CN116882032B

CN116882032B - Building design atlas digitizing and visualizing method and device for applying same

Info

Publication number: CN116882032B
Application number: CN202311127701.7A
Authority: CN
Inventors: 赵广坡; 刘宜丰; 赵一静; 黄扬; 刘冠军; 白蜀珺
Original assignee: China Southwest Architectural Design and Research Institute Co Ltd
Current assignee: China Southwest Architectural Design and Research Institute Co Ltd
Priority date: 2023-09-04
Filing date: 2023-09-04
Publication date: 2023-11-17
Anticipated expiration: 2043-09-04
Also published as: CN116882032A

Abstract

The invention discloses a method and a device for digitizing and visualizing a building design atlas and applying the same, wherein the method comprises the following steps: extracting the graphic identification information and the graphic parameter information of each graphic block from the paper building design drawing set; establishing a first atlas knowledge atlas corresponding to the paper architectural design atlas according to the extracted graphic identification, parameter information and logic relation thereof; establishing an atlas BIM model based on the graphic parameter information of each image block, reading the annotation data of the user about the atlas BIM model, and determining the attribute information of the annotation data; and retrieving corresponding entity nodes in the first atlas knowledge graph according to the graphic identification information of each atlas BIM model, and storing attribute information to the labeling nodes associated with the retrieved nodes to obtain a second atlas knowledge graph. According to the invention, the second atlas knowledge graph related to the BIM entity is constructed, so that the two-dimensional atlas information is digitally stored and structurally expressed, and a good foundation is laid for the expansion application of the atlas knowledge.

Description

Building design atlas digitizing and visualizing method and device for applying same

Technical Field

The invention relates to the technical field of building design atlas processing, in particular to a method and a device for digitizing and visualizing a building design atlas and applying the method and the device.

Background

The standard design drawing set (hereinafter referred to as "drawing set") of the building engineering is a drawing which is uniformly compiled according to standardized requirements for engineering construction buildings, structures, engineering facilities, devices and the like in China, industry and places, and is one of basic references of the building engineering design. The standard atlas can be directly applied to the architectural engineering design drawing, only the reference information is needed to be marked, repeated drawing work of a designer is reduced, and some mandatory design regulations, specifications and the like are restrained in the design process in a standard atlas reference mode, so that engineering quality is guaranteed. Meanwhile, the standard atlas can also penetrate through the whole industrial chain from design to construction through the convention of the standard practice, so that the floor application is realized.

The main content of the existing atlas consists of two-dimensional illustration and dimension marking, and is supplemented with text/form description to supplement three-dimensional dimension information, building material information and the like so as to express design information. However, for three-dimensional stereo architecture, the existing atlas still has more contents such as reinforcing steel bar detail structure, complex node connection mode and the like, and can only be understood by relying on the space imagination capability and practical working experience of a user, so that designers need to spend a great deal of time and effort to learn the atlas, understanding deviation easily occurs, and different designers can understand different from one designer to another. In addition, after the atlas is quoted to the formal design drawing, the subsequent cost and constructors also need to consult the atlas for relevant quotes to finish respective work, so that a great deal of learning time is consumed, and meanwhile, the accuracy of the subsequent cost and construction is possibly influenced due to insufficient understanding depth.

Therefore, the existing atlas has complicated information, lacks structural expression, is unfavorable for the retrieval, extraction and reuse of the atlas knowledge, and a designer can only select a proper atlas according to design experience and does not fully exert the use value of the atlas.

Disclosure of Invention

The invention aims to solve the problems that the existing building engineering standard design atlas is complicated in information and lacks of structural expression, is unfavorable for atlas knowledge retrieval, extraction and reuse and has low use value, and provides a building design atlas digitizing and visualizing method and device for applying the same.

In order to achieve the above object, the present invention provides the following technical solutions:

a method of digitizing a building design atlas, comprising:

extracting the graphic identification information and the graphic parameter information of each graphic block from the paper building design drawing set;

establishing a first atlas knowledge atlas corresponding to the paper architectural design atlas according to the extracted graphic identification information, graphic parameter information and the logic relation between the extracted information;

Establishing an atlas BIM model based on the graphic parameter information of each image block, reading the annotation data of a user about the atlas BIM model, and determining the attribute information of the annotation data;

and searching corresponding entity nodes in the first atlas knowledge graph according to the graph identification information corresponding to each atlas BIM model, adding labeling nodes according to the labeling data on the basis of the entity nodes, and storing attribute information of the labeling data to obtain a second atlas knowledge graph, wherein the second atlas knowledge graph and the atlas BIM model are used for realizing the digitization of the paper architectural design atlas.

According to a specific embodiment, in the above method for digitizing a building design atlas, the establishing a first atlas knowledge atlas corresponding to the paper building design atlas according to the extracted graphic identification information, the graphic parameter information and the logical relationship between the extracted information includes:

defining a atlas knowledge atlas body in advance according to the logic relation between the graphic identification information and the graphic parameter information;

storing the extracted graph identification information and the graph parameter information into a predefined atlas knowledge atlas body;

Wherein, the atlas knowledge atlas ontology includes: the method comprises the steps of establishing knowledge links among five class levels through logical relations.

According to a specific embodiment, in the above building design atlas digitizing method, the graphic identification information includes: specialty code, atlas number, page number, atlas number;

the graphic parameter information includes: geometric information and non-geometric information, the geometric information comprising: text labeling, dimension labeling, profile information and geometric construction, wherein the non-geometric information comprises: relational annotation, model annotation and universal text annotation.

According to a specific embodiment, in the above method for digitizing a building design atlas, the establishing an atlas BIM model based on the graphic parameter information of each tile includes:

acquiring a first coordinate system corresponding to a current image block and a second coordinate system corresponding to BIM software, and determining a coordinate conversion relation between the first coordinate system and the second coordinate system;

and establishing the atlas BIM model according to the geometric information, the relation type annotation and the coordinate conversion relation.

According to a specific embodiment, in the above method for digitizing a building design atlas, the attribute information of the labeling data includes: labeling data category, data position and data content;

The labeling data category comprises: relational annotation, text marking, dimension marking and profile information.

