阅读说明：本技术 一种建筑物三维模型语义化的方法及系统 (Building three-dimensional model semantization method and system ) 是由 刘俊伟 彭贵堂 王金兰 于 2021-09-23 设计创作，主要内容包括：本发明属于三维建筑物语义建模技术领域,提供了一种建筑物三维模型语义化的方法及系统,所述方法包括：S1、对每个建筑物三维模型进行预处理；S2、针对每个三维模型,分别计算得到三维模型的几何特征量；S3、基于三维模型的几何特征量,将三维模型各组成面自动识别为建筑物主要结构组件,并生成对应的语义描述信息；S4、对于建筑物主要结构组件,分别生成对应的附加语义描述信息,还生成对整个建筑物外观的语义描述信息；S5、构造建筑物三维模型群的完整语义网；S6、对上述语义网进行分析,进而得到对建筑物三维模型群的语义描述信息。通过本发明能够对建筑物三维模型进行自动的语义建模,提高建筑物三维模型语义化的准确性和智能化程度。(The invention belongs to the technical field of semantic modeling of three-dimensional buildings, and provides a building three-dimensional model semantization method and a system, wherein the method comprises the following steps: s1, preprocessing the three-dimensional model of each building; s2, respectively calculating the geometric characteristic quantity of each three-dimensional model to obtain the geometric characteristic quantity of the three-dimensional model; s3, automatically identifying each component surface of the three-dimensional model as a main structural component of the building based on the geometric characteristic quantity of the three-dimensional model, and generating corresponding semantic description information; s4, generating corresponding additional semantic description information for the main structural components of the building respectively, and generating semantic description information of the appearance of the whole building; s5, constructing a complete semantic net of the building three-dimensional model group; and S6, analyzing the semantic net to further obtain semantic description information of the building three-dimensional model group. The invention can automatically carry out semantic modeling on the building three-dimensional model and improve the semantic accuracy and the intelligent degree of the building three-dimensional model.)

1. A building three-dimensional model semantization method is characterized by comprising the following steps:

s1, preprocessing each building three-dimensional model, and converting geometric expressions in each building three-dimensional model into geographic expressions;

s2, respectively calculating the geometric characteristic quantity of the three-dimensional model aiming at each three-dimensional model, and describing the geometric characteristics of each component surface of the three-dimensional model;

s3, automatically identifying each component surface of the three-dimensional model into corresponding main structural components of the building, including wall surfaces, floors and roofs, based on the geometric characteristic quantity of the three-dimensional model, and generating semantic description information of different structural components;

s4, respectively generating corresponding additional semantic description information including orientation and height for each building main structure component, and simultaneously generating semantic description information including building floor area for the whole building appearance;

s5, taking the main structural components of each building, semantic description information and additional semantic description information corresponding to the main structural components, and the semantic description information of the whole building as nodes, taking topological structure relations among different buildings as edges, and constructing a complete semantic net of a three-dimensional model group of the building;

s6, analyzing the complete semantic net of the building three-dimensional model group to further obtain semantic description information of the building three-dimensional model group, wherein the semantic description information comprises semantic description of the number of buildings of the building group, semantic description of the floor area of the building group and semantic description of communication relation among the buildings in the building group.

2. The method for building three-dimensional model semantization according to claim 1, wherein in S1, each building three-dimensional model is preprocessed, specifically comprising the following steps:

s11, for any building three-dimensional model, converting the geometric expression of the building three-dimensional model which is infinitely approximated to the geometric shape by a triangulation network into the geographic expression of points, lines and/or surfaces;

and S12, for any surface in the building three-dimensional model, extracting the geographic coordinates corresponding to a plurality of corner points of the surface, and storing the geographic coordinates corresponding to the corner points as a coordinate string according to a certain sequence so as to express each surface in the building three-dimensional model.

3. The method according to claim 1, wherein the geometric feature quantities of the three-dimensional model are calculated in S2, respectively, for each three-dimensional model, and specifically the method includes extracting each geometric figure surface forming the appearance of the three-dimensional model from the three-dimensional model, sequentially extracting vertices of each geometric figure surface in a clockwise direction, and obtaining coordinates of each geometric figure surface in the coordinate system, and calculating a unit normal vector of each geometric figure surface in the coordinate system, wherein the feature quantities of each geometric figure surface include vertex coordinate data of the geometric figure surface and unit normal vector data of the geometric figure surface, and the feature quantities of all the geometric figure surfaces form the geometric feature quantities of the three-dimensional model.

