Method, device and storage medium for identifying component differences in different versions of BIM models

Abstract

The application discloses a component difference identification method and device in different version BIM models and a storage medium. The method comprises the following steps: selecting a component to be compared from different version BIM models; extracting contour lines of the component to be compared in each version BIM model; sampling the contour lines to obtain a key point set of the contour lines; sampling a surface of the component to be compared to obtain a description point set of the surface; calculating a logarithmic distance between each key point in the key point set and all points in the description point set; constructing a statistical histogram of the component to be compared according to the logarithmic distance; and identifying component differences according to the statistical histograms of the components to be compared in different version BIM models. The application can identify subtle differences of components, and improve identification efficiency and identification accuracy.

Description

Technical Field

[0001] This invention belongs to the field of building information modeling technology, and in particular relates to a method, device and storage medium for identifying component differences in different versions of BIM models. Background Technology

[0002] With the development of CIM technology, the application of BIM models is becoming increasingly widespread. Automatically identifying differences between different versions of BIM models plays a crucial role in driving model data consumption, such as online BIM model review, rapid quantity verification, and scheme comparison. Due to differences in application industries, software, and vendors, BIM models exhibit multi-source heterogeneity. In the model data application phases of engineering construction management and operation and maintenance, to achieve BIM model integration, grid models are generally used for unified description, enabling business applications based on BIM models in open environments such as browsers. Frequent modifications to BIM models due to business needs result in numerous versions. In application scenarios such as BIM model delivery, change review, and scheme comparison, it is necessary to identify the differences of the same model component in different versions of BIM models, with the key being the identification of subtle design differences. Currently, the differences between different model versions are mainly identified through manual observation, which is inefficient and lacks accuracy.

[0003] The difficulty in BIM model comparison lies in comparing geometric information. Due to differences in the principles of geometric modeling and the diverse sources of models, there is currently no clearly efficient method for comparing geometric information. Existing model comparison techniques are mainly divided into two categories:

[0004] One type is comparison based on model semantic information description. For example, the comparison of models based on IFC format and XDB format is achieved by comparing text files, which is essentially a text comparison method.

[0005] Another type of comparison is based on the model itself. For example, methods based on geometric similarity, topological similarity, perspective projection, and contours primarily involve selecting model feature definitions, such as topological structure or distance statistical histograms. Methods that statistically describe models based on feature definitions mainly classify models, identifying similar models with different parameters. This requires a stable description of component categories, a large number of sampling points, and computation time, and must ignore subtle differences between similar models. Research indicates that current model comparison techniques mainly rely on pattern recognition for model classification, characterized by ignoring minor differences or topologically invariant pose differences between similar models. Summary of the Invention

[0006] The purpose of this invention is to provide a method, device, and storage medium for identifying differences in components in different versions of BIM models, in order to solve the problem of identifying differences in similar components in different versions of BIM models, especially the identification of subtle differences.

[0007] This invention solves the above-mentioned technical problems through the following technical solution: a method for identifying component differences in different versions of BIM models, comprising the following steps:

[0008] Select components to be compared from different versions of BIM models;

[0009] Extract the outline of the component to be compared in each version of the BIM model;

[0010] The contour line is sampled to obtain the key point set of the contour line;

[0011] The surface of the component to be compared is sampled to obtain a set of descriptive points on the surface;

[0012] Calculate the logarithmic distance between each key point in the key point set and all points in the description point set;

[0013] Construct a statistical histogram of the component to be compared based on the logarithmic distance;

[0014] Differences between components are identified by statistical histograms of the components to be compared in different versions of BIM models.

[0015] Furthermore, the specific process for extracting the outline of the component to be compared in each version of the BIM model is as follows:

[0016] The edge information of the component to be compared is extracted cell by cell to obtain the edge information set. Among them, e i For the information of the i-th edge, Let be the coordinates of the starting point of the i-th edge. Let n be the coordinates of the endpoint of the i-th edge. i Let t be the normal to the cell corresponding to the i-th edge, and t be the number of edges in the component to be compared.