According to a specific embodiment, in the above method for digitizing a building design atlas, on the basis of the entity node, adding a labeling node according to the labeling data, where the labeling node is used to store attribute information of the labeling data, the method includes:

on the basis of the graph nodes retrieved based on the graph identification information, adding marking nodes in the first graph set knowledge graph according to the marking data types, wherein the marking nodes comprise character marking nodes, dimension marking nodes and profile information nodes; storing the data location and the data content in the added annotation node;

and finding out page nodes or graph nodes corresponding to the profile information nodes based on the relational annotation, and establishing an association relationship between the two nodes through a logical association symbol so as to store the relational annotation.

In another aspect of the present invention, there is provided a method of building design atlas visualization, the method comprising:

the second knowledge graph of each professional paper building design atlas is established in advance by adopting the building design atlas digitizing method;

Acquiring an atlas visual identifier captured by a user through shooting equipment, and retrieving to obtain a corresponding page node in the second knowledge graph; the atlas visualization identifier includes: an atlas number and page number;

and reading a plurality of graph node graph numbers contained in the page node based on the page node, determining a plurality of graph identification information, reading a corresponding atlas BIM model based on the graph identification information, carrying out light loading on the read atlas BIM model, and presenting the loaded light atlas BIM model and the corresponding knowledge graph node information to a user.

According to a specific embodiment, in the building design atlas visualization method, the lightweight loading of the read atlas BIM model includes:

performing format conversion on the atlas BIM model, and converting the atlas BIM model into a model in an FBX format;

and adjusting the position and angle of the model in the FBX format so that the position and angle of the loaded model are the same as those of the original model.

According to a specific embodiment, in the building design atlas visualization method, the acquiring the atlas visualization identifier captured by the user through the shooting device includes:

when the user captures and selects the atlas visual identification through the shooting device, the atlas name and the page number selected by the user are acquired based on the deep learning technology and the OCR technology.

According to a specific embodiment, in the building design atlas visualization method, the method further includes: information reloading flow after the atlas BIM model is loaded in a lightweight mode;

the information reloading process comprises the following steps: judging whether attribute information of the labeling data is stored in the corresponding graph node or not through the second knowledge graph retrieval, if yes, traversing the attribute information of the labeling data, wherein the attribute information of the labeling data comprises the following steps: labeling data category, data position and data content; the labeling data category comprises: relational annotation, text marking, dimension marking and profile information;

and sequentially reloading the data content at the position corresponding to the BIM of the lightweight atlas according to the type and the position of the marked data through a thread js code library.

According to a specific embodiment, in the building design atlas visualization method, when the corresponding page node is found in the second knowledge graph, and there is a section labeling node associated with the corresponding page node, the method further includes:

the attribute information of the section labeling node is read, and the position and the content of the section labeling are determined based on the attribute information;

Creating a plane, the position of the created plane coincides with the position of the section, dividing the lightweight atlas BIM model into two parts through the plane,

and acquiring a hiding instruction of a user on one side of the plane relative to the light-weight atlas BIM model, hiding a part of the model based on the hiding instruction, and presenting the section.

In another aspect of the present invention, there is provided a method for applying a building design atlas, comprising:

by adopting the building design atlas visualization method, a light atlas BIM model and corresponding knowledge graph node information are loaded;

learning, teaching application of the atlas knowledge is implemented based on the lightweight atlas BIM model, and/or,

and based on the knowledge graph node information retrieval, acquiring atlas knowledge, downloading a BIM model required by design, and carrying out rapid forward design.

In another aspect of the present invention, there is provided a building design atlas digitizing apparatus, the apparatus comprising:

the extraction unit is used for extracting the graph identification information and the graph parameter information of each graph block from the paper building design graph set;

the first processing unit is used for storing the extracted graphic identification information and the graphic parameter information of each graphic block into a predefined atlas knowledge atlas body and establishing a first atlas knowledge atlas corresponding to the paper architectural design atlas;

The second processing unit is used for establishing an atlas BIM model based on the graphic parameter information of each image block, reading the annotation data of a user about the atlas BIM model and determining the attribute information of the annotation data;

and the third processing unit is used for retrieving corresponding entity nodes in the first atlas knowledge graph according to the graph identification information corresponding to each atlas BIM model, adding labeling nodes according to the labeling data on the basis of the entity nodes, storing attribute information of the labeling data to obtain a second atlas knowledge graph, and realizing the digitization of the paper architectural design atlas through the second atlas knowledge graph and the BIM model.

According to a specific embodiment, in the architecture design atlas digitizing apparatus, the attribute information of the labeling data includes: labeling data category, data position and data content;

the labeling data category comprises: relational annotation, text marking, dimension marking and profile information;

the third processing unit is configured to: on the basis of the retrieved graph nodes, adding annotation nodes in the first graph set knowledge graph according to the annotation data types, wherein the annotation nodes comprise character annotation nodes, dimension annotation nodes and profile information nodes; storing the data location and the data content in the added annotation node;

And finding out page nodes corresponding to the profile information nodes based on the relational annotation, and establishing an association relationship between the two nodes through a logical association symbol so as to store the relational annotation.

In another aspect of the present invention, there is provided a building design atlas visualization apparatus including:

the storage unit is used for establishing a second knowledge graph of each professional paper building design atlas in advance by adopting the building design atlas digitizing method;

the acquisition unit is used for acquiring an atlas visual identifier captured by a user through the shooting equipment and retrieving the corresponding page node in the second knowledge graph; the atlas visualization identifier includes: an atlas number and page number;

and the visual loading unit is used for reading a plurality of graph node graph numbers contained in the page node based on the page node, determining a plurality of graph identification information, reading a corresponding atlas BIM model based on the plurality of graph identification information, carrying out light loading on the read atlas BIM model, and presenting the loaded light atlas BIM model and the corresponding knowledge graph node information to a user.

According to a specific embodiment, in the architectural design atlas visualization apparatus described above, the visualization loading unit is further configured to: judging whether attribute information of the labeling data is stored in the corresponding graph node or not through the second knowledge graph retrieval, if yes, traversing the attribute information of the labeling data, wherein the attribute information of the labeling data comprises the following steps: labeling data category, data position and data content; the labeling data category comprises: relational annotation, text marking, dimension marking and profile information;

According to a specific embodiment, in the above building design atlas visualization apparatus, when the user selects the atlas name and the page number through a picture, the acquiring unit is configured to acquire the atlas name and the page number selected by the user based on a deep learning technique and an OCR technique.