4. The method for semanticizing the three-dimensional model of building according to claim 1, wherein in S3, based on the geometric feature quantities of the three-dimensional model, the components of the three-dimensional model are automatically identified as the corresponding main structural components of the building, comprising the following steps:

s31, in the coordinate system of the three-dimensional model, designating the plane determined by the X coordinate axis and the Y coordinate axis as a standard plane, and calculating unit normal vector data on the standard plane;

s32, respectively calculating the cosine values of included angles between the unit normal vectors of the geometric figure surfaces forming the appearance of the three-dimensional model and the unit normal vectors of the standard plane according to the geometric characteristic quantity of the three-dimensional model, and further obtaining the numerical values of the included angles between the geometric figure surfaces forming the appearance of the three-dimensional model and the standard plane, wherein the specific calculation process is described by the following formula:，

wherein the content of the first and second substances,is a unit normal vector of the geometric figure surface,the number of geometric figure surfaces forming the appearance of the three-dimensional model,is a unit normal vector on a standard plane,is the cosine value of the included angle between the unit normal vector of each geometric figure surface and the unit normal vector of the standard plane,the included angle value between each geometric figure surface and the standard plane is obtained;

and S33, identifying each geometric figure surface as a corresponding building main structural component based on the included angle value between each geometric figure surface forming the appearance of the three-dimensional model and the standard plane.

5. The method for building three-dimensional model semantization according to claim 4, wherein in S33, based on the included angle value between each geometric figure surface and the standard plane, which form the appearance of the three-dimensional model, each geometric figure surface is identified as a corresponding building main structural component, and the method specifically comprises the following steps:

s331, clustering numerical values of included angles between each geometric figure surface forming the appearance of the three-dimensional model and a standard plane, and dividing each geometric figure surface forming the appearance of the three-dimensional model into different categories based on clustering results of the numerical values of the included angles;

s332, according to different categories into which the geometric figure surfaces forming the appearance of the three-dimensional model are divided, preliminarily identifying the geometric figure surfaces as main structural components of the building, wherein the categories respectively comprise a wall category, a floor category and a roof category;

s333, respectively acquiring texture mapping images corresponding to different geometric mapping surfaces for each geometric mapping surface contained in the wall surface category, and performing similarity matching on the characteristics of the texture mapping images and the characteristics of the texture mapping images of the building wall surface, door and window pre-stored in the system, so as to further identify each geometric mapping surface as the wall surface category, the door category and the window category.

6. The method of claim 1, wherein in step S4, for each building main structural component, generating corresponding additional semantic description information, and generating semantic description information for the whole building appearance, specifically including generating, for each geometry plane composing the three-dimensional model appearance, additional semantic description information about its orientation based on the direction of unit normal vector of the geometry plane, generating semantic description information about its height by calculating corresponding segment length of the geometry plane on Z coordinate axis in the coordinate system, and generating semantic description information about building area by calculating projected area of the three-dimensional model of the building on X0Y plane in the coordinate system.

7. The method according to claim 1, wherein the topological structure relationship between the different buildings in S5 includes a connection relationship, an adjacent relationship, and a separation relationship between the buildings, the connection relationship further includes a geometric structure and a connection direction of a connection part between every two buildings, the adjacent relationship describes a position relationship between every two buildings that are directly adjacent and not connected, the adjacent relationship further includes an adjacent direction, the separation relationship describes a position relationship between every two buildings that are not connected and not adjacent, and the relative position relationship between two buildings.

8. The method as claimed in claim 1, wherein the step S6 of analyzing the complete semantic web of the three-dimensional building model group to obtain semantic description information of the three-dimensional building model group specifically includes traversing each node and each edge of the semantic web, generating semantic description information of the number of buildings of the building group by counting the number of nodes of the semantic web, generating semantic description of the floor area of the building group and description of the connection relationship between the buildings in the building group according to the topological structure relationship between the different buildings described by the edge of the semantic web, and combining the semantic description information on the nodes of the semantic web.