[0017] The vertices of the components to be compared are sorted according to the priority order of X / Y / Z, and the edge associated with each vertex is stored.

[0018] The edges associated with each vertex are deduplicated to obtain the remaining edges;

[0019] All remaining edges are spliced ​​together to obtain the outline of the component to be compared.

[0020] Preferably, the specific implementation process for deduplicating the edges associated with each vertex is as follows:

[0021] For any vertex, extract the non-repeating edges associated with that vertex;

[0022] Calculate the included angle θ between the normals of the two overlapping edges, when If the condition is met, retain the two duplicated edges; otherwise, remove the two duplicated edges. The included angle threshold,

[0023] The remaining edges are composed of non-repeating edges and retained repeating edges.

[0024] Furthermore, the specific implementation process for sampling the contour line is as follows:

[0025] The key point set is obtained by uniformly sampling along the contour line according to the set sampling interval; or,

[0026] The key point set is obtained by sampling along the contour line according to the set sampling quantity.

[0027] Furthermore, the surface of the component to be compared is sampled using a Poisson disk uniform sampling method.

[0028] Furthermore, the specific formula for calculating the logarithmic distance is as follows:

[0029]

[0030] Among them, l ij Key points on contour line C And the j-th point d in the description point set j The logarithmic distance.

[0031] Furthermore, the statistical histogram of the component to be compared is constructed as follows:

[0032] For the key point p i Normalize each log distance in the log distance set to obtain the normalized distance set;

[0033] The normalized distance set is divided into M equal parts, and the statistical value of each part is calculated. The specific calculation formula is as follows:

[0034]

[0035]

[0036] in, For key point p i The logarithmic distance set is the statistical value of the k-th division of the histogram, where D is the description set, n is the number of points in the description set, and L is the distance between the logarithmic distance set and the logarithmic distance set. ijFor the normalized distance in the normalized distance set, (b k-1 ,b k [] represents the k-th division after dividing M into equal parts, where k∈[1,M];

[0037] A statistical histogram of the key points is formed based on the statistical values ​​of each equal division. The expression of the statistical histogram is as follows:

[0038]

[0039] in, For key point p i The statistical histogram, the items in [] represent key points p. i The statistical values ​​of the kth division of the statistical histogram, where D is the set of descriptive points;

[0040] Calculate the statistical histograms of all key points and generate a statistical histogram for the component to be compared. The specific expression is as follows:

[0041]

[0042] Where H is the statistical histogram of the component to be compared, and r is the number of key points in the key point set.

[0043] Furthermore, the specific implementation process for identifying component differences based on the statistical histograms of the components to be compared from different versions of the BIM model is as follows:

[0044] The shape distance of the components to be compared is calculated based on the statistical histograms of the components from different versions of BIM models. The specific calculation formula is as follows:

[0045]

[0046] Where e represents the shape distance, and the statistical histogram of component A to be compared in one version of the BIM model. r is the number of keypoints in the keypoint set. For key point p i The logarithmic distance set is the statistical value of the k-th division of the histogram, where D is the descriptive point set. The statistical histogram of component B to be compared in another version of the BIM model is also shown.

[0047] When e≤e θ When there is no difference between component A and component B, there is a difference; otherwise, there is a difference, where e θ This is the distance threshold.

[0048] Preferably, e θ =10 -3 .

[0049] Furthermore, the method also includes:

[0050] When the components are identical, the center points of the two components to be compared are determined by calculating the coordinate range of the two components to be compared; and / or, scaling is determined by calculating the distance between the maximum and minimum vertices of the AABB bounding boxes of the two components to be compared; and / or, rotation is determined by calculating the vector angle between the two components to be compared.

[0051] Based on the same inventive concept, embodiments of the present invention provide an electronic device, the electronic device comprising:

[0052] At least one processor; and,

[0053] A memory communicatively connected to the at least one processor; wherein,

[0054] The memory stores instructions that can be executed by the at least one processor. When the instructions are executed by the at least one processor, the at least one processor performs the component difference identification method in different versions of the BIM model as described above.

