Column sample identification method and device, electronic equipment and readable storage medium

Abstract

The present application relates to the technical field of engineering construction, and discloses a column sample identification method, device, electronic equipment and readable storage medium. Wherein, the method comprises: obtaining a column sample drawing to be identified; identifying text information in the column sample drawing to obtain a plurality of text entities; associating the plurality of text entities with column sample graph elements in the column sample drawing to generate a column sample directed graph; and based on the column sample directed graph, performing traversal combination on the column sample graph elements and the text entities to obtain a plurality of target column samples. Through the implementation of the technical scheme of the present application, comprehensive analysis of the column sample is realized, and column sample identification in multiple scenarios is supported, thereby overcoming the limitations of column sample identification.

Description

Technical Field

[0001] This invention relates to the field of engineering and construction technology, specifically to a method, device, electronic device, and readable storage medium for identifying column details. Background Technology

[0002] In architectural engineering, column detail drawings typically include the vast majority of column details, meaning most are drawn on a single sheet of paper, with a smaller portion included in the column plan. The issue of multiple names for a single column detail drawing often arises when the drawing is done separately and includes tables. Examples include: different elevations on the same drawing have different longitudinal reinforcement; some floors have longitudinal reinforcement while others do not; using codes to associate elevations generates different cross-sections and reinforcement specifications for different floors; using brackets or other markers to associate elevations and generate various other information; and associating names with elevations and generating various other information, etc.

[0003] Regarding the issue of multiple interpretations in column detail drawings, most related technologies rely on hard coding for single scenarios, which can only solve local problems and cannot comprehensively analyze multiple interpretations, resulting in certain limitations in the recognition of column detail drawings. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method, apparatus, electronic device and readable storage medium for identifying column detail drawings, in order to solve the problem of difficulty in achieving comprehensive analysis of multiple interpretations of a single column detail drawing.

[0005] According to a first aspect, embodiments of the present invention provide a method for identifying column detail drawings, comprising: acquiring a column detail drawing to be identified; identifying text information in the column detail drawing to obtain multiple text entities; associating the multiple text entities with column detail elements in the column detail drawing to generate a column detail directed graph; and traversing and combining the column detail elements and the text entities based on the column detail directed graph to obtain multiple sets of target column details.

[0006] The column detail recognition method provided in this invention identifies multiple text entities by recognizing text information in the column detail drawing, and associates each text entity with the column detail primitives to generate a directed graph of the column detail. This directed graph determines various information represented by the column detail primitives, facilitating a comprehensive interpretation of the information of each column detail in the drawing. By traversing and combining the column detail primitives and text entities through the directed graph, multiple sets of target column details are obtained, achieving comprehensive analysis of multiple meanings in a single column detail drawing. This supports column detail recognition in various scenarios and overcomes the limitations of traditional column detail recognition.

[0007] In conjunction with the first aspect, in the first embodiment of the first aspect, the step of associating multiple text entities with column detail elements in the column detail drawing to generate a directed column detail graph includes: determining whether each text entity is associated with the column detail element; when one or more text entities are not associated with the column detail, removing the text entities that are not associated with the column detail element to obtain multiple target associated entities; associating the multiple target associated entities with the column detail element to generate the directed column detail graph.

[0008] The column pattern recognition method provided in this embodiment of the invention determines whether there are irrelevant text entities by detecting the correlation between each text entity and the column pattern primitives, removes irrelevant text entities to determine the target associated entity, and generates a corresponding column pattern directed graph based on the target associated entity and the column pattern primitives, thus ensuring the effective association between text entities and column pattern primitives.

[0009] In conjunction with the first embodiment of the first aspect, in the second embodiment of the first aspect, the step of associating multiple target associated entities with the column detail primitives to generate the column detail directed graph includes: obtaining constraint values ​​corresponding to each target associated entity; combining multiple constraint values ​​to obtain multiple target constraint values; and constraining the column detail primitives based on the multiple target constraint values ​​to obtain the column detail directed graph.

[0010] The column detail identification method provided in this embodiment of the invention obtains multiple target constraint values ​​by combining the constraint values ​​corresponding to each target associated entity, thereby determining all constraint value combinations. Then, the target constraint values ​​are used to constrain the column detail primitives, thus ensuring that the target constraint values ​​meet all scenarios to the greatest extent.

[0011] In conjunction with the second implementation of the first aspect, in the third implementation of the first aspect, when the constraint value is an elevation segment, the step of combining multiple constraint values ​​to obtain multiple target constraint values ​​includes: performing intersection processing and / or remainder processing on multiple elevation segments to obtain multiple target elevation segments.

