be reduced, and the degree of model deformation should be minimized. Specifically, in this application, any triangular mesh can be selected, and its three vertices can be set as follows: , , Calculate the unit normal vector of the triangular mesh. Because there may be a vertex. When adjacent triangles have different areas, in order to prioritize showcasing the main trend of the building's exterior surface, the vertices need to be calculated using area weighting. The normal vector. During the weighting process, larger triangular meshes contribute more weight to the vertex normal vector, but this does not affect the final vertex curvature. For example, compared to the vertex... The adjacent grids are all flat large triangular grids, at which point the vertices... The normal vector of the vertex is aligned with the normal vector of the adjacent triangular mesh, with an included angle of 0 and a curvature of 0; it is also aligned with the vertex. The adjacent grids are all complex small triangular meshes, at which point the vertices... The direction of the normal vector of the vertex differs greatly from that of the normal vector of the adjacent triangular mesh, and the included angle is also large, resulting in a large curvature; with the vertex When adjacent triangles contain both large and small triangles (such as decorative lines), the normal vector of the small triangle may have a larger angle with the normal vector of the large triangle, but due to the smaller area weighting, the angle at the vertices is smaller. The influence of the normal vector is small, and the normal vector of the large triangular mesh ultimately dominates. The curvature will not increase inexplicably due to the small triangular mesh, nor will it approach 0 due to the large triangular mesh. However, it will be slightly larger than the curvature of the triangular mesh in the flat area, reflecting local details.

[0065] Based on the above requirements, as a possible implementation, step 301 can be implemented through the following process: calculate the unit normal vector of each adjacent triangle at the endpoint; multiply the unit normal vector of each adjacent triangle by its area to obtain the weighted normal vector of each adjacent triangle; sum all the weighted normal vectors and normalize the summation result to obtain the average weighted normal vector of the endpoint; calculate the spatial angle between the unit normal vector and the average weighted normal vector of each adjacent triangle; calculate the average value of the spatial angle as the curvature characteristic value.

[0066] As an example, vertices Unit normal vector based on area weighting Satisfy the following formula:

[0067]

[0068] Where k represents the vertex The number of adjacent triangular meshes, where j represents the number of adjacent vertices. The j-th adjacent triangular mesh, where j is greater than or equal to 1. Represents the vertex The area of ​​the j-th adjacent triangular mesh. Represents the vertex The unit normal vector of the j-th adjacent triangular mesh. This represents the magnitude of the area-weighted normal vector.

[0069] As yet another example, vertices curvature eigenvalues Satisfy the following formula:

[0070]

[0071] Represents the calculation of vertices normal vector Normal vectors of adjacent triangular meshes The included angle (Usually the dot product formula is used) The curvature eigenvalue is obtained by summing and averaging the included angles. The larger the average included angle, the greater the curvature of that vertex. It can reflect the degree of curvature at the vertex. The larger the curvature characteristic value, the more distorted the local surface of the building is (such as sharp corners, raised decorations, etc.). The smaller the included angle, the smaller the curvature, and the flatter the area of ​​the building's exterior surface.

[0072] The calculation process of the unit normal vector in the embodiments of this application can refer to the prior art, and will not be described in detail here.

[0073] Step 302: Calculate the second inhibition factor of the target edge based on the maximum value of the curvature eigenvalues ​​of the two endpoints.

[0074] Among them, the second inhibition factor is positively correlated with the maximum value.

[0075] As an example, combining the second inhibitory factor in step 301 above. Satisfy the following formula:

[0076]

[0077] in, This represents the curvature influence factor, controlling the degree to which curvature amplifies the cost; an empirical value of 1.5 is used. This represents the maximum curvature at the two endpoints of an edge. If one endpoint has a larger curvature, it is assumed that the triangular mesh region connected by this edge may contain details such as corners or sharp edges, and the suppression factor is increased. This indicates that the protection of high curvature regions is enhanced by quadratic expansion, which makes edge folding more costly and thus lowers the priority.

[0078] It should be noted that, The selection mechanism ensures that as long as any endpoint of an edge is in a high-curvature region, the entire edge will receive the corresponding protection strength. The curvature influence factor β, as a preset parameter, determines the contribution of curvature features to the final suppression effect. By... By performing squaring operations, non-linear cost increases occur in high curvature regions, thereby providing stronger protection for feature areas such as sharp corners and raised decorations during the simplification process.

[0079] Besides using the maximum value of the curvature characteristic values ​​of the two endpoints, those skilled in the art can also use other parallel implementation methods, such as: taking the arithmetic mean of the curvature characteristic values ​​of the two endpoints as the calculation benchmark; or taking the weighted average of the curvature characteristic values ​​of the two endpoints, wherein the weights are dynamically adjusted according to the triangle density of the region where the endpoints are located; or establishing an evaluation system based on the sum of the curvature characteristic values ​​of the endpoints. This application does not limit this.