[0055] Based on the same inventive concept, embodiments of the present invention provide a non-transitory computer-readable storage medium storing computer instructions that, when executed by at least one processor, cause the at least one processor to perform the component difference identification method in different versions of BIM models as described above.

[0056] Beneficial effects

[0057] Compared with the prior art, the advantages of the present invention are as follows:

[0058] The present invention provides a method, electronic device and storage medium for identifying component differences in different versions of BIM models. It uses statistical histograms to describe the features of components to be compared in different versions of BIM models, transforming the problem of component differences or similarity measurement in different versions into a differential measurement of two statistical histograms. Then, by calculating the distance, the differential comparison of the two statistical histograms is realized, which can identify subtle differences in components and improve the identification efficiency and accuracy.

[0059] This invention demonstrates high accuracy in identifying the impact of noise from design modifications and surface subdivisions. By employing a uniform sampling method, it achieves highly accurate comparison results with a smaller number of samples, thus improving comparison efficiency and saving time. It has significant practical applications in scheme comparison and model verification. Attached Figure Description

[0060] To more clearly illustrate the technical solution 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 one embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 This is a flowchart of the component difference identification method in an embodiment of the present invention;

[0062] Figure 2 This is an example diagram of component outline extraction in an embodiment of the present invention;

[0063] Figure 3 This is a comparison diagram of feature descriptions in an embodiment of the present invention. Detailed Implementation

[0064] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. 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.

[0065] The technical solutions of this application will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0066] To overcome the problems of low cost-effectiveness, large identification errors of similar components, and limited applicability of existing methods, this invention provides a difference identification method to solve or partially solve the problem of component model comparison. It is especially suitable for the field of civil engineering, because there are many non-standard components in this field, whose shape characteristics are difficult to describe in a parametric way, and may have arbitrary topological structures.

[0067] Taking the identification of component differences in BIM models V1.0 and V1.3 as examples, this invention illustrates the effectiveness of a component difference identification method proposed in this paper for different versions of BIM models.

[0068] like Figure 1 As shown in this embodiment, a method for identifying component differences in different versions of a BIM model includes the following steps:

[0069] Step 1: Select the components to be compared from the V1.0 and V1.3 BIM models.

[0070] According to national or industry classification and coding standards for BIM, each component should have a unique identification code throughout the entire life cycle of the project, and the twin iteration of the BIM model can be tracked through the identification code of the component.

[0071] In this embodiment, component identification codes are used as the basis for rapid component classification and comparison object determination. In the V1.0 and V1.3 BIM models: different codes indicate newly added or deleted components; identical codes indicate the same component. The differences between the two are compared to determine the component to be compared. A single BIM model contains multiple components. If a component exists in V1.0 but not in V1.3, it indicates that the component was deleted in V1.3; if a component is absent in V1.0 but present in V1.3, it indicates that the component was added in V1.3. Deleted components are not compared; newly added components are simply identified.

[0072] This invention uses a key point distance context description method to define the component model. It uniformly samples along the contour boundary of the component to form key points describing the component model. Then, it collects shape description points on the surface of the component model using a Poisson disk uniform sampling method, and constructs a distance context description between each key point and the description points.

[0073] Step 2: Extract the outline of the component to be compared in each version of the BIM model.

[0074] For the components to be compared in both the V1.0 and V1.3 BIM models, the outlines of the components are extracted. The specific implementation method is as follows:

[0075] Step 2.1: Extract the edge information of the component to be compared cell by cell to obtain the edge information set E:

[0076]

[0077] Among them, e i For the information of the i-th edge, Let be the coordinates of the starting point of the i-th edge. Let n be the coordinates of the endpoint of the i-th edge. i Let t be the normal to the cell corresponding to the i-th edge, and t be the number of edges in the component to be compared.