[0012] In conjunction with the second embodiment of the first aspect, in the fourth embodiment of the first aspect, when the constraint value is a name, the step of combining multiple constraint values ​​to obtain multiple target constraint values ​​includes: determining a target constraint name from the multiple names; integrating the other names with the target constraint name to obtain an integrated name, wherein the integrated name is consistent with the target constraint name.

[0013] In conjunction with the second embodiment of the first aspect, in the fifth embodiment of the first aspect, when the constraint value is a label, the step of combining multiple constraint values ​​to obtain multiple target constraint values ​​includes: classifying different labels to obtain label classification results; and associating and combining the labels based on the label classification results to generate target constraint values ​​corresponding to each label.

[0014] The column sample identification method provided in this embodiment of the invention facilitates the combination of different column sample information by performing corresponding combination processing on different constraint values, thereby supporting column sample identification for all usage scenarios to the greatest extent.

[0015] In conjunction with any one of the second to fifth embodiments of the first aspect, in the sixth embodiment of the first aspect, before constraining the column detail primitives based on the plurality of target constraint values ​​to obtain the column detail directed graph, the method includes: determining whether there are duplicate constraint values ​​among the plurality of target constraint values; when there are duplicate constraint values, merging the duplicate constraint values ​​to obtain the merged target constraint values.

[0016] The column sample identification method provided in this embodiment of the invention detects duplicate constraint values ​​in the target constraint values ​​and merges the duplicate constraint values, thereby avoiding repeated combination processing of duplicate constraint values ​​and improving identification efficiency.

[0017] In conjunction with the first aspect, in the seventh embodiment of the first aspect, the step of traversing and combining the primitives and the text entities based on the directed graph of the column pattern to obtain multiple sets of target column patterns includes: determining constrained nodes and unconstrained nodes based on the directed graph of the column pattern; performing reverse traversal starting from constrained nodes that have no successor nodes to obtain backtracking nodes; and generating target column patterns for each constrained node based on the information represented by the backtracking nodes and the unconstrained nodes.

[0018] The column sample identification method provided in this embodiment of the invention performs reverse traversal through constrained nodes to determine backtracking nodes that can characterize column sample information, and combines the information represented by the backtracking nodes and unconstrained nodes to determine the target column sample, thus ensuring the integrity of the target column sample information.

[0019] In conjunction with the seventh embodiment of the first aspect, in the eighth embodiment of the first aspect, generating a target column sketch for each constrained node based on the information represented by the backtracking node and the unconstrained node includes: obtaining display settings for the backtracking node and the unconstrained node; controlling the display of the information represented by the backtracking node and the unconstrained node based on the display settings, and generating the target column sketch.

[0020] The column sample identification method provided in this embodiment of the invention combines the display settings of backtracking nodes and unconstrained nodes to control the display of the information represented by the backtracking nodes and unconstrained nodes in order to generate the target column sample, thereby ensuring the display integrity of the target column sample.

[0021] According to a second aspect, embodiments of the present invention provide a column detail identification device, comprising: an acquisition module for acquiring a column detail drawing to be identified; an identification module for identifying text information in the column detail drawing to obtain multiple text entities; an association module for associating the multiple text entities with column detail elements in the column detail drawing to generate a column detail directed graph; and a generation module for traversing and combining the column detail elements and the text entities based on the column detail directed graph to obtain multiple sets of target column details.

[0022] According to a third aspect, embodiments of the present invention provide an electronic device, including: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the column sample identification method described in the first aspect or any embodiment of the first aspect.

[0023] According to a fourth aspect, embodiments of the present invention provide a computer-readable storage medium storing computer instructions for causing a computer to perform the column sample identification method described in the first aspect or any embodiment of the first aspect.

[0024] It should be noted that the beneficial effects of the column sample identification device, electronic device, and computer-readable storage medium provided in the embodiments of the present invention can be found in the description of the corresponding content in the column sample identification method, and will not be repeated here. Attached Figure Description

[0025] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0026] Figure 1 A schematic diagram showing multiple names on a single drawing of a column detail drawing is provided.

[0027] Figure 2 This is a flowchart of a column sample identification method according to an embodiment of the present invention;

[0028] Figure 3This is another flowchart of a column sample identification method according to an embodiment of the present invention;

[0029] Figure 4 A schematic diagram of a column detail diagram in an embodiment of the present invention is shown;

[0030] Figure 5 A schematic diagram of a directed graph of a column sample is shown in an embodiment of the present invention;

[0031] Figure 6 This illustrates another scenario diagram of a column outline in an embodiment of the present invention;

[0032] Figure 7 Another schematic diagram of the column-scale directed graph in an embodiment of the present invention is shown;

[0033] Figure 8 This illustration shows another scenario of a column detail diagram in an embodiment of the present invention;

[0034] Figure 9 This shows another schematic diagram of the directed graph of the column sample in an embodiment of the present invention;