[0080] Based on the above technical solution, this application constructs a mechanism for identifying and protecting sharp features of the model shape by calculating a second inhibition factor based on the endpoint curvature feature value. This mechanism can accurately capture geometric features such as corners and edges on the building's outer surface. By taking the maximum value of the endpoint curvature as the judgment criterion, it ensures that edges related to sharp features are fully protected. This effectively postpones the folding operation of high curvature areas during model simplification, preventing the smooth deformation of the building outline features during simplification and maintaining the geometric accuracy of the building model.

[0081] like Figure 4 As shown, in one possible implementation, the present application can obtain the second folding cost by modifying the first folding cost based on the first inhibition factor and the second inhibition factor. In this case, the process of calculating the second folding cost in step 103 above includes the following steps:

[0082] Step 401: Subtract the second preset value from the sum of the first and second inhibition factors of the target edge to obtain the combined inhibition factor.

[0083] Step 402: The product of the combined suppression factor and the first folding cost is used as the second folding cost of the target edge.

[0084] As an example, the cost of the second fold Satisfy the following formula:

[0085]

[0086] It should be noted that, due to the first inhibitory factor given above... Second inhibitory factor All are greater than 1. It must be greater than 2, therefore The coefficient before should be used To ensure that the total cost remains constant when folding the triangular mesh edges in flat areas of the building, the edge folding cost is increased in areas with rich details, thus reducing the folding priority and preventing the loss of details and model deformation. Therefore, based on the second folding cost determined in this application, when the model is in a flat area with little curvature change, the values ​​of ω and δ both approach 1. At this time, the combined suppression coefficient is about 1, and the second folding cost is basically consistent with the first folding cost, without affecting the conventional simplification process. When the model is in an area with dense details or drastic shape changes, at least one of ω and δ is significantly greater than 1. At this time, the combined suppression coefficient increases accordingly, resulting in a corresponding increase in the second folding cost, thereby effectively reducing the edge folding priority in this area.

[0087] It should be noted that the second folding cost can also be achieved in other ways, such as: constructing a combined suppression coefficient by weighted summation of the two suppression factors, where the weight coefficients are dynamically adjusted according to the model feature protection requirements; or using the maximum value of the two suppression factors as the combined suppression coefficient; or multiplying the two suppression factors to construct the combined suppression coefficient. This application does not limit this.

[0088] Based on the above technical solution, this application constructs a complete cost correction system by combining two suppression factors and multiplying them by the base cost. This achieves a synergistic effect of detail density protection and shape feature protection. A linear combination method ensures a superimposed protection effect in regions with both high detail density and high curvature. Subtracting a preset value ensures that the correction coefficient is close to one in flat regions, without affecting the conventional simplification process, while producing significant suppression in feature regions, thus forming an intelligent differentiated processing strategy for different regions of the model.

[0089] like Figure 5 As shown, in one possible implementation, the process of simplifying the edges in the first triangular mesh model in step 104 above includes:

[0090] Step 501: Perform heap sort on the edges according to the order of their second fold cost from smallest to largest.

[0091] The process of heap sort can refer to existing technologies, and this application does not limit it.

[0092] Step 502: Select the edge with the lowest second folding cost in the current sorting and fold it, and update the connection relationship between the remaining edges and vertices to obtain a new triangular mesh model.

[0093] As one possible implementation, step 502 can be implemented through the following process: delete the currently required folding edge, the two endpoints of the folding edge, and the two triangles with the folding edge as a common edge; calculate the positions of the two vertices of the folding edge based on a preset vertex generation algorithm to generate new vertices; update the topological connection relationship of the triangular mesh model based on the new vertices and the other two vertices of the two triangles to form new triangles.

[0094] Step 503: Recalculate the second folding cost of each edge in the new triangular mesh model.

[0095] Step 504: Iterate through steps 501 to 503 until the preset number of iterations is reached, or the simplification error of the new triangular mesh model is less than the first threshold. Output the current triangular mesh model as the second triangular mesh model.

[0096] The simplification process mainly includes three steps: edge priority sorting, mesh topology update, and termination judgment. Edge priority sorting is implemented using a heap sort algorithm to ensure that the edge with the lowest second-fold cost is selected for operation in each iteration. Mesh topology update includes edge folding and generating new topological relationships. The termination judgment is determined based on a preset upper limit for the number of iterations and a simplification error threshold. After each edge folding operation, due to changes in the local geometry and topological relationships of the model, the second-fold cost of the affected region needs to be recalculated, including updating the basic fold cost of the relevant edges and the corresponding suppression factors. This dynamic update mechanism ensures that simplification decisions are always made based on the current state of the model, maintaining the accuracy and rationality of the simplification process.