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

1. extracting the graphic identification information and the graphic parameter information of each graphic block from a paper building design graphic set, storing the extracted graphic identification information and graphic attribute information of each graphic block into a predefined graphic set knowledge graph body, establishing a graphic set knowledge graph, and realizing the digital storage of two-dimensional graphic set information and the structured expression of graphic set knowledge through the graphic set knowledge graph; on the basis, a BIM model is built based on graphic parameter information corresponding to a graphic block, labeling data of a user on the atlas BIM model is read, attribute information of the labeling data is stored in a first knowledge graph to obtain a second atlas knowledge graph associated with the BIM entity, on the basis of realizing digital storage and structural expression of two-dimensional atlas information, the BIM model is linked to realize complete digital storage of the atlas information, and a foundation is laid for atlas visual application display;

2. The second knowledge graph is used for stably storing the annotation data of the user about the atlas BIM model, and the application scene of the visual expression of the BIM model is expanded in an auxiliary way through the knowledge graph technology of digital and structured storage, so that the problem that the annotation information is very easy to lose in the model conversion process when the BIM model is effectively solved;

3. based on the expandability of the knowledge graph, the graph body of the graph set knowledge graph can be used as a basic framework for the digital programming of the future graph set, and is favorable for sustainable storage and utilization of the graph set knowledge.

Drawings

FIG. 1 is a flowchart of a method for digitally storing a set of architectural design drawings in accordance with an exemplary embodiment of the present invention;

FIG. 2 is a first schematic diagram of paper atlas data according to an exemplary embodiment of the invention;

FIG. 3 is a diagram illustrating classification of graphical parameter information according to an exemplary embodiment of the present invention;

FIG. 4 is a second schematic diagram of paper atlas data according to an exemplary embodiment of the invention;

FIG. 5 is a third schematic diagram of paper atlas data according to an exemplary embodiment of the invention;

FIG. 6 is a diagram of an atlas knowledge-graph ontology architecture, according to an exemplary embodiment of the present invention;

FIG. 7 is a flowchart of a method of architectural design atlas visualization interaction in accordance with an exemplary embodiment of the present invention;

fig. 8 is a schematic block diagram of a building design atlas visualization interaction, rapid design methodology in accordance with an exemplary embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to test examples and specific embodiments. It should not be construed that the scope of the above subject matter of the present invention is limited to the following embodiments, and all techniques realized based on the present invention are within the scope of the present invention.

Example 1

FIG. 1 illustrates a method for digitizing a set of architectural design drawings in accordance with an exemplary embodiment of the present invention, comprising:

s1, extracting graphic identification information and graphic parameter information of each graphic block from a paper building design drawing set;

s2, establishing a first atlas knowledge atlas corresponding to the paper architectural design atlas according to the extracted graphic identification information, the graphic parameter information and the logic relation between the extracted information;

s3, establishing an atlas BIM model based on the graphic parameter information of each image block, reading annotation data of a user about the atlas BIM model, and determining attribute information of the annotation data;

s4, searching corresponding entity nodes in the first atlas knowledge graph according to the graph identification information corresponding to each atlas BIM model, adding labeling nodes according to the labeling data on the basis of the entity nodes, and storing attribute information of the labeling data to obtain a second atlas knowledge graph, wherein the second atlas knowledge graph and the BIM model are used for realizing the digitization of the paper architectural design atlas.

The graphic identification information of each graphic block is a symbol of each graphic block, and is used for distinguishing and identifying the symbol of each graphic block from symbols of other graphic blocks, such as a paper atlas example diagram shown in fig. 2, and an atlas number source is arranged at the lowest part of the atlas page so as to distinguish the atlas from other atlas; reference numerals will be made to each block in the drawings to illustrate what number of blocks are, for example: the numbers (1) and (5) are two different blocks; meanwhile, based on parameterization building BIM model technology, the graphic parameter information is the relevant information for building a parameterized model, and comprises the following steps: size, geometry, etc.

It can be appreciated that, in order to increase the practicality of the created model and make the model easy to understand, the user will typically make a label on the created atlas BIM model, and typically the user will make data such as dimensions, profile information, etc. on the model to assist the user in making a full spatial understanding of the three-dimensional structure corresponding to the model; however, the BIM model has high requirements on hardware and is easy to cause problems of blocking and the like, so that under a specific application scene, light format conversion is usually required to be carried out on the BIM model, but marking information is very easy to lose in the model conversion process when the BIM model is used, therefore, in the embodiment, marking data are stably stored by acquiring marking data of a user on the model and storing attribute information of the marking data in a knowledge graph in a linkage way, the application scene of visual expression of the BIM model is assisted to be expanded by using a digital and structurally stored knowledge graph technology, the heavy load of visual BIM model marking is assisted, and the problem that the marking information is very easy to lose in the model conversion process when the BIM model is effectively solved.

In summary, the digitizing method provided by the embodiment has the following advantages:

1. extracting the graphic identification information and the graphic parameter information of each graphic block from a paper building design graphic set, storing the extracted graphic identification information and graphic attribute information of each graphic block into a predefined graphic set knowledge graph body, establishing a first graphic set knowledge graph, and realizing the digital storage of two-dimensional graphic set information and the structured expression of graphic set knowledge through the graphic set knowledge graph; meanwhile, a BIM model is built based on graphic parameter information corresponding to a graphic block, marking data of a user on the atlas BIM model is read, attribute information of the marking data is stored in a first knowledge graph to obtain a second atlas knowledge graph related to the BIM entity, and on the basis of realizing digital storage and structural expression of two-dimensional atlas information, the linkage BIM model realizes complete digital storage of the atlas information, so that a foundation is laid for atlas visual application display;

Example 2

In one possible implementation manner, S1 specifically includes: analyzing, extracting and classifying the existing two-dimensional atlas information:

specifically, the image block is taken as a basic unit, and the image identification information and the image parameter information corresponding to each image block are extracted, wherein the image identification information comprises: specialty, atlas number, page number, atlas number;

specifically, as shown in fig. 2 to 5, the parameter information is divided into geometric information and non-geometric information according to the graphic parameter information type. Geometric information is also named model information, which refers to graphic information and a small amount of graphic annotation information used for describing a specific building structure in a drawing, and non-geometric information is text expression information. Geometric information can be classified into text labeling, dimension labeling, profile information and geometric construction according to constituent contents. The non-geometric information is classified into digitalized information and non-digitalized information according to whether digitalized display is possible or not, and digitalized information is classified into a relational annotation and a model annotation according to a digitalized presentation method. The non-digitalized information is a general text annotation which is not related to the visual model, wherein the paper building design atlas is an existing standard paper atlas, and for convenience of presentation, fig. 2, fig. 4 and fig. 5 provided by the application only show partial geometric information and partial non-geometric information corresponding to the standard atlas.