[0035] Figure 10 This illustration shows another scenario of a column outline diagram in an embodiment of the present invention;

[0036] Figure 11 This is another schematic diagram of the directed graph of the column sample in an embodiment of the present invention;

[0037] Figure 12 A schematic diagram illustrating the type system for different nodes in an embodiment of the present invention is shown;

[0038] Figure 13 This is another flowchart of the column sample identification method according to an embodiment of the present invention;

[0039] Figure 14 This is a structural block diagram of a column sample identification device according to an embodiment of the present invention;

[0040] Figure 15 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] To reduce the length of drawings, column detail drafters often represent component information for multiple column details in a single detail drawing. This leads to the problem of multiple names for a single drawing, where information from a detail drawing (not limited to annotations) is used in a limited number of permutations and combinations to generate multiple sets of information, each of which can fully describe a set of component information.

[0043] For example, in the same drawing, different elevations have different longitudinal reinforcements, such as... Figure 1 As shown in Figure ①; in the same drawing, some floors have longitudinal reinforcement, while others do not, such as... Figure 1 As shown in Figure ②; within the same drawing, codes are used to associate elevations, generating information on different cross-sections and rebar specifications for different floors, such as... Figure 1 As shown in Figure ③, the elevation is associated with markers such as parentheses, and various other information is generated by associating these markers with other information, such as... Figure 1 As shown in Figure ④, various information is generated by associating names with elevations and other information.

[0044] Based on this, the technical solution of this invention constructs a directed graph of column details to determine various information represented by the column detail primitives, facilitating a comprehensive interpretation of the information of each column detail in the drawing. By traversing and combining column detail primitives and text entities through the directed graph of column details, multiple sets of target column details are obtained, achieving comprehensive analysis of multiple meanings in a single column detail drawing, supporting column detail recognition in various scenarios, and overcoming the limitations of column detail recognition.

[0045] According to an embodiment of the present invention, an embodiment of a method for identifying column samples is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0046] This embodiment provides a method for identifying column samples, which can be used in electronic devices such as mobile phones, tablets, computers, servers, etc. Figure 2 This is a flowchart of a column sample identification method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0047] S11, Obtain the column detail drawing to be identified.

[0048] Column detail drawings are architectural structural drawings created by engineering drafters. These column detail drawings contain component information for multiple sets of column details.

[0049] In some implementations, users can upload the column detail drawing to be identified to an electronic device via an external storage medium, and the electronic device can then read the column detail drawing from the external storage medium.

[0050] In some implementations, users can draw column detail drawings using drawing tools deployed on an electronic device and export them after completion. The electronic device can then read the exported column detail drawings.

[0051] Of course, the detailed drawings of this column can also be obtained through other means, which are not specifically limited here.

[0052] S12 identifies text information in column detail drawings and obtains multiple text entities.

[0053] The electronic device is equipped with a text recognition model, which can be a convolutional recurrent neural network (CRNN), a recurrent neural network (RNN), or other models, without specific limitations here.

[0054] The column detail drawing includes text information that characterizes the column detail information, including text annotations and symbols. The electronic device uses this text recognition model to identify the text elements in the column detail drawing, aggregates the text elements into text entities, and uses these text entities to represent the parameter information of the column detail.

[0055] S13: Associate multiple text entities with column detail elements in the column detail drawing to generate a directed column detail graph.

[0056] A directed graph for column details is used to represent the relationship between text entities and column detail elements. The electronic device extracts each group of column detail elements from the column detail drawing. Since the information interpreted by each text entity can be considered as the association information with the column detail elements, multiple sets of association information are obtained by associating each text entity with related column detail elements. These multiple sets of association information are then combined to form a directed graph for column details.

[0057] S14. Based on the directed graph of the column outline, the column outline primitives and text entities are traversed and combined to obtain multiple sets of target column outlines.

[0058] The directed graph of a column detail consists of multiple nodes, each representing the position of components and text entities within the column detail primitives. The target column detail is generated based on the constraints of the directed graph.

[0059] The nodes in the directed graph of the column layout are traversed to identify multiple target nodes that do not have child nodes. Each target node is used as a starting point to perform a reverse traversal until there are no parent nodes. The nodes obtained by the reverse traversal of each target node and their corresponding text entities are combined to obtain multiple sets of information to be expressed by the column layout. The hidden line removal control is performed on the information to be expressed by each set of column layouts to generate the corresponding target column layout.

[0060] The column detail recognition method provided in this embodiment identifies multiple text entities by recognizing text information in the column detail drawing, and associates each text entity with the column detail primitives to generate a directed graph of the column detail. This directed graph determines various information represented by the column detail primitives, facilitating a comprehensive interpretation of the information of each column detail in the drawing. By traversing and combining the column detail primitives and text entities through the directed graph, multiple sets of target column details are obtained, achieving comprehensive analysis of multiple meanings in a single column detail drawing. This supports column detail recognition in various scenarios and overcomes the limitations of traditional column detail recognition.