[0078] Assume that both the V1.0 and V1.3 BIM models are triangular mesh models, meaning each cell is a triangular face. For example... Figure 2 The mesh diagram of the component to be compared is shown below. Figure 2It can be seen that the component to be compared consists of 4 triangular faces and has 12 edges, i.e., t = 12, where t is equal to the product of the number of cells and the number of edges in each cell. The normal n of the cell corresponding to the i-th edge. i This refers to the normal of the triangle face. For example, the normal of the edge formed by starting point P2 and ending point P3 is the normal of the triangle face P1P2P3.

[0079] The coordinates of the vertices of each triangle face in a BIM model are known. BIM models are typically exported from design software, and a 3D coordinate system can be constructed within the design software. Based on this 3D coordinate system, the coordinates of the vertices of each cell in the BIM model can be obtained. Alternatively, a 3D coordinate system can be constructed on the component to be compared, such as... Figure 2 As shown, by constructing a three-dimensional coordinate system with vertex P0 as the origin, the coordinates of the other vertices of the component can be obtained.

[0080] Step 2.2: Sort the vertices of the components to be compared according to the priority order of X / Y / Z, and store the edges associated with each vertex.

[0081] by Figure 2 For example, the component to be compared has 6 vertices, P0 to P5. For the coordinates of the 6 vertices, they are first sorted by X in ascending order. If X values ​​are equal, they are then sorted by Y values ​​in ascending order. If both X and Y values ​​are equal, they are finally sorted by Z values ​​in ascending order. The vertices sorted according to the priority of X / Y / Z are P5, P0, P2, P4, P1, and P3. For each vertex, the edges associated with that vertex are stored. For example, for P5, the associated edges are P5P0, P5P1, P5P4, and P5P1.

[0082] Step 2.3: Perform deduplication on the edges associated with each vertex to obtain the remaining edges.

[0083] Specifically, step 2.31: For any vertex, extract the non-repeating edges associated with that vertex.

[0084] For example, for vertex P5, its non-repeating edges are P5P0 and P5P4, which are both boundary edges and need to be included in the contour line.

[0085] Step 2.32: Calculate the included normal angle θ between the two overlapping edges, based on... Determine whether to retain duplicate borders, where, The included angle threshold,

[0086] By controlling The value of can control the density of the outline. When the value is 0, only edges generated by θ = 90° are accepted; When θ is within the range of 45° to 135°; When θ can be 0° to 180°, it can be any angle between 0° and 180°. In this embodiment, let θ be... However, when the obtained outline is empty, The value is recommended to be set to 1, which is generally due to the presence of spherical components.

[0087] For example, the repeating edge P5P1 of vertex P5 represents the common edge of two triangular faces. In this case, it is necessary to determine the angle θ between the normals of the two edges. If the two overlapping edges are not found, retain them; that is, take the common edge of the two triangular faces as the outline. Otherwise, remove the two overlapping edges. Since triangular faces P5P0P1 and P1P4P5 are coplanar, θ = 0, and edge P5P1 is discarded.

[0088] Step 2.33: The remaining edges are formed by the non-repeating edges and the retained repeating edges.

[0089] When the two duplicate edges are retained, the remaining edges of vertex P5 are P5P0, P5P1, and P5P4; when the two duplicate edges are removed, the remaining edges of vertex P5 are P5P0 and P5P4.

[0090] Step 2.4: Sew together all remaining edges to obtain the outline of the component to be compared.

[0091] After the processing in step 2.3, the number of edges of the component to be compared is reduced. The remaining edges are sorted and connected according to the connection order of the endpoints of the edges to obtain the final outline.

[0092] Step 3: Sample the contour line to obtain the key point set of the contour line. in Let r be the i-th key point on the contour line C, and r be the number of key points on the contour line C.

[0093] There are two ways to sample key points on the contour line: First, uniform sampling is performed along the contour line according to the set sampling interval, so that key point data of the same accuracy can be obtained for different types of components, ensuring that the accuracy is not affected by the volume and shape of the component; Second, sampling is performed along the contour line according to the set sampling quantity, that is, the number of points in the key point set is fixed.

[0094] This embodiment uses a set sampling interval for uniform sampling. By reasonably controlling the sampling interval of key points, it is possible to perform equal-density sampling on components of different sizes, without causing insufficient key points due to a fixed number of samples.