In one possible implementation manner, S2 specifically includes: defining a atlas knowledge atlas body in advance according to the logic relation between the graphic identification information and the graphic parameter information;

It can be understood that the digital application of the paper atlas is most commonly performed in a text description and picture display mode at present, and the paper atlas is displayed to a user through a webpage end (see a website for establishing a label library), and a common database is adopted for storing knowledge items, so that the electronic query of the atlas is facilitated. However, the knowledge correlation retrieval, the visual application, the knowledge reuse of the atlas and the like of the atlas are not supported by the related technology, the value of the atlas is not really exerted, and the knowledge atlas body is established by adopting the knowledge atlas technology and combining with the BIM technology, so that the deep utilization of the atlas can be effectively supported.

Fig. 6 shows an example architecture of an atlas knowledge atlas body predefined in an embodiment of the present invention, where according to the two-dimensional atlas knowledge architecture, the atlas knowledge atlas body is composed of five class levels of specialty, atlas, page, graph, and model information, class level information is described by class attributes, and knowledge links are established between class levels by "has_part_of" relationship and "related_to" relationship, so as to form an atlas knowledge network. Specifically, the atlas knowledge atlas ontology comprises two kinds of triples, (entity) - { attribute: attribute value }, (entity) - [ relationship ] - (entity); the method comprises the following steps:

(1) (entity) - { Attribute: attribute value }

(1) (specialty) - { specialty class: XX }, specialty class typically includes five major specialty of building, structure, water supply and drainage, heating and ventilation, and electromechanics

(2) (atlas) - { atlas number: XX }

(atlas) - { atlas name: XX }

(3) (Page) - { Page number: XX })

(Page) - { text annotation: XX }, classifying according to section 1 information, namely storing text annotation information which cannot be digitally displayed on each page

(4) (Picture) - { Picture number: XX }, each page of picture name of the existing picture set covers two modes of words and sequence numbers, which is unfavorable for computer storage and use, so that pictures are recoded and ordered by picture number attributes from left to right, from top to bottom, from number 1, and picture number ordering is performed by integer increment.

(Picture) - { Picture name: XX })

(graph) - { model annotation: XX }, classifying according to section 1 information, i.e. storing model annotation information digitally displayed for each page

(5) (text annotation) - { text annotation Point 1:XX }

(text annotation) - { text annotation Point 2:XX }

(text endorsement) - { size value: XX })

(6) (size endorsement) - { size endorsement Point 1:XX }

(size endorsement) - { size endorsement Point 2:XX }

(size Annotation) - { size value: XX }

(7) (section) - { section setpoint 1:XX }

(section) - { section setpoint 2:XX }

(section) - { section setpoint 3:XX }

(section) - { section name: XX }

(2) (entity) - [ relationship ] - (entity)

(1) (specialty) - [ has_part_of ] - (atlas)

(2) (atlas) - [ has_part_of ] - (page)

(3) (Page) - [ has_part_of ] - (figure)

(4) (diagram) - [ has_part_of ] - (literal notation)

(diagram) - [ has_part_of ] - (dimension marking)

(diagram) - [ has_part_of ] - (cross section).

Accordingly, the storing the extracted graphic identification information and the graphic attribute information of each tile into the predefined atlas knowledge network architecture includes:

and storing the graphic identification information to a professional, atlas, page and graph level, storing the universal text annotation to a page level, and storing the model annotation to a graph level.

In a possible implementation manner, in S3, the building an atlas BIM model based on the graphics parameter information of each tile specifically includes:

obtaining a first coordinate system corresponding to a current image block and a second coordinate system corresponding to BIM software, determining a coordinate conversion relation between the first coordinate system and the second coordinate system,

and establishing the atlas BIM model according to the geometric information, the relational annotation information and the coordinate conversion relation of the image blocks.

Preferably, to ensure that the built atlas BIM model more conforms to the expression of existing atlas and can facilitate subsequent lightweight model transformations, the built atlas BIM model can follow the following drawing rules/methods:

(1) Size marking and value taking rule:

fixed size: drawing set marking size as drawing size

Interval size: taking critical values by single interval constraint, if a is more than 300, a is more than or equal to 300, and drawing size is 300; the double-interval constraint drawing size takes the middle value, for example, b is more than or equal to 300 and less than or equal to 400, and the drawing size takes 350.

Conditional dimensions: and (3) taking values according to drawing proportion and constraint conditions, wherein if the dimension marking condition is that a takes 300mm when b is less than 300mm, and when b is more than 300mm, a takes 500mm, the condition values are taken according to the drawing dimension of b sides.

The atlas has no rule for value of specified size information: the height information of the wall, column and other types of components is not specified in the building diagram set, the height information is required to be adjusted according to the actual design content, and the corresponding height is required to be presented in the three-dimensional diagram set to show the detailed geometric structure. The model drawing personnel should draw the component height according to the actual component proportion.

(2) Sizing rules:

and (3) fixed size marking: and adding dimension marking information into the atlas BIM model according to the atlas annotation position, wherein the dimension marking characters are real drawing dimensions.

Section size marking: according to the annotation position of the atlas, adding dimension marking information into the atlas BIM model, wherein the dimension marking text needs to replace the real drawing dimension with the atlas dimension information, and if the real drawing dimension b=350 is marked as 300-400.

And (3) marking the following conditions: and adding dimension marking information into the atlas BIM model according to the atlas annotation position, wherein dimension marking characters are value information under different constraint conditions.

The atlas is not specified size-labeled: without labeling, the designer has the general knowledge of selecting the dimensions according to the actual design requirements in the case of random sizing.