[0061] This embodiment provides a method for identifying column samples, which can be used in electronic devices such as mobile phones, tablets, computers, servers, etc. Figure 3 This is a flowchart of a column sample identification method according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0062] S21, Obtain the column detail drawing to be identified. For detailed instructions, please refer to the relevant descriptions in the above embodiments; they will not be repeated here.

[0063] S22, Identify the text information in the column detail drawing to obtain multiple text entities. For detailed explanations, please refer to the relevant descriptions in the above embodiments; they will not be repeated here.

[0064] S23, associate multiple text entities with column detail elements in the column detail drawing to generate a directed column detail graph.

[0065] Specifically, step S23 may include:

[0066] S231, determine whether each text entity is associated with the column detail element.

[0067] All text entities should be associated with column feature primitives. However, text noise (i.e. irrelevant information) is inevitably identified during the text recognition process. This text noise has no relation to the column feature primitives.

[0068] If a text entity can be integrated into a column detail element, it indicates that there is an association between the text entity and the column detail element. The electronic device integrates each text entity with the column detail element to determine whether all text entities can be integrated into the column detail element. When there is a text entity that cannot be integrated into the column detail element, it indicates that the text entity is not associated with the column detail element, and step S232 is executed; otherwise, it indicates that all text entities are associated with the column detail element. In this case, all text entities can be identified as target associated entities and associated with the column detail element.

[0069] S232, when one or more text entities are not associated with the column outline, remove the text entities that are not associated with the column outline elements to obtain multiple target associated entities.

[0070] When one or more text entities are not associated with the column outline, these one or more text entities are removed to obtain the remaining text entities, which are then identified as the target associated entities.

[0071] S233 associates multiple target entities with column detail primitives to generate a directed column detail graph.

[0072] Based on the relationship between the target associated entity and the column detail primitive, a directed line segment is constructed between the column detail primitive and the target associated entity. Based on this directed line segment, a directed acyclic graph is formed, which is the column detail directed graph.

[0073] Specifically, step S233 above may include:

[0074] (1) Obtain the constraint values ​​corresponding to each target associated entity.

[0075] (2) Combine multiple constraint values ​​to obtain multiple target constraint values.

[0076] (3) Constrain the column detail primitives based on multiple target constraint values ​​to obtain the column detail directed graph.

[0077] Different target associated entities correspond to different constraint values. All constraint values ​​are ranged and organized to generate independent subsets based on the constraint values. These independent subsets are used to constrain the column detail primitives, resulting in a column detail directed graph that can represent multiple sets of associated information.

[0078] In some optional implementations, the constraint value can be an elevation segment, and step (2) above can include: performing intersection processing and / or remainder processing on multiple elevation segments to obtain multiple target elevation segments.

[0079] When a column detail element has multiple elevation segment constraints, intersection and / or remainder operations are performed on the different elevation segments. Specifically, suppose there are two elevation segments A and B, where B contains A. In this case, performing intersection and remainder operations on these two elevation segments A and B yields two new target elevation segments C and D. Here, C = A ∩ B, and D = BA. The newly generated target elevation segment C is obviously contained within elevation segments A and B, and can continue to constrain elevation segments A and B. The other target elevation segment D is only contained within elevation segment B, meaning it can only constrain elevation segment B.

[0080] by Figure 4 Taking the column detail drawing as an example, two of the 14 longitudinal bars are constrained by two target-related entities, and the constraint values ​​of these two target-related entities are both in the elevation range of "-0.10 to 5.70". There is also another target-related entity ( Figure 4 (Not shown in the image) is used for the concentrated constraints of the elevation segment, which represents the elevation segment "-0.01 to 10.00". For this column detail element, it specifically has two elevation segments: "-0.01 to 5.70" and "5.70 to 10.00", and the second elevation segment obviously needs to be obtained by subtracting the two elevation segments.

[0081] like Figure 5 As shown, the root represents the column detail drawing, the circle represents the longitudinal reinforcement, and the black represents text entities. Considering the complexity of the illustration, stirrups and other non-contributing information are not drawn. Intersection and remainder operations are performed on the elevation segments "-0.10~5.70" and "-0.01~10.00" to generate target elevation segments: "-0.010~5.700" and "5.700~10.000". Observing the directed graph of the column detail drawing based on these target elevation segments, it can be found that by backtracking from nodes "-0.010~5.700" and "5.700~10.000" respectively, collecting the traversed nodes yields two sets of information: one set contains two longitudinal reinforcements associated with the elevation segment "-0.010~5.700", and the other set does not contain any.