[0095] The advantage of using contour line key points is that it ensures relatively stable model features, can accurately reflect whether the model has deformed, and is more sensitive to changes in the model before and after modifications. The density of the acquired contour lines and the number of key points are easy to control, offering a high degree of freedom.

[0096] Step 4: Sample the surface of the component to be compared to obtain the set of descriptive points of the surface.

[0097] A common method for obtaining a descriptive point set D by sampling the surface of a component is to perform random sampling within a triangular mesh. When the number of samples is large enough, the component can be completely described. However, the number of "sufficient" samples is uncertain for different types of components, making it difficult to guarantee the universality of the method. In addition, the process of surface subdivision or simplification of the component model is not uniform—subdivision or simplification is more frequent in areas requiring more detailed description, such as curved surfaces, while standardized components such as beams, slabs, and columns are not subdivided. Therefore, this results in denser local sampling points on the component, leading to greater randomness in the results. However, for human visual recognition models, the contribution of each point on the component is the same.

[0098] This invention provides a complete comparison of components, requiring only the comparison of the complete differences between two components. Taking into account the influence of the number of samples and uniformity, a Poisson disk uniform sampling method is used to uniformly sample the surface of the component model. This allows for a stable description of the model with fewer sampling points and also resists the effects of surface subdivision or simplification. The distribution of description points is optimized by setting a fixed sampling radius.

[0099] In this embodiment, a Poisson disk uniform sampling method is used to collect 100,000 points on the surface of the component.

[0100] Step 5: Calculate each keypoint in the keypoint set. The logarithmic distance to all points in the point set D is calculated using the following formula:

[0101]

[0102] Among them, l ij Key points on contour line C And the j-th point d in the description point set j Logarithmic distance, key points on contour line C logarithmic distance set n is the number of points in the point set.

[0103] Step 6: Construct a statistical histogram of the components to be compared based on the logarithmic distance.

[0104] Logarithmic distance is formed about key points The distance context description, and the construction method of the statistical histogram featuring logarithmic distance, is as follows:

[0105] Step 6.1: For key point p i logarithmic distance set L i Each logarithmic distance l in ij After normalization, we obtain the normalized distance set L, where each element is denoted as L. ij L = {L i1 ,L i2 ,…,L ij ,…,L in}, L ij Values ​​between [0, 1];

[0106] Step 6.2: Divide the normalized distance set L into M equal parts and calculate the statistical value of each part. The specific calculation formula is as follows:

[0107]

[0108]

[0109] in, For key point p i The logarithmic distance set is the statistical value of the k-th division of the histogram, where D is the description point set and n is the number of points in the description point set. k-1 ,b k Let be the k-th division after dividing M into equal parts, where k∈[1,M]. Assume n=9, M=3, then L={L i1 ,L i2 ,…,L ij ,…,L i9 The three equal parts after dividing M into equal parts are respectively And the statistical values ​​within each equal division are respectively The number of, where j takes values ​​from 1 to 9;

[0110] Step 6.3: Generate a statistical histogram of the key points based on the statistical values ​​of each equal division. The expression for the statistical histogram is:

[0111]

[0112] in, For key point p i The statistical histogram, the items in [] represent key points p. i The statistical values ​​of the kth division of the statistical histogram, where D is the set of descriptive points;

[0113] Step 6.4: Repeat steps 6.1 to 6.3 to calculate the statistical histograms of all key points, forming a statistical histogram of the component to be compared.

[0114] Equation (4) represents the current element L in L. ij Whether a key point falls within the kth division is determined by the proportion of context distances falling into each statistical histogram using equations (3) and (4). For all key points, the statistical method is as follows: for different key points, the same histogram group interval is used, i.e., the same M value is used (different M values ​​can also be selected for each key point). Therefore, in the implementation process, only the distance context of each key point needs to be calculated to form r statistical histograms. By merging the r statistical histograms according to the same group interval, the statistical histogram of the component to be compared can be obtained:

[0115]

[0116] Where H is the statistical histogram of the component to be compared, the items in [] represent the statistical values ​​of the kth division of the statistical histogram of the component to be compared, and r is the number of key points in the key point set.