(3) The model stores rules:

the model ID is in the format of "professional_atlas number_page number_atlas number", and the model ID is stored in an rfa format file in the atlas BIM model database.

Further, in the process of constructing a three-dimensional BIM model by using the two-dimensional atlas content, there is a conversion between the two-dimensional coordinate system and the three-dimensional coordinate system, so that it is necessary to determine the coordinate system of the two-dimensional atlas and the coordinate system of the three-dimensional BIM model, and in this embodiment, the two coordinate systems are defined as the atlas coordinate system and the BIM model coordinate system respectively:

s31, defining a local coordinate system:

i. establishing a local coordinate system for each different image block of each page in the image set;

the local coordinate system of each tile comprises three mutually perpendicular axes; one of the axes is oriented to the right in the horizontal direction of the drawing, and is called an axis A; one axis takes the vertical upward direction of the drawing sheet as the direction, and is called an axis B; one axis is oriented inward of the drawing sheet and perpendicular to axes a and B, referred to as axis C;

Axes a, B, C represent x-axis, y-axis, z-axis in any combination, depending on the content of the tile representation.

S32, defining an overall coordinate system:

the whole coordinate system is obtained by adopting BIM model coordinate system of BIM design software.

Establishing mapping of an atlas coordinate system and a model coordinate system, and mapping a local origin to an origin of a whole coordinate system;

mapping the x-axis of the local coordinate system to the x-axis of the global coordinate system;

the y-axis of the local coordinate system is mapped to the y-axis of the global coordinate system.

In one possible implementation manner, the attribute information of the annotation data includes: labeling data category, data position and data content;

In a possible implementation manner, in S4, on the basis of the entity node, a labeling node is added according to the labeling data, and the method is used for storing attribute information of the labeling data, and includes:

on the basis of the graph nodes retrieved based on the graph identification information, adding marking nodes in the first knowledge graph according to the marking data types, wherein the marking nodes comprise character marking nodes, dimension marking nodes and profile information nodes; storing the data location and the data content in the added annotation node;

Accordingly, the method and the device for identifying the model in the BIM effectively solve the problem that the marked data is easy to lose in the model conversion process when the BIM is used for carrying out follow-up visual model marking heavy load by acquiring the attribute information (position and content) of the marked data of the user on the model and storing the attribute information in the knowledge graph in a linked mode.

In one possible implementation manner, a tab panel is built in the BIM to assist a user in making model annotations, and after the user annotates, annotation data of the user is read to determine attribute information of the user, which specifically includes:

i. acquiring the file name of a current BIM model family document;

ii, establishing a tab panel, wherein options comprise two types of text labels and dimension labels;

if the user selects the text label, the user clicks the model label position to obtain a text label point 1, then clicks the text label position to obtain a text label point 2, inputs text content, and then automatically draws the text label in BIM design software;

Recording the three-dimensional position of the text mark point 1 clicked by the user in the iii, the text mark 2 in the BIM model coordinate system, the text mark content and the model file name thereof by the program;

if the user selects the dimension marking, the user clicks two points of the dimension marking of the model to obtain a dimension marking point 1 and a dimension marking point 2, then clicks a dimension marking position to obtain a dimension marking point 3, inputs a dimension value, and then automatically draws the dimension marking in BIM design software;

program records vi obtained point of sizing 1, point of sizing 2, the three-dimensional position of the dimension marking point 3 in the BIM model coordinate system, the dimension value and the model file name thereof;

and (3) the program searches the graph nodes in the knowledge graph through the file names according to the obtained content, and adds text marking nodes, dimension marking nodes and attribute content thereof in the graph nodes through the 'has_part_of' relationship.

The above flow can be implemented by secondary development in BIM design software, taking automatic implementation in Revit by means of c# code secondary development as an example:

for i. the file name of the family file can be obtained by the getexterterlfilereference () function and GetAbsolutePath () function of the Revit API;

the fabrication of tab panels requires three steps: creating a tab using the CreateRibbonTab () function tab, creating a tab panel using the CreateRibbonPanel, and creating two buttons using the AddItem () function;

User input can be obtained through uidocumet.selection.pickpoint (), a text annotation object is created through a text note.create () function, the attribute of the number text annotation object is defined, and finally the text annotation object is added into the Revit through the create.newtext () function;

for iv, the information may be stored in the form of a list or dictionary.

Model section acquisition and map relation automatic storage, wherein the graph centered section symbol relates to relation type annotation information display and subsequent visual display and interaction. Therefore, the embodiment of the application establishes a set of profile acquisition scheme and stores the profile acquisition scheme in the knowledge graph in a linkage way for subsequent visual interaction. The specific technical scheme is as follows:

i. acquiring the file name of a current BIM model family document;

ii, establishing a button panel, clicking a button to enter a process of recording profile information of the atlas;

thirdly, a user clicks three points in sequence in a three-dimensional view angle of BIM design software according to the profile position of the atlas, a profile is built, the three points are respectively a profile marking point 1, a profile marking point 2 and a profile marking point 3, and a profile name is input at the position of the profile marking point 3;

iv. The user enters the associated page number_map number for this section;

Recording three-dimensional positions of the section marking point 1, the section marking point 2 and the section marking point 3 which are obtained in the step iii in a BIM model coordinate system by a program, and the section name, the model file name and the atlas page_figure;

the program searches the graph nodes in the knowledge graph through the file names, and adds the profile nodes and attribute contents thereof in the graph nodes through the 'has_part_of' relationship. Further, the model file name professional_atlas number is complemented to the page number_atlas number acquired in the step vi, the section view node or the page node is searched in the knowledge graph through the professional_atlas number_page number_atlas number, and the relation between the section view node and the section view node/the page node is established through the relation of 'related_to'.

for i, a rib panel can be created by the UIControlldAppl.CreateRibbonPanel () function, and then a button can be added to the panel by the AddItem () function;

for ii, firstly, three points clicked by a user are obtained through a UIdocument.selection.PickPoint () function, then, the three points are sequentially transmitted into the function through a plane.CreateByThreeP () function, a plane can be created, finally, a transaction is started, the created plane is put into an active view, and the transaction is submitted;

Firstly, obtaining a plane through a PickObject () function, then obtaining the mouse position of a user at any time, and carrying out vector calculation to obtain a translation vector XYZ of the plane, and preferably, transmitting the translation vector into the plane through a TransLate () function to realize translation of the plane;

for the program, a neo4j.driver library is needed, the library can read and manage knowledge graph data through C#, and C# codes are queried and created by using Match and Create commands through the neo4j.driver library.