[0082] In some alternative implementations, the constraint value can be a name, and step (2) above may include:

[0083] (21) Determine the target constraint name from multiple names.

[0084] (22) Integrate other names with the target constraint name to obtain an integrated name, which is consistent with the target constraint name.

[0085] The target constraint name is used to constrain the information represented by the column detail primitive. This target constraint name is a name with exclusive characteristics. Among the multiple target-related entities corresponding to the column detail primitive, there may be multiple names. The electronic device determines the target constraint name heuristically and integrates it with other names to obtain the final integrated name, which is the target constraint name. In other words, the result of integrating any other non-exclusive name with the target constraint name is a target constraint name with exclusive characteristics.

[0086] Here, we take a specific scenario as an example, such as Figure 6 The name shown is "KZ7a only". The column detail drawing links multiple names "KZ7" and "KZ7a" through a table, but near a certain column detail element, it is further explained by the text entity "KZ7a only". "KZ7a only" can be interpreted as a name with exclusive characteristics, while "KZ7" and "KZ7a" are names without exclusive characteristics.

[0087] Let "!KZ7a" represent "only KZ7a is shown", and use the idea of ​​range intersection and merging to inspire: ! When KZ7a and KZ7 intersect and merge, their intersection is KZ7a, with no remainder; ! The intersection of KZ7a and KZ7a is still KZ7a, with no remainder; KZ7a and KZ7 have no intersection and no remainder. That is, after merging and associating, the entire column-specific directed graph shrinks to only parameter KZ7a.

[0088] Specifically, see the directed graph of the column. Figure 7 Wherein, the root represents the column detail drawing, which is related to... Figure 6 The column outline shown has a 1:1 representation relationship, and the node "KZ7a only" is for... Figure 6 The Chinese character element "Show Only KZ7a" is represented here by the constraint value "!KZ7a", which is attached to "Show Only KZ7a". In other words, the character element "Show Only KZ7a" refers to the information source of "!KZ7a", representing a parent-child relationship. "KZ7" and "KZ7a" originate from the capture association relationship between the table and column detail drawings. For table identification and the establishment of association relationships, please refer to the relevant industry standards; details will not be elaborated here.

[0089] Since the name belongs to a discrete range, a heuristic approach can be used to find multiple ranges of information represented by the bar chart. Specifically, we can define the intersection and union of "!KZ7a" and "KZ7" as KZ7a, with an empty remainder; let the intersection and union of "!KZ7a" and "KZ7a" also result in KZ7a, with an empty remainder; merge the two KZ7as to obtain a single child node KZ7a. Thus, after integration, the information in the bar chart is reduced to a single range node KZ7a. During reverse traversal, this range node only contains "KZ7a", thereby filtering (rejecting) information.

[0090] In some alternative implementations, the constraint value can be a marker, and step (2) above may include:

[0091] (23) Classify the different tags to obtain the tag classification results.

[0092] (24) Based on the label classification results, the labels are associated and combined to generate target constraint values ​​corresponding to each label.

[0093] The markers include bracket markers and sequence markers. These two types of markers are generally not mixed, but there are still some scenarios where they are used interchangeably. To address this, electronic devices can classify the identified markers and then associate and combine them based on the classification results to determine the target constraint value used to represent each marker. For example, merging sequence 1 with "null", sequence 2 with "()", and sequence 3 with "[]", then for sequence 1, the target constraint value is "null", for sequence 2, it is "()", and for sequence 3, it is "[]".

[0094] by Figure 8 Taking the column detail diagram shown as an example, a directed column detail diagram can be generated based on the above method, such as... Figure 9As shown. Black nodes are all text entities associated with the original column detail element; 20C12 is interpreted as longitudinal reinforcement specification; C8@200 is interpreted as stirrup specification; the name node GBZ14 (GBZ14a) is interpreted as two name values, each marked with parentheses, namely, no parentheses (Null) and with parentheses (small parentheses); the elevation segment "11.120~69.960; 74.500" is interpreted as two elevation value ranges: 11.120~69.960 and 11.120~74.500, respectively, with column label constraints. The column labels are divided into sequence numbers 1 and 2. A heuristic approach is needed to associate and combine the column labels. For example, associating sequence number 1 with "null" and sequence number 2 with "()", after sequence numbers 1 and 2 are associated with "null" and "()", they lose their eligibility as backtracking start nodes. At this point, the information of the column detail element will be interpreted as two sets of information.

[0095] By combining different constraint values, it is easy to obtain different column sample information, which supports column sample recognition for all usage scenarios to the greatest extent.

[0096] As an optional implementation, before step (3), the following may also be included:

[0097] (31) Determine whether there are duplicate constraint values ​​among multiple target constraint values.