[0117] Step 7: Identify the differences between the components by using the statistical histograms of the components to be compared in the V1.0 and V1.3 BIM models.

[0118] The result of feature description of the components is a statistical histogram. The problem of measuring the similarity of component models is transformed into a difference measurement of the two statistical histograms. Since the present invention aims to solve the problem of complete comparison of component models, the Euclidean distance can be directly selected to measure the distance between the two statistical histograms, and a threshold can be set for comparison.

[0119] Specifically, the shape distance of the components to be compared is calculated based on the statistical histograms of the components in the V1.0 and V1.3 BIM models. The specific calculation formula is as follows:

[0120]

[0121] Where e represents the shape distance, and the statistical histogram of component A to be compared in the V1.0 version BIM model. For key point p i The statistical histogram of the kth division of the histogram, the statistical histogram of component B to be compared in the V1.3 version BIM model.

[0122] The smaller the shape distance, the higher the similarity. (e is defined as...) θ The distance threshold is defined when e ≤ e θ In this case, there should be no difference between component A to be compared in the V1.0 version BIM model and component B to be compared in the V1.3 version BIM model; otherwise, there will be a difference. For example... Figure 3 The distance distribution map shown.

[0123] In this embodiment, e θ =10 -3 During surface subdivision, the similarity metric e = 1.7 * 10^6 -4 It can be determined that the component has not been modified in the two versions of the model; after the component is redesigned, the shape description changes, and the similarity measure is e = 0.019, which indicates that the component has been modified in the two versions of the model.

[0124] This invention solves the problem of identifying model differences between different versions of the same type of component, identifies subtle design differences of the same type of component, and can resist the influence of noise caused by surface subdivision and simplification. Based on the model mesh representation, fixed key points and random description points are selected to form a statistical histogram of the distance context description of the key points, and comparison is made based on the statistical histogram.

[0125] The problem with comparing component models using the above method is that it is invariant to translation, rotation, and scaling. This is a significant advantage for applications that only classify models; however, in the application scenario of this invention, model version changes are generally made on the original model, and the displacement, angle, and scaling of local components should all be identified.

[0126] Therefore, for two components A and B that are identical, further comparison is needed:

[0127] (1) Calculate the coordinate range P of components A and B. max and P min For the center point P of components A and B center =(P max +P min Compare ) / 2 to determine whether a translation has occurred.

[0128] Among them, P max and P min Let P be the maximum and minimum vertices of the AABB bounding box of component A or component B, respectively. P can be obtained by iterating through the vertex coordinates of component A or component B. max =(x max ,y max ,z max ) and P min =(x min ,y min ,z min ).

[0129] (2) Using the distance ratio d between the two components A and B A / d B To help measure whether scaling has occurred, the calculation formula is:

[0130]

[0131]

[0132] Where, d A Let d be the distance between the largest and smallest vertices of component A. B Let be the distance between the largest and smallest vertices of component B, and let dis(,) represent the distance function between the two points.

[0133] (3) Using the included angle between the vectors of the two components A and B To determine whether a rotation has occurred, the vector calculation formula is:

[0134] Vector of component A

[0135] Vector of component B

[0136] The above description only discloses specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or modifications that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying differences in components in different versions of BIM models, characterized in that, The method comprises the following steps: selecting a component to be compared from different versions of BIM models; extracting the contour line of the component to be compared in each version of BIM model; sampling the contour line to obtain a key point set of the contour line; sampling the surface of the component to be compared to obtain a description point set of the surface; calculating the logarithmic distance between each key point in the key point set and all points in the description point set; constructing a statistical histogram of the component to be compared according to the logarithmic distance; identifying the component difference according to the statistical histogram of the component to be compared in different versions of BIM models; The specific implementation process of extracting the contour line of the component to be compared in each version of BIM model is as follows: extracting edge information of the to-be-compared component cell by cell to obtain an edge information set wherein, is the i-th edge information, is a start point coordinate of the i-th edge, is an end point coordinate of the i-th edge, is a normal of the i-th edge corresponding cell, is the number of edges in the to-be-compared component; sorting the vertices of the component to be compared according to the priority order of X / Y / Z, and storing the edge lines associated with each vertex; de-duplicating the edge lines associated with each vertex to obtain remaining edge lines; splicing all remaining edge lines to obtain the contour line of the component to be compared.