9. A system for semanticizing a three-dimensional model of a building, for implementing the method according to any one of claims 1 to 8, comprising the following modules:

the model preprocessing module is used for preprocessing each building three-dimensional model and converting geometric expressions in each building three-dimensional model into geographic expressions;

the characteristic extraction module is used for respectively calculating geometric characteristic quantity of each three-dimensional model aiming at each three-dimensional model, is used for describing the geometric characteristics of each component surface of the three-dimensional model, and specifically comprises the steps of calculating vertex coordinate data of each geometric figure surface forming the appearance of the three-dimensional model and unit normal vector data of each geometric figure surface;

the first semantic generation module is used for automatically identifying main structural components of each three-dimensional model appearance, including wall surfaces, floors, roofs, windows and doors, based on the geometric characteristic quantity of each three-dimensional model, generating semantic description information of different structural components, respectively generating additional semantic description information including orientation and height aiming at the main structural components of each three-dimensional model appearance, and simultaneously generating semantic description information of the whole building appearance, including the floor area of the building;

and the second semantic generation module is used for constructing a complete semantic web of the building three-dimensional model group based on the main structural components of the three-dimensional model appearance and the corresponding semantic description information, performing traversal analysis on the semantic web, further obtaining the semantic description information of the building three-dimensional model group, and realizing automatic semantic annotation on the traditional building three-dimensional model.

Technical Field

The invention belongs to the technical field of semantic modeling of three-dimensional buildings, and particularly relates to a building three-dimensional model semantization method and system.

Background

With the continuous maturity of three-dimensional modeling software and the rapid development of three-dimensional scanning equipment, the work of acquiring three-dimensional models becomes easier and easier, and at the same time, the development of three-dimensional model management and application is accelerated, especially in the field of three-dimensional buildings, a large number of three-dimensional building models are accumulated, the traditional three-dimensional building models lack semantic description information, and different models are in a logically mutually dispersed relationship and are not connected with each other, so that only the structure, the appearance and the like of the three-dimensional models can be checked, but related analysis and other application work can not be carried out based on the three-dimensional models, the three-dimensional model semantization is taken as a key technology of the three-dimensional model management and application, and in fact, according to the characteristics of the three-dimensional models, a process of establishing a corresponding relationship between specific text semantics and the three-dimensional models by a certain method is adopted in the prior art, the corresponding relationship between the three-dimensional model and the semantics thereof is generally established in a manner of manually checking and labeling the three-dimensional model, but when the number of the three-dimensional model is large, the problems of low efficiency and insufficient intelligence of the model semantics often occur.

Disclosure of Invention

Aiming at the technical problems, the invention provides a building three-dimensional model semantization method and a building three-dimensional model semantization system, which are used for realizing automatic semantic annotation on a traditional building three-dimensional model and obtaining the semantization building three-dimensional model, so that the traditional building three-dimensional model built by modeling software such as 3DMAX, CAD and the like can exert the utility to the greatest extent, and the semantization three-dimensional model is used for designing, planning, analyzing and implementing all aspects in a building group, thereby meeting the requirements of managing and analyzing buildings.

In order to achieve the above object, the present invention provides a method for building three-dimensional model semantization, which comprises the following steps:

preprocessing each building three-dimensional model, and converting geometric expressions in each building three-dimensional model into geographic expressions;

step two, respectively calculating geometric characteristic quantities of the three-dimensional models aiming at each three-dimensional model, and describing the geometric characteristics of each component surface of the three-dimensional models;

automatically identifying each component surface of the three-dimensional model into corresponding main structural components of the building, including wall surfaces, floors and roofs, based on the geometric characteristic quantity of the three-dimensional model, and generating semantic description information of different structural components;

step four, generating corresponding additional semantic description information including orientation and height for each building main structure component, and generating semantic description information of the whole building appearance including the building floor area;

fifthly, constructing a complete semantic net of a building three-dimensional model group by taking the main structural components of each building, semantic description information corresponding to the main structural components, additional semantic description information and semantic description information of the whole building as nodes and topological structure relations among different buildings as edges;

analyzing the complete semantic net of the building three-dimensional model group to further obtain semantic description information of the building three-dimensional model group, wherein the semantic description information comprises semantic description of the number of buildings of the building group, semantic description of the floor area of the building group and semantic description of communication relation among the buildings in the building group.