[0098] (32) When there are duplicate constraint values, merge the duplicate constraint values ​​to obtain the target constraint value after merging.

[0099] Multiple target constraint values ​​are compared to determine if any are duplicated. If duplicate constraints are found, they are merged into one. Figure 4 and Figure 6 As shown. After merging, the target constraint value is obtained, which is the duplicate constraint value.

[0100] By detecting duplicate constraint values ​​in the target constraint values ​​and merging them, the repeated combination of duplicate constraint values ​​is avoided, thereby improving the recognition efficiency.

[0101] As an optional implementation, the constraint value can also be a centralized label associated with the elevation. Figure 10 Taking the column layout diagram shown as an example, text recognition can yield the text entity "7C14(Total)\n(Only for layers 14-30)\n10C25\n(Only for layers 31-32)". Based on this text entity, a directed graph with the text entity as its root can be generated, such as... Figure 11 As shown.

[0102] Considering the similarity of association or matching operations, we will use the same data structure to describe the association results. That is, whether it's a text entity, a value, or floor information, it needs to inherit from a base class (node) capable of performing association operations. A type system is now introduced to describe the specialized representations of different nodes, such as... Figure 12 As shown.

[0103] exist Figure 12 In the static object type description (i.e., static UML) shown, the top-level "edge," "node," and "graph" represent the necessary concepts contained in a graph data structure. A "graph" includes a list of nodes, with edges connecting them. The following sections... Figure 12 The concepts of each node in the process are described as follows:

[0104] (1) Constraint Node: The node in the graph data structure concept is specialized into a "constraint node". "Constraint" means that each node forms a constraint on the column detail information to specialize the representation range of the column detail information.

[0105] (2) Entities and values: Constraint nodes are divided into entity constraint nodes (“entities”) and value constraint nodes (“values”) according to the information source of the column detail drawing. The source of entities is the symbols that appear in the form of CAD elements in the column detail drawing (such as longitudinal reinforcement circles, stirrup lines, and text blocks). The source of values ​​is the literal meaning value of the text CAD elements in the column detail drawing after being read and received. It can be seen that all values ​​are generated by the interpretation of entities.

[0106] (3) Entities are further divided into symbol entities and text entities. Text entities are text elements in CAD drawings. Symbol entities have more diverse sources and are generally formed by CAD line elements. They mainly include stirrup forms outlined by lines; longitudinal reinforcement points represented by circles, fills or regular polygons; and cross sections formed by lines surrounding stirrups and longitudinal reinforcements.

[0107] (4) Values ​​can be further divided into “column mark”, “specification value (static value)” and “range value”. The specification value is the steel bar specification (number of bars, diameter, arrangement and other steel bar business attributes). The column mark is the mark added to the specification during the writing of the steel bar specification, such as “C8@200(A6@200)” where the brackets are the mark added to the specification. The range value represents the column name (such as GZ1) and elevation (such as -0.050 to 2.950).

[0108] Based on the analysis of massive amounts of drawings and the entity relationship set, the following specifications for association operations are provided:

[0109] (1) Any value must be constrained to an entity, and an entity can also constrain another entity. Entities include longitudinal bars, stirrups, details, and other symbols and text annotations themselves.

[0110] (2) Only values ​​can be associated and merged, such as level 1-3, column labels, etc.

[0111] (3) Values ​​can constrain entities and values, but conversely, entities do not constrain values;

[0112] (4) Column labels belong to values;

[0113] (5) The values ​​of non-column labels are divided into static values ​​and range values. Among them, the name and elevation are range values, and the rest are static values.

[0114] (6) Column labels include serial number labels (1,2,3) and bracket labels (small brackets, square brackets, curly brackets, no brackets).

[0115] S24, based on the directed graph of the column outline, the column outline primitives and text entities are traversed and combined to obtain multiple sets of target column outlines. For detailed explanations, please refer to the relevant descriptions in the above embodiments; they will not be repeated here.

[0116] The column detail recognition method provided in this embodiment determines the existence of irrelevant text entities by detecting the correlation between each text entity and the column detail primitives. Irrelevant text entities are then removed to identify target associated entities. A corresponding directed graph of the column detail is then generated based on the target associated entities and the column detail primitives, ensuring the effective association between text entities and column detail primitives. By combining the constraint values ​​corresponding to each target associated entity to obtain multiple target constraint values, all constraint value combinations can be determined. These target constraint values ​​are then used to constrain the column detail primitives, ensuring that the target constraint values ​​satisfy all scenarios to the greatest extent possible.