In summary, the method provided by the embodiment of the invention has the following advantages:

1. for the atlas information except the geometric structure, a knowledge body is established through a knowledge graph, so that the atlas knowledge structural storage and the complex semantic description are realized, the related information is stored in a graph database mode, the digital storage of the atlas is realized, and the efficient retrieval of the atlas information is assisted; for geometric construction information, a BIM model is established through a BIM technology, so that the storage of the geometric construction information is realized, on the basis, the entity association relation between a knowledge graph and the established BIM model is established through the mode of acquiring labeling data on the model and storing the labeling data in the graph, and the linkage between the knowledge graph and the BIM model is realized, so that the complete digital storage of the atlas information is realized based on the knowledge graph and the BIM technology, and a foundation is laid for the visualization application display of the atlas;

2. On the basis of realizing informatization storage of atlas information based on a knowledge graph and BIM technology, attribute information of the annotation data is stored in a knowledge graph data layer in a linkage way by acquiring the annotation data of a user about a model, the annotation data is stored stably, and the visualization reloading of the model annotation is completed in an auxiliary way.

Example 3

In another aspect of the present invention, as shown in fig. 7, there is provided a building design atlas visualization method, including:

s501, a second knowledge map of each professional paper building design atlas is established in advance by adopting the building design atlas digitizing method of the embodiment 1 or the embodiment 2;

s502, obtaining an atlas visual identifier captured by a user through shooting equipment, and retrieving the atlas visual identifier from the second knowledge graph to obtain a corresponding leaf node; the atlas visualization identifier includes: an atlas number and page number;

s503, based on the page node, reading a plurality of graph node graph numbers contained in the page node, determining a plurality of graph identification information, based on a plurality of graph identification information, reading a corresponding atlas BIM model, carrying out light loading on the read atlas BIM model, and presenting the loaded light atlas BIM model and the corresponding knowledge graph node information to a user.

It can be understood that the BIM model can display design information in three dimensions, but has high requirements on hardware, is easy to cause problems such as jamming and the like, and is difficult to be applied to webpages, AR equipment, mobile equipment and the like, so that the model is used in a field Jing Shouxian; therefore, in this embodiment, an atlas visualization method is provided, in which the knowledge graph and the BIM technology are used to complete the digital storage of the two-dimensional atlas information, and on this basis, when the user selects to read the atlas, the corresponding entity node and the atlas BIM model thereof are found in the second knowledge graph, the BIM model is converted into a lightweight model, and the entity node and the lightweight atlas BIM model thereof are visually presented to the user.

Example 4

In a possible implementation manner, in S503, the light-weighted loading is performed on the read atlas BIM model, which specifically includes: and carrying out format conversion on the atlas BIM model, converting the atlas BIM model into a model in the FBX format, and adjusting the position and angle of the model in the FBX format so that the position and angle of the loaded model are the same as those of the original model.

Specifically, the BIM model in rfa format is a closed model, and because the BIM model contains abundant construction and design information, the BIM model is quite large in size and difficult to directly display in a webpage or AR application. Meanwhile, the method for directly showing the rfa family file in the browser is not supported at the present stage. Thus, to address this problem, and to avoid repetitive modeling, we need to format the rfa model into a presentable format suitable for browser or AR applications.

FBX is a commonly used lightweight model format. The method has the advantage of high compression, and can compress the complex three-dimensional model to a relatively small file size, so that the size and complexity of the model can be greatly reduced, and the loading speed and the display effect of the model on mobile terminal equipment are improved. Meanwhile, the method has the advantage of customization, supports editing in different three-dimensional software, and can be used for conveniently customizing the model in a personalized way. Therefore, the defect of visual interaction of the rfa file is perfectly overcome by the FBX format.

Furthermore, the embodiment provides a set of construction method of the lightweight model, which not only can keep the advantages of high compression and customization of the FBX, but also solves the defect of losing key information, and meanwhile, the whole process is automatic. The specific method is as follows:

the first step: converting the rfa format file into an FBX file. There are various solutions for converting rfa format files into FBX files. Taking Revit as an example, converting by adopting Revit API, so that file change can be avoided as much as possible; specifically, calling a Revit API through C#, configuring an FBXExportoptions configuration item, and performing format conversion through Export;

and a second step of: loading an FBX model through an open source three.js library; three. Js is an open source JavaScript 3D library that can be used to create a variety of 3D scenes and interactive applications; introducing a thread_js package into the JavaScript code, and calling the THREE_FBXLoader () function to put the path of the FBX model into the function so as to realize the loading of the file;

And a third step of: adjusting the position and angle of the loaded FBX model to enable the position and angle of the loaded model to be the same as those of the rfa model; in order to prevent the position of the loaded file from changing, the fbx model position needs to be ensured to be at the origin of a coordinate system; by setting the position. Set (0, 0, 0) of the loaded model, the relocation is possible.

In one possible implementation method, the method further includes: s504, carrying out information reloading flow after the atlas BIM model is loaded in a lightweight mode;

the information reloading process comprises the following steps: judging whether attribute information of the labeling data is stored in the corresponding graph node or not through the second knowledge graph retrieval, if yes, traversing the labeling attribute information, wherein the attribute information of the labeling data comprises the following steps: labeling data category, data position and data content; the labeling data category comprises: relational annotation, text marking, dimension marking and profile information;

and sequentially reloading the data content at the corresponding position of the lightweight BIM according to the type and the position of the marked data through a thread js code library.

Specifically, S504 specifically includes:

s5041: reading a knowledge graph, and retrieving corresponding labeling information through the file name of the FBX model; since automated programs are written in JS language now, a new method is required to retrieve knowledge maps. Installing neo4j-driver packages through npm, starting knowledge graph query transactions by using session () functions, and finally searching all labeling information through a MATCH () RETURN n' command;

S5042: all the labeling information is circulated in turn, and the following third to sixth steps are performed for each labeling information.