As a preferred technical solution of the present invention, in step three, based on the geometric feature quantity of the three-dimensional model, each component plane of the three-dimensional model is automatically identified as a corresponding building main structural component, which specifically includes the following steps:

firstly, in a coordinate system of the three-dimensional model, a plane determined by an X coordinate axis and a Y coordinate axis is designated as a standard plane, and unit normal vector data on the standard plane is calculated;

secondly, respectively calculating the cosine values of included angles between the unit normal vectors of the geometric figure surfaces forming the appearance of the three-dimensional model and the unit normal vectors of the standard plane according to the geometric characteristic quantity of the three-dimensional model, and further obtaining the numerical values of the included angles between the geometric figure surfaces forming the appearance of the three-dimensional model and the standard plane, wherein the specific calculation process is described by the following formula:

，

wherein the content of the first and second substances,is a unit normal vector of the geometric figure surface,the number of geometric figure surfaces forming the appearance of the three-dimensional model,is a unit normal vector on a standard plane,is the cosine value of the included angle between the unit normal vector of each geometric figure surface and the unit normal vector of the standard plane,the included angle value between each geometric figure surface and the standard plane is obtained;

and thirdly, identifying each geometric figure surface as a corresponding building main structure component based on the numerical value of an included angle between each geometric figure surface forming the appearance of the three-dimensional model and a standard plane.

Compared with the prior art, the invention has the following beneficial effects:

the invention provides a building three-dimensional model semantization method and a system thereof, aiming at the traditional building three-dimensional model, firstly, the model is standardized and preprocessed, then, the geometric characteristic quantity forming the appearance of the three-dimensional model is calculated to formally describe the structural characteristic of the model, then, based on the geometric characteristic quantity, the main structural component of the appearance of the three-dimensional model is automatically identified, and corresponding semantic description information is generated, in order to obtain the semantic information about a building group from the appearance structure and the semantic information of each building, a complete semantic net of the building three-dimensional model group is also constructed, and the semantic information about the building group is generated by analyzing the complete semantic net The method has the advantages of high efficiency and good accuracy, and the intelligent degree of the semantic work of the model is also improved.

Drawings

FIG. 1 is a flowchart illustrating the overall steps of a method for building three-dimensional model semantics of the present invention;

FIG. 2 is a flowchart illustrating the overall steps of the present invention for pre-processing each building three-dimensional model;

FIG. 3 is a flowchart illustrating the overall steps of the major structural components of the present invention for automatically identifying the appearance of a three-dimensional model;

FIG. 4 is a flowchart illustrating the detailed steps of identifying the primary structural components of the three-dimensional model appearance based on the included angle values in accordance with the present invention;

FIG. 5 is a block diagram of a system for building three-dimensional model semantics.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

It will be understood that, as used herein, the terms "first," "second," and the like may be used herein to describe various elements, but these elements are not limited by these terms unless otherwise specified. These terms are only used to distinguish one element from another. For example, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script, without departing from the scope of the present application.

Referring to fig. 1, the present invention provides a method for semanticizing a three-dimensional model of a building, which is implemented by the following steps:

preprocessing each building three-dimensional model, and converting geometric expressions in each building three-dimensional model into geographic expressions;

step two, respectively calculating geometric characteristic quantities of the three-dimensional models aiming at each three-dimensional model, and describing the geometric characteristics of each component surface of the three-dimensional models;

automatically identifying each component surface of the three-dimensional model into corresponding main structural components of the building, including wall surfaces, floors and roofs, based on the geometric characteristic quantity of the three-dimensional model, and generating semantic description information of different structural components;

step four, generating corresponding additional semantic description information including orientation and height for each building main structure component, and generating semantic description information of the whole building appearance including the building floor area;

fifthly, constructing a complete semantic net of a building three-dimensional model group by taking the main structural components of each building, semantic description information corresponding to the main structural components, additional semantic description information and semantic description information of the whole building as nodes and topological structure relations among different buildings as edges;

analyzing the complete semantic net of the building three-dimensional model group to further obtain semantic description information of the building three-dimensional model group, wherein the semantic description information comprises semantic description of the number of buildings of the building group, semantic description of the floor area of the building group and semantic description of communication relation among the buildings in the building group.