[0117] This embodiment provides a method for identifying column samples, which can be used in electronic devices such as mobile phones, tablets, computers, servers, etc. Figure 13 This is a flowchart of a column sample identification method according to an embodiment of the present invention, such as... Figure 13 As shown, the process includes the following steps:

[0118] S31, Obtain the column detail drawing to be identified. For detailed instructions, please refer to the relevant descriptions in the above embodiments; they will not be repeated here.

[0119] S32, Identify the text information in the column detail drawing to obtain multiple text entities. For detailed explanations, please refer to the relevant descriptions in the above embodiments; they will not be repeated here.

[0120] S33: Associate multiple text entities with column detail elements in the column detail drawing to generate a directed column detail graph. For detailed explanations, please refer to the relevant descriptions in the above embodiments; they will not be repeated here.

[0121] S34. Based on the directed graph of the column outline, the column outline primitives and text entities are traversed and combined to obtain multiple sets of target column outlines.

[0122] Specifically, step S34 above may include:

[0123] S341, Determine constrained and unconstrained nodes based on the column-matrix directed graph.

[0124] Constrained nodes are nodes constrained by text entities, while unconstrained nodes are nodes without constraint information. By parsing the text entities and the column detail diagram in the generated column detail directed graph, multiple constrained nodes and multiple unconstrained nodes can be identified.

[0125] S342, starting from the constraint node that has no successor node, perform a reverse traversal to obtain the backtracking node.

[0126] Every reverse traversal must begin with a node that has no successor, and proceed in reverse order according to the association direction of the column-shaped directed graph until no node remains. All nodes encountered from the starting point to the end point are designated as backtracking nodes.

[0127] It should be noted that a starting constraint node is not allowed to contain multiple constraint values ​​(such as multiple tags and ranges). If it contains multiple constraint values, these values ​​need to be arranged and combined. When any constraint node contains multiple successor nodes of different types, all of these successor nodes must be traversed before the reverse traversal can continue.

[0128] S343, based on the information represented by the backtracking nodes and unconstrained nodes, generates a target column sample for each constrained node.

[0129] By combining all backtracking nodes with all unconstrained nodes, all target column outlines can be obtained. Specifically, if the directed graph of the column outline contains M incoherent ranges and N incoherent labels as constraint values, the directed graph of the column outline expresses M*N sets of information under the constraints of the ranges and labels. Based on the association relationship of the directed graph of the column outline, it is only necessary to randomly select one of the M range nodes, combine it with any one of the N labels, and add all the unconstrained nodes, and perform a reverse traversal until there are no parent nodes. The combination of the information of all nodes is the information expressed by the column outline under the current constraint. Based on this information, the target column outline under the current constraint can be generated.

[0130] Specifically, step S343 above may include:

[0131] (1) Obtain the display settings for backtracking nodes and unconstrained nodes.

[0132] Display settings are used to characterize the visualization of backtracking nodes, specifically including whether a node is visible or invisible. Technicians can configure the display settings for each backtracking node and unconstrained node, and the electronic device can respond to these settings by determining the display settings information for each node.

[0133] (2) Based on the display settings, control the display of information represented by backtracking nodes and unconstrained nodes to generate target column samples.

[0134] Based on the display settings of each node, nodes that are set to be visible can be visualized, while nodes that are set to be invisible can be hidden. This enables display control over backtracking nodes and unconstrained nodes, thereby generating the corresponding target column details.

[0135] This section combines the display settings of backtracking nodes and unconstrained nodes to control the display of the information represented by the backtracking nodes and unconstrained nodes in order to generate the target column detail, thereby ensuring the display integrity of the target column detail.

[0136] The column sample identification method provided in this embodiment uses a reverse traversal of constrained nodes to determine the backtracking nodes that can characterize the column sample information. The target column sample is determined by combining the information represented by the backtracking nodes and the unconstrained nodes, thus ensuring the integrity of the target column sample information.

[0137] This embodiment also provides a column sample identification device, which is used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0138] This embodiment provides a device for identifying column samples, such as... Figure 14 As shown, it includes:

[0139] The acquisition module 41 is used to acquire the column detail drawings to be identified.

[0140] The recognition module 42 is used to recognize the text information in the column detail drawing and obtain multiple text entities.

[0141] The association module 43 is used to associate multiple text entities with column detail elements in the column detail drawing to generate a directed column detail graph.

[0142] The generation module 44 is used to traverse and combine column outline primitives and text entities based on the directed graph of the column outline to obtain multiple sets of target column outlines.

[0143] Optionally, the aforementioned associated module 43 may include:

[0144] The first judgment submodule is used to determine whether each text entity is associated with the column outline element.

[0145] The removal submodule is used to remove text entities that are not associated with column outlines when one or more text entities are not associated with column outlines, thus obtaining multiple target associated entities.

[0146] The association submodule is used to associate multiple target associated entities with column detail primitives to generate a directed column detail graph.