S5043: if the labeling information is character labeling, executing a fourth step; if the labeling information is size labeling, executing a fifth step;

s5044: acquiring a labeling point of a character label, a character position of the character label and a character label value, and drawing the character label to a corresponding position of a model through a three.js API; creating a 3D straight line in the three.js, firstly determining the material of the straight line by instantiating the three.linebasic material (), then putting the marked coordinate points obtained in the knowledge graph on the three.buffergeometry (). Setfrom points () object to determine the shape of the straight line, finally putting the material of the straight line and the shape of the straight line into the three.line (), so as to generate a marked line at a specified position, and binding the marked line to the FBX model through an add () function; the text is created in the thread js, and the text and the model can be bound by only putting the content of the text and the position information of the text on the three.

S5045: acquiring two labeling points of the dimension labeling, the text position of the dimension labeling and the value of the dimension labeling, and drawing the dimension labeling to the corresponding position of the model through a three. Js API;

S5046: judging whether all the labeling information is processed, if so, maintaining the model with the heavy load size.

In one possible implementation method, the embodiment of the application provides a picture retrieval function for a user based on VR, AR and MR technologies, and when the user can shoot a picture with an atlas visual identifier (namely, a picture containing an atlas name and a page number) through shooting equipment, the method automatically acquires the atlas name and the page number selected by the user based on a deep learning technology and an OCR technology when the picture of the user is received.

Specifically, VR, AR, and MR technologies are currently developed, and a common feature of the technologies is that a visual identifier is required to establish a mapping between the real world and the virtual world. In order to avoid the transformation of visual identification on a large number of existing paper atlas, the embodiment of the application provides an atlas visual identification acquisition method based on OCR technology and deep learning.

In fact, the drawing set number_page number recorded in the drawing frame at the right lower corner of each page of the drawing set is the unique identification of the drawing set, and the drawing frame is regular in format and quite easy to identify, so that in the embodiment, the paper drawing set number_page number is used as the drawing set visual identification, a user can shoot a certain page of the drawing set browsed at present through an OCR character recognition technology and a target detection deep learning technology, the target detection deep learning technology automatically locates the drawing frame area, and the OCR character recognition technology automatically extracts characters in the area.

The extracted text contains an atlas number, the currently browsed page number of the atlas, and all atlas information of the page is retrieved and acquired in the knowledge graph by the atlas number_page number. The implementation steps are as follows:

(1) Visual identification acquisition

The first step: deploying a deep learning model on a background server, deploying a trained target detection model and an OCR model on the background server by using a deep learning framework such as TensorFlow or PyTorch;

and a second step of: the method comprises the steps that an image uploading API is used at a mobile phone end, an image is uploaded to a background server through HTTP or other protocols at the mobile phone end, after the server receives the image, preprocessing and post-processing are carried out, and the image is input into a deep learning model for image recognition;

and a third step of: the mobile phone terminal shoots a picture of a certain page of the atlas;

fourth step: the atlas picture is transmitted to the background through the API of the second step;

fifth step: using the target detection model which is well arranged in the first step in the background to detect the position of a picture frame in the atlas picture;

sixth step: picture cropping is carried out through a PL library of python; specifically calling a crop function in PL, inputting the frame position detected in the fifth step, and outputting the frame picture of the shot album picture;

Seventh step: the local picture frame is subjected to character recognition by using an OCR technology, specifically, a CV2 library can be used for reading the picture frame, and then a pytessaact.

(2) Knowledge graph retrieval and model acquisition based on visual identification

The first step: establishing connection with neo4j through JS, and starting a transaction;

and a second step of: transmitting CyPher statement to neo4j through JS, and searching all data of the current atlas and the current page number in a third knowledge graph due to the known atlas name and the current page number, wherein startNodeName represents the atlas name, endNodeName represents the current page number, and the content searched by the releatednode from the knowledge graph is ' MATCH ' (startNode { name: $startNodeName) - [ - > (endNode { name: $endNodeName) - [ ] - > (releatedNode) RETURN relatedNode ';

and a third step of: and circulating the relatedNode to obtain a BIM model, a 3D model, jump information and other information.

In summary, the embodiment realizes the atlas visual interaction based on the lightweight model and the visual identifier, including VR, AR, MR scene interaction and web interaction; the established visual identification is combined with an AR/MR technology based on the identification, so that the view of the lightweight model of the atlas and related contents in the scene of using the mobile terminal handheld device can be realized; and (3) based on the established visual identification, embedding the lightweight model obtained by searching in the knowledge graph into html to realize that the web end views the lightweight model of the atlas and related content.

In a possible implementation manner, in the architectural design atlas visualization method, when the acquired page number has a profile labeling node associated with the acquired page number, the method further includes:

creating a plane, wherein the position of the created plane coincides with the position of the section, dividing the light BIM model into two parts through the plane,

and acquiring a hiding instruction of a user on one side of the plane relative to the lightweight BIM model, hiding a part of the model based on the hiding instruction, and presenting the section.

Specifically, for three-dimensional visual expression of a two-dimensional atlas, since a projection view and a section view of the same component in the two-dimensional atlas are not usually on the same page, and page skip is involved, when a handheld device shoots a section page, the user needs to skip to a section position corresponding to a lightweight model, so that the present patent proposes a section skip method of atlas visual interaction to improve the use convenience of the visual atlas, and the specific method is as follows:

the first step: loading a visualization model through an FBXLoader () function;

And a second step of: three-dimensional positions of a model section marking point 1, a section marking point 2 and a section marking point 3 stored in a knowledge graph in a BIM model coordinate system are obtained, three points p1, p2 and p3 are declared in a thread. Js through new THREE. Vector3 (X, Y and Z), and plane. SetFromCoplanarPoints (p 1, p2 and p 3) are called to create an infinitely-extended plane, and the plane position coincides with the section position; then call webglrenderer.

And a third step of: binding a mouse click event to the hot spot through the thread. Js; the plane (section) divides the model into two parts which are respectively positioned at two sides of the plane (section); clicking the model on one side by a mouse, hiding the model, and then viewing the section; the specific code implementation method comprises the following steps: firstly, creating a raycaster object through a new three. Raycaster () function, then creating a mouse vector through new three. Vector2 (), and then monitoring a mouse click event through a window. Addeventlister ('click', onClick, false) code; in the event, the position of the mouse vector is firstly obtained, and then the difference value between the mouse vector and the camera vector is calculated, so that the raycaster object is updated. The raycaster object is a ray, and the intersection of the ray and the part of the segmented model is judged; thus hiding the model by its material.