Further, in step one, the three-dimensional building model in the present embodiment is constructed by the 3DMAX model, and is constructed according to actual geographic coordinates, including orientation, scale, and the like, in the initial stage of the construction of the 3DMAX model. The 3DMAX building three-dimensional models are all triangular surfaces, namely, one wall is formed by at least two triangular surfaces. Therefore, the coordinate system of the three-dimensional building models needs to be preprocessed to convert the geometric expression in each three-dimensional building model into a geographic expression.

Specifically, referring to fig. 2, the preprocessing is performed on each building three-dimensional model in the step one, and the method further includes the following steps:

the method comprises the following steps that firstly, for any building three-dimensional model, the geometric expression which is infinitely approximated to the geometric shape by a triangulation network in the building three-dimensional model is converted into the geographic expression of points, lines and/or surfaces;

and secondly, extracting the geographic coordinates corresponding to a plurality of corner points of any surface in the building three-dimensional model, and storing the geographic coordinates corresponding to the corner points as a coordinate string according to a certain sequence so as to express each surface in the building three-dimensional model.

That is, for each face in each building three-dimensional model, firstly selecting adjacent triangles with parallel normal vectors to carry out triangle network merging, merging the triangle networks by removing the shared edge of the two adjacent triangles, judging whether the vertexes are in the same plane after merging, if the vertexes are in the same plane, continuing merging until no triangle is merged, thereby realizing merging the triangle networks into the face. Further, after extracting the geographic coordinates of each face corner point, storing the geographic coordinates as a coordinate string in a certain sequence, namely, for obtaining a polygon with continuous and closed vertices, the polygon is described by sequential point sequence combination, wherein P = { P1, P2, p3.. Taking a wall surface as an example, a triangular net contained in the wall surface is merged into a surface, and then the geographic coordinates of four intersection points of the wall surface are extracted to describe the wall surface. The processing procedure of the embodiment can be beneficial to the subsequent extraction of the feature vectors, and the structure and the number of the geometric objects of the three-dimensional model are not modified.

Furthermore, in the second step, the geometric characteristic quantity of each three-dimensional model is calculated and obtained for each three-dimensional model, and the calculation process comprises the steps of firstly extracting each geometric figure surface forming the appearance of the three-dimensional model on the three-dimensional model, and then sequentially extracting the geometric figure surfaces in the clockwise directionTaking out the top point of each geometric figure surface, obtaining the coordinate of each geometric figure surface in the coordinate system, then calculating the unit normal vector of each geometric figure surface in the coordinate system, finally, the characteristic quantity of each geometric figure surface is composed of the top point coordinate data of the geometric figure surface and the unit normal vector data of the geometric figure surface, and the data organization form of the characteristic quantity of the geometric figure surface is thatAnd the characteristic quantities of all the geometric figure surfaces form the geometric characteristic quantity of the three-dimensional model.

Specifically, the geometric features of the geometric figure surfaces forming the appearance of the three-dimensional model can be described in a formalized manner by acquiring the geometric feature quantity of the three-dimensional model, so that in the subsequent steps, the geometric figure surfaces forming the appearance of the three-dimensional model can be automatically identified into different building appearance structural components based on the geometric feature quantity of the three-dimensional model, and semantic description information of the building appearance structural components is generated.

Furthermore, the geometric characteristic quantity of each three-dimensional model appearance is obtained in the third step, that is, each geometric figure surface and unit normal vector data thereof composing the three-dimensional model are obtained, in order to identify each geometric figure surface as different building appearance structure components, including wall surface, floor, roof, door, window, etc., the geometric characteristic quantity of each three-dimensional model needs to be analyzed, considering that in actual life, the included angle values of different building appearance structure components and the building bottom surface are obviously different, for example, the included angle value of the wall surface, the door, the window and the building bottom surface is usually ninety degrees, the included angle value of the floor and the building bottom surface is usually zero degrees, and the included angle value of the roof and the building bottom surface is usually between zero degrees and ninety degrees, therefore, according to the different included angle values of each geometric figure surface and the building three-dimensional model bottom surface, the geometry map can be initially identified as a different building facade element.