[0097] It should be noted that, apart from the simplification processes described above, each step can be implemented using corresponding parallel technical solutions. For example, in terms of sorting algorithms, in addition to heap sort, quicksort, merge sort, or other comparison-based sorting algorithms can also be used. Regarding the termination condition settings, besides the number of iterations and the simplification error threshold, the model data reduction ratio, the maximum folding cost threshold, or processing time limits can also be used as auxiliary judgment criteria. As for the iteration control strategy, different strategies such as fixed-ratio simplification, hierarchical simplification, or multi-precision simplification can be adopted; this application does not limit these strategies.

[0098] Optionally, after outputting the second triangular mesh model, this application can export the second triangular mesh model as a new OBJ file for easy and quick opening and continued use later.

[0099] Based on the above technical solutions, this application constructs a complete simplification process through heap sorting, iterative folding, and dynamic updating, achieving automation and optimization of the simplification process. Sorting by cost ensures that each fold selects the edge with the least impact on the model, and recalculating the cost of the affected region guarantees the real-time accuracy of the simplification decision. The termination condition provides a balance between simplification quality and efficiency, enabling the simplification process to complete automatically within a preset accuracy range, outputting an optimized model that meets data volume requirements while preserving key features.

[0100] like Figure 6 As shown, in one possible implementation, the process of constructing the first triangular mesh model in step 101 above can be specifically implemented through the following steps:

[0101] Step 601: Obtain the initial three-dimensional point cloud data of the building's exterior surface obtained by scanning the building's exterior surface using a lidar device.

[0102] Step 602: Perform outlier filtering on the initial 3D point cloud data to remove noise and obtain the processed 3D point cloud data.

[0103] In one possible implementation, the point cloud data of buildings is arranged in dense, regular blocks, with small neighborhood distances between points. In contrast, the point cloud data of birds and dust particles are scattered, isolated points and dynamic targets; they may only appear at a certain moment in a laser detection and ranging (LiDAR) scan, with very few point clouds (possibly only 1-2 points), no continuous spatial distribution pattern, and large spatial distances from surrounding points. Therefore, this application can remove them by outlier filtering.

[0104] Optionally, step 602 can be implemented through the following process: setting a distance threshold based on the neighborhood distance between each point cloud data in the initial 3D point cloud data; deleting target points from the initial 3D point cloud data to obtain processed 3D point cloud data, wherein the target point is a point whose number of point cloud data within a radius of the distance threshold centered on the target point is less than the second threshold. The specific process of outlier filtering can be found in relevant technologies, and will not be elaborated upon here.

[0105] Step 603: Convert the processed 3D point cloud data into a triangular mesh based on the Poisson surface reconstruction algorithm.

[0106] Step 604: Perform geometric processing, semantic mapping, and parametric modeling on the triangular mesh to obtain the first triangular mesh model.

[0107] In other words, this application can convert point cloud data into triangular meshes based on Poisson surface reconstruction technology, and construct an initial BIM model of the building wall by combining geometric processing, semantic mapping, and parametric modeling. The processes of Poisson surface reconstruction algorithm, geometric processing, semantic mapping, and parametric modeling can be found in existing technologies, and will not be elaborated upon here.

[0108] Based on the above technical solution, this application defines the update rules for the mesh topology through specific folding operation steps, ensuring the consistency and correctness of the model's geometry and topological relationships during the simplification process. By standardizing the deletion of old elements and the generation of new elements, mesh tearing, holes, or non-manifold structures that may occur during simplification are avoided. This rigorous operation process guarantees the integrity and usability of the simplified model, providing a reliable geometric foundation for subsequent BIM applications.