2. The method of claim 1, wherein the method further comprises: The specific implementation process of de-duplicating the edge lines associated with each vertex is as follows: for any vertex, extracting non-duplicate edge lines associated with the vertex; calculating a normal included angle of the two repeated edge lines when , the two repeated edge lines are reserved; otherwise, the two repeated edge lines are removed; wherein, is an included angle threshold, ; composing remaining edge lines from non-duplicate edge lines and retained duplicate edge lines.

3. The method of claim 1, wherein the method further comprises: The specific implementation process of sampling the contour line is as follows: uniformly sampling along the contour line according to a set sampling interval to obtain the key point set; or sampling along the contour line according to a set sampling number to obtain the key point set.

4. The method of claim 1, wherein the method further comprises: The surface of the component to be compared is sampled by using a Poisson disk uniform sampling method.

5. The method of claim 1, wherein, The specific calculation formula of the logarithmic distance is as follows: ； wherein, is the key point on the contour line C and the logarithmic distance of the jth point in the set of points is described.

6. The method of claim 1, wherein the method further comprises: The construction method of the statistical histogram of the component to be compared is as follows: normalizing each log distance in the set of log distances of the key points to obtain a set of normalized distances; M-partitioning the normalized distance set, calculating the statistical value of each partition, and the specific calculation formula is as follows: ； ； wherein, is the log distance set of the key points is the statistical value of the kth quantile of the histogram of the log distance set of the key points, is the description of the point set, is the number of points in the point set, is the normalized distance in the normalized distance set, is the kth quantile after M quantification, ; forming the statistical histogram of the key point according to the statistical value of each partition, and the expression of the statistical histogram is as follows: ； wherein, is a key point of a statistical histogram, the entry in [ ] represents the statistical value of the k-th division of the statistical histogram of a key point is a point set;​ calculating the statistical histogram of all key points, and forming the statistical histogram of the component to be compared, and the specific expression is as follows: ； wherein, is a statistical histogram of the to-be-compared member, is the number of keypoints within the set of keypoints.

7. The method of claim 1, wherein the method further comprises: The specific implementation process of identifying the component difference according to the statistical histogram of the component to be compared in different versions of BIM models is as follows: calculating the shape distance of the component according to the statistical histogram of the component to be compared in different versions of BIM models, and the specific calculation formula is as follows: ； wherein, is the shape distance, wherein the statistical histogram of the component A to be compared in the version BIM model , , is the number of keypoints within the set of keypoints, is the set of keypoints is the statistical value of the logarithmic distance set at the k-th division of the histogram, is the statistical histogram of the component B to be compared in the version BIM model ; When there is no difference between the to-be-compared member A and the to-be-compared member B, otherwise there is a difference, wherein is a distance threshold value.

8. The method of claim 7, wherein the method further comprises: 。 9. The method of claim 1-8, wherein, The method further comprises: when there is no difference in the component, judging whether the center points of the two components to be compared are translated by calculating the coordinate ranges of the two components to be compared; and / or, judging whether scaling occurs by calculating the distance between the maximum vertex and the minimum vertex of the AABB bounding box of the two components to be compared; and / or, judging whether rotation occurs by calculating the vector angle of the two components to be compared.

10. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor executes the component difference identification method in different versions of BIM models according to any one of claims 1-9.

11. A non-transitory computer-readable storage medium, comprising: The non-transitory computer readable storage medium stores computer instructions which, when executed by at least one processor, cause the at least one processor to perform the method for identifying differences in components in different versions of BIM models according to any one of claims 1-9.

Images