Example 5

by adopting the building design atlas visualization method of the above embodiment 4, loading a light atlas BIM model and corresponding knowledge-graph node information;

and (3) learning and teaching application of the atlas knowledge are realized based on the lightweight atlas BIM model, and/or the atlas knowledge is acquired based on the knowledge atlas node information retrieval, and the BIM model required by the design is downloaded for rapid forward design.

Specifically, as shown in fig. 8, on the basis of the architectural design atlas visualization method (web interaction) provided in embodiment 4, when using corresponding software, a designer may obtain atlas knowledge based on the retrieved knowledge atlas node information; and/or when a designer checks the lightweight model of the atlas at the web end, clicking a model downloading button, and searching and extracting the BIM model corresponding to the lightweight model in the knowledge graph by the program according to the visual identification of the model, wherein the designer can realize quick design by only modifying the model family file according to the design size and then linking the model family file into the design model, thereby realizing quick design based on the visual atlas BIM model.

Example 6

In another aspect of the present invention, there is provided a building design atlas digitizing storage apparatus for implementing the building design atlas digitizing method according to embodiment 1 or embodiment 2, the apparatus comprising:

and the third processing unit is used for adding annotation nodes or labeling nodes on the basis of the graph nodes according to the graph parameter information category in the first knowledge graph according to the labeling data category on the basis of the retrieved graph nodes, and storing the rest attribute information in the added nodes.

Example 7

In another aspect of the present invention, there is provided a building design atlas visualization interaction apparatus for implementing the building design atlas visualization method described in embodiment 3 or embodiment 4, the apparatus comprising:

In one possible implementation, the visualization loading unit is further configured to: judging whether attribute information of the labeling data is stored in the corresponding graph node or not through the second knowledge graph retrieval, if yes, traversing the labeling attribute information, wherein the attribute information of the labeling data comprises the following steps: labeling data category, data position and data content; the labeling data category comprises: relational annotation, text marking, dimension marking and profile information;

In one possible implementation, when the user selects the album name and the page number by picture, the acquisition unit is configured to acquire the album name and the page number selected by the user based on the deep learning technique and the OCR technique.

The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.

Claims

1. A method for digitizing a set of architectural designs, comprising:

searching corresponding entity nodes in the first atlas knowledge graph according to the graph identification information corresponding to each atlas BIM model, adding labeling nodes according to the labeling data on the basis of the entity nodes, and storing attribute information of the labeling data to obtain a second atlas knowledge graph, wherein the second atlas knowledge graph and the atlas BIM model are used for realizing the digitization of the paper architectural design atlas;

The step of establishing the first atlas knowledge atlas includes:

and storing the extracted graph identification information and the graph parameter information into a predefined atlas knowledge-graph body to obtain the first atlas knowledge-graph.

2. The architectural design atlas digitizing method of claim 1, wherein the atlas knowledge atlas ontology comprises: the method comprises the steps of establishing knowledge links among five class levels through logical relations.

3. The architectural design atlas digitizing method of claim 1, wherein the graphical identification information comprises: specialty code, atlas number, page number, atlas number;

4. A method of digitizing a building design atlas according to claim 3, wherein the building atlas BIM model based on the graphic parameter information for each tile comprises:

5. The method for digitizing a set of architectural designs according to any one of claims 1 to 4, wherein the attribute information of the annotation data includes: labeling data category, data position and data content; the labeling data category comprises: relational annotation, text marking, dimension marking and profile information.

6. The method for digitizing a set of architectural designs according to claim 5, wherein the adding, based on the entity node, a labeling node according to the labeling data for storing attribute information of the labeling data includes:

7. A method of building design atlas visualization, the method comprising:

a second knowledge graph of each professional paper building design atlas is established in advance by adopting the building design atlas digitizing method of any one of claims 1-6;

8. The architectural design atlas visualization method of claim 7, wherein lightweight loading the read atlas BIM model comprises:

9. The architectural design atlas visualization method of claim 7, wherein the obtaining an atlas visualization identifier captured by a user via a photographing device comprises:

10. The architectural design atlas visualization method of claim 7, further comprising: information reloading flow after the atlas BIM model is loaded in a lightweight mode;

11. The architectural design atlas visualization method of claim 7, wherein when the corresponding page node is retrieved from the second knowledge-graph to have a profile annotation node associated therewith, the method further comprises:

12. A method of building design atlas application, comprising:

loading a light atlas BIM model and corresponding knowledge graph node information by adopting the architectural design atlas visualization method according to any one of claims 7-11;

13. A building design atlas digitizing apparatus, the apparatus comprising:

the third processing unit is used for retrieving corresponding entity nodes in the first atlas knowledge graph according to the graph identification information corresponding to each atlas BIM model, adding labeling nodes according to the labeling data on the basis of the entity nodes, storing attribute information of the labeling data to obtain a second atlas knowledge graph, and realizing the digitization of the paper architectural design atlas through the second atlas knowledge graph and the atlas BIM model;

Wherein, the first processing unit is specifically configured to:

14. The architectural design atlas digitizing apparatus of claim 13, wherein the attribute information of the annotation data comprises: labeling data category, data position and data content;

15. A building design atlas visualization device, comprising:

the storage unit is used for establishing a second knowledge graph of each professional paper building design atlas by adopting the building design atlas digitizing method of any one of claims 1-6 in advance;

16. The architectural design atlas visualization device of claim 15, wherein the visualization loading unit is further configured to: judging whether attribute information of the labeling data is stored in the corresponding graph node or not through the second knowledge graph retrieval, if yes, traversing the attribute information of the labeling data, wherein the attribute information of the labeling data comprises the following steps: labeling data category, data position and data content; the labeling data category comprises: relational annotation, text marking, dimension marking and profile information;

17. The architectural design atlas visualization apparatus according to claim 15, wherein when the user selects the atlas name and the page number by picture, the acquisition unit is configured to acquire the atlas name and the page number selected by the user based on a deep learning technique and an OCR technique.