[0109] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0110] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for rapidly constructing a building exterior surface model based on BIM technology, characterized in that, The method includes: A first triangular mesh model is constructed based on the building's exterior surface data. The first triangular mesh model is a model composed of multiple triangles spliced ​​together to represent the shape of the building's exterior surface. The first folding cost of each edge of each triangle in the first triangular mesh model is calculated based on the quadratic error metric method. Based on a first suppression factor and / or a second suppression factor for each edge, the first folding cost of each edge is corrected to obtain a second folding cost for each edge; wherein, the first suppression factor is used to suppress the folding cost of high detail density regions, and the second suppression factor is used to suppress the folding cost of high shape change regions; The edges in the first triangular mesh model are simplified based on the second folding cost of each edge to obtain the second triangular mesh model; The method further includes, before correcting the first folding cost based on the first and second inhibition factors: Calculate the average area of ​​the adjacent triangles at each of the two endpoints of the target edge; the adjacent triangles refer to triangles with one of the two endpoints of the target edge as a vertex; the target edge is any side in the first triangular mesh model; Calculate the average of the average areas of the adjacent triangles at each of the two endpoints; Based on the average of the average areas, a first inhibition factor for the target edge is calculated, and the first inhibition factor for the target edge is negatively correlated with the average area of ​​the adjacent triangles; Calculate the curvature characteristic values ​​of the two endpoints of the target edge, whereby the curvature characteristic values ​​characterize the average degree of change of the surface at the endpoints; the target edge is any edge in the first triangular mesh model. Based on the maximum value of the curvature feature values ​​of the two endpoints, a second inhibition factor for the target edge is calculated, and the second inhibition factor is positively correlated with the maximum value.

2. The method for rapid construction of building exterior surface models based on BIM technology according to claim 1, characterized in that, The calculation of the first inhibition factor of the target edge based on the average of the average area includes: Determine the hyperparameters regulating the first inhibitory factor; Calculate the ratio of the first inhibition factor regulation hyperparameter to the average value of the average area; The sum of the ratio and the first preset value is used as the first inhibition factor.

3. The method for rapid construction of building exterior surface models based on BIM technology according to claim 1, characterized in that, The calculation of the curvature characteristic values ​​of the two endpoints of the target edge includes: Calculate the unit normal vector of each adjacent triangle at the endpoint; Multiply the unit normal vector of each adjacent triangle by its area to obtain the weighted normal vector of each adjacent triangle; Summing all weighted normal vectors and normalizing the summation result yields the average weighted normal vector at the endpoints. Calculate the spatial angle between the unit normal vector of each adjacent triangle and the average weighted normal vector; The average value of the spatial angles is calculated as the curvature characteristic value.

4. The method for rapid construction of building exterior surface models based on BIM technology according to claim 1, characterized in that, The step of correcting the first folding cost of each edge based on the first suppression factor and / or the second suppression factor of each edge to obtain the second folding cost of each edge includes: The combined inhibition factor is obtained by subtracting the second preset value from the sum of the first inhibition factor and the second inhibition factor of the target edge. The product of the combined suppression factor and the first folding cost is used as the second folding cost of the target edge.

5. The method for rapid construction of building exterior surface models based on BIM technology according to claim 1, characterized in that, The step of simplifying the edges in the first triangular mesh model based on the second folding cost of each edge to obtain the second triangular mesh model includes: Step 1: Perform heap sort on the edges according to their second fold cost in ascending order; Step 2: Select the edge with the lowest folding cost in the current sorting and fold it, then update the connection relationship between the remaining edges and vertices to obtain a new triangular mesh model; Step 3: Recalculate the second folding cost of each edge in the new triangular mesh model; Iteratively execute steps one through three above until the preset number of iterations is reached, or the simplification error of the new triangular mesh model is less than the first threshold, and output the current triangular mesh model as the second triangular mesh model.

6. The method for rapid construction of building exterior surface models based on BIM technology according to claim 5, characterized in that, Select the edge with the second smallest folding cost in the current sort and fold it, including: Delete the currently required fold edge, the two endpoints of the fold edge, and the two triangles that share the fold edge as a common edge; The positions of the two vertices of the folded edge are calculated based on a preset vertex generation algorithm to generate new vertices; Based on the new vertex and the other two vertices of the two triangles, the topological connection relationship of the triangular mesh model is updated to form a new triangle.

7. The method for rapid construction of building exterior surface models based on BIM technology according to claim 1, characterized in that, The construction of the first triangular mesh model based on the building's external surface data includes: Acquire initial 3D point cloud data of the building's exterior surface obtained by scanning the exterior surface using a lidar device; The initial 3D point cloud data is filtered to remove noise and outlier points, resulting in the processed 3D point cloud data. The processed 3D point cloud data is converted into a triangular mesh based on the Poisson surface reconstruction algorithm; The first triangular mesh model is obtained by performing geometric processing, semantic mapping, and parametric modeling on the triangular mesh.

8. The method for rapid construction of building exterior surface models based on BIM technology according to claim 7, characterized in that, The outlier filtering process performed on the initial 3D point cloud data to remove noise results in the following processed 3D point cloud data: A distance threshold is set based on the neighborhood distance between each point cloud data in the initial three-dimensional point cloud data; Delete the target point from the initial three-dimensional point cloud data to obtain the processed three-dimensional point cloud data, wherein the target point is a point whose number of point cloud data within a radius of the target point and the distance threshold is less than a second threshold.

Images