Specifically, referring to fig. 3, in step three, based on the geometric feature quantity of the three-dimensional model, automatically identifying each component plane of the three-dimensional model as a corresponding building main structural component specifically includes the following steps:

firstly, in a coordinate system of the three-dimensional model, a plane determined by an X coordinate axis and a Y coordinate axis is designated as a standard plane, and unit normal vector data on the standard plane is calculated;

secondly, respectively calculating the cosine values of included angles between the unit normal vectors of the geometric figure surfaces forming the appearance of the three-dimensional model and the unit normal vectors of the standard plane according to the geometric characteristic quantity of the three-dimensional model, and further obtaining the numerical values of the included angles between the geometric figure surfaces forming the appearance of the three-dimensional model and the standard plane;

and thirdly, identifying each geometric figure surface as a corresponding building main structure component based on the numerical value of an included angle between each geometric figure surface forming the appearance of the three-dimensional model and a standard plane.

For the sake of understanding, the specific calculation process in the second step is described by the following formula:

，

wherein the content of the first and second substances,is a unit normal vector of the geometric figure surface,the number of geometric figure surfaces forming the appearance of the three-dimensional model,is a unit normal vector on a standard plane,unit method for each geometric figure surfaceThe cosine value of the included angle between the vector and the unit normal vector of the standard plane,and (4) representing the included angle value between each geometric figure surface and the standard plane.

Further, referring to fig. 4, in the third step, based on the numerical value of the included angle between each geometric figure surface forming the appearance of the three-dimensional model and the standard plane, each geometric figure surface is identified as a corresponding building main structural component, and the method specifically includes the following steps:

firstly, clustering numerical values of included angles between each geometric figure surface forming the appearance of the three-dimensional model and a standard plane, and dividing each geometric figure surface forming the appearance of the three-dimensional model into different categories based on clustering results of the numerical values of the included angles;

secondly, according to different categories into which various geometric figure surfaces forming the appearance of the three-dimensional model are divided, preliminarily identifying the geometric figure surfaces as main structural components of the building, wherein the categories respectively comprise a wall category, a floor category and a roof category;

and thirdly, respectively acquiring texture mapping images corresponding to different geometric figure surfaces for each geometric figure surface contained in the wall surface category, and performing similarity matching on the characteristics of the texture mapping images and the characteristics of the texture mapping images of the building wall surface, door and window pre-stored in the system, thereby further identifying each geometric figure surface as the wall surface category, the door category and the window category.

Specifically, in the third step, the geometric figure surfaces belonging to the wall surface category need to be further analyzed, so that the geometric figure surfaces are continuously and respectively classified into the wall surface category, the door category, and the window category. The analysis processing process comprises the steps of obtaining texture mapping images corresponding to different geometric figure surfaces, obtaining feature data of the texture mapping images through a feature extraction algorithm, for example, an image texture feature extraction method based on a co-occurrence matrix is used, the method combines frequency domain statistical features and spatial distribution characteristics of the images, firstly extracts local frequency domain information of the images through wavelet transformation, then combines overall structure features of the images to construct a wavelet gray level co-occurrence matrix for extracting image texture features, then carries out similarity matching on the features of the texture mapping images and the features of the texture mapping images of building doors and windows stored in the system in advance, and respectively identifies the geometric figure surfaces contained in wall categories as wall surfaces, doors and windows in corresponding building structure components when the similarity is greater than a threshold value specified by the system.

Further, in the fourth step, for each main structural component of the three-dimensional model appearance, generating corresponding additional semantic description information and generating semantic description information of the whole building appearance, specifically, generating additional semantic description information about the orientation of each geometric figure surface forming the three-dimensional model appearance based on the direction of the unit normal vector of the geometric figure surface, for example, the direction of the unit normal vector of the geometric figure surface of the wall surface a is the positive direction of the X axis, the orientation of the wall surface a in practical situations is south, and semantic description information about the height of the wall surface a is generated by calculating the corresponding line segment length of the geometric figure plane on the Z coordinate axis in the coordinate system, and the semantic description information about the floor area of the building is generated by calculating the projection area of the three-dimensional model of the building on the X0Y plane in the coordinate system.

Specifically, before generating semantic description information about height and semantic description information about floor area of a building, first, a corresponding distance of a geometric figure plane on a Z coordinate axis in a coordinate system and a projection area of a three-dimensional model of the building in the coordinate system are respectively obtained, and then, according to a scale relationship between the three-dimensional model determined in the step one and an actual building, numerical value conversion is performed on the distance data and the area data, and finally, a corresponding height numerical value and an area numerical value of the three-dimensional model of the building under an actual condition are obtained.

Further, in the fifth step, the main structural components forming the appearance of each building, the semantic description information and the additional semantic description information corresponding to the main structural components, and the semantic description information about the whole appearance of the building are acquired, in order to obtain the semantic information about the building group according to the data information about the appearance of the building, the main structural components of the appearance of each building and the semantic description information thereof are respectively used as node data, and the topological structure relationship among different buildings is used as an edge to construct a complete semantic net of the three-dimensional model group of the building, the topological structure relationship includes the connection relationship, the adjacent relationship and the separation relationship among the buildings, the connection relationship specifically includes the geometric structure and the connection direction of the connection part between every two buildings, the adjacent relationship describes the position relationship between every two buildings which is directly adjacent and not connected, the method specifically comprises the adjacent direction, the position relation that every two buildings are not connected and not adjacent is described by the separation relation, and the relative position relation of the two buildings is specifically further included.

Specifically, the above topological structure relationship is described by way of example, and the connection relationship between two buildings is illustrated by way of example, building a and building B are communicated through a corridor with a cubic geometry, and the position of building a is fixed, and building B is in the direction of the south of building a; the adjacent relationship between two buildings is, for example, that there is no other building between building a and building C, and the position of building a is fixed, building C is in the direction forty degrees north of building a; for example, the above example is used to explain the topological relationship, but not to limit the topological relationship, in the direction of the building a, and the building D is located in the west of the building a, and the semantic information about the building group can be obtained by analyzing the complete semantic web of the building three-dimensional model group.

Further, in the sixth step, the complete semantic net of the building three-dimensional model group is analyzed to generate semantic description information of the building three-dimensional model group, specifically including traversing each node and each edge of the semantic net, generating semantic description information of the number of buildings of the building group by counting the number of nodes of the semantic net, generating the semantic description information of the building group according to the topological structure relationship between different buildings described by the edges of the semantic net, and combining the semantic description information on the nodes of the semantic net, specifically including the total floor area of the building group, the communication relationship between the buildings in the building group, and the like. The total floor area of the building group is the sum of the floor areas of the buildings recorded on all the nodes, and the communication relation among the buildings is obtained by analyzing the edges of the connecting nodes.

The semantic web is traversed to acquire, analyze and summarize structural component data and semantic information forming the appearance of the building and topological relations among different buildings, so that semantic information about a building group can be acquired, and related analysis work can be further performed based on a semantic three-dimensional model, so that the traditional three-dimensional model of the building can play a role to the greatest extent.

Referring to fig. 5, the present invention further provides a building three-dimensional model semantization system, which is used for implementing the building three-dimensional model semantization method described in the foregoing, and specifically includes the following modules:

the model preprocessing module is used for preprocessing each building three-dimensional model and converting geometric expressions in each building three-dimensional model into geographic expressions;

the characteristic extraction module is used for respectively calculating geometric characteristic quantity of each three-dimensional model aiming at each three-dimensional model, is used for describing the geometric characteristics of each component surface of the three-dimensional model, and specifically comprises the steps of calculating vertex coordinate data of each geometric figure surface forming the appearance of the three-dimensional model and unit normal vector data of each geometric figure surface;

the first semantic generation module is used for automatically identifying main structural components of each three-dimensional model appearance, including wall surfaces, floors, roofs, windows and doors, based on the geometric characteristic quantity of each three-dimensional model, generating semantic description information of different structural components, respectively generating additional semantic description information including orientation and height aiming at the main structural components of each three-dimensional model appearance, and simultaneously generating semantic description information of the whole building appearance, including the floor area of the building;

and the second semantic generation module is used for constructing a complete semantic web of the building three-dimensional model group based on the main structural components of the three-dimensional model appearance and the corresponding semantic description information, performing traversal analysis on the semantic web, further obtaining the semantic description information of the building three-dimensional model group, and realizing automatic semantic annotation on the traditional building three-dimensional model.

The detailed functional and technical details of the modules may refer to the description of the method embodiment, and are not described herein again.

It should be understood that, although the steps in the flowcharts of the embodiments of the present invention are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a portion of the steps in various embodiments may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included therein.