[0147] Optionally, the aforementioned association submodule can be used to: obtain the constraint values ​​corresponding to each target associated entity; combine multiple constraint values ​​to obtain multiple target constraint values; and constrain the column detail primitives based on the multiple target constraint values ​​to obtain the column detail directed graph.

[0148] Specifically, when the constraint value can be an elevation segment, the above-mentioned association submodule can be used to: perform intersection processing and / or remainder processing on multiple elevation segments to obtain multiple target elevation segments.

[0149] Specifically, when the constraint value can be a name, the above-mentioned association submodule can be used to: determine the target constraint name from multiple names; integrate other names with the target constraint name to obtain an integrated name, which is consistent with the target constraint name.

[0150] Specifically, when the constraint value can be a label, the above-mentioned association submodule can be used to: classify different labels to obtain label classification results; and associate and combine labels based on the label classification results to generate target constraint values ​​corresponding to each label.

[0151] Optionally, the aforementioned associated module 43 may further include:

[0152] The second judgment submodule is used to determine whether there are duplicate constraint values ​​among multiple target constraint values.

[0153] The merge submodule is used to merge duplicate constraint values ​​when duplicate constraint values ​​exist, and obtain the target constraint value after merging.

[0154] Optionally, the above-mentioned generation module 44 may include:

[0155] The node determination submodule is used to determine constrained and unconstrained nodes based on the column-matrix directed graph.

[0156] The traversal submodule is used to perform reverse traversal starting from a constraint node that has no successor node, and obtain the backtracking node.

[0157] The generation submodule is used to generate target column samples for each constrained node based on the information represented by the backtracking node and the unconstrained node.

[0158] Optionally, the above-mentioned generation submodule is specifically used to: obtain the display settings for backtracking nodes and unconstrained nodes; control the display of the information represented by backtracking nodes and unconstrained nodes based on the display settings, and generate the target column sample.

[0159] In this embodiment, the column sample identification device is presented in the form of a functional unit. Here, a unit refers to an ASIC circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0160] The further functional descriptions of each module and submodule are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0161] The column detail recognition device provided in this embodiment identifies multiple text entities by recognizing text information in the column detail drawing, and associates each text entity with the column detail graphic elements to generate a directed graph of the column detail. This directed graph determines various information represented by the column detail graphic elements, facilitating a comprehensive interpretation of the information of each column detail in the drawing. By traversing and combining the column detail graphic elements and text entities through the directed graph, multiple sets of target column details are obtained, achieving comprehensive analysis of multiple meanings in a single column detail drawing. This supports column detail recognition in various scenarios and overcomes the limitations of traditional column detail recognition.

[0162] This invention also provides an electronic device having Figure 14 The identification device shown is for the column sample.

[0163] Please see Figure 15 , Figure 15 This is a schematic diagram of the structure of an electronic device provided in an optional embodiment of the present invention, such as... Figure 15As shown, the electronic device may include: at least one processor 501, such as a central processing unit (CPU), at least one communication interface 503, memory 504, and at least one communication bus 502. The communication bus 502 is used to enable communication between these components. The communication interface 503 may include a display screen or a keyboard; optionally, the communication interface 503 may also include a standard wired interface or a wireless interface. The memory 504 may be high-speed volatile random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory 504 may also be at least one storage device located remotely from the aforementioned processor 501. The processor 501 may be combined with... Figure 14 The described apparatus stores an application program in memory 504, and processor 501 calls the stored HA202208894-

[0164] The program code stored in device 504 is used to execute any of the above method steps.

[0165] The communication bus 502 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 502 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 15 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0166] The memory 504 may include volatile memory, such as random-access memory (RAM); the memory may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 504 may also include a combination of the above types of memory.

[0167] The processor 501 can be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and an NP.

[0168] The processor 501 may further include a hardware chip. This hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0169] Optionally, the memory 504 is also used to store program instructions. The processor 501 can invoke the program instructions to implement the column pattern identification method shown in the above embodiments of this application.

[0170] This invention also provides a non-transitory computer storage medium storing computer-executable instructions that can execute the column sample identification method in any of the above method embodiments. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.

[0171] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for identifying column samples, characterized in that, include: Obtain the detailed drawing of the column to be identified; The text information in the column detail drawing is identified to obtain multiple text entities; Associating multiple text entities with column detail elements in the column detail drawing to generate a directed column detail graph includes: determining whether each text entity is associated with a column detail element; When one or more of the text entities are not associated with the column outline, the text entities that are not associated with the column outline are removed to obtain multiple target associated entities; the constraint values ​​corresponding to each target associated entity are obtained; the multiple constraint values ​​are combined to obtain multiple target constraint values; the column outline is constrained based on the multiple target constraint values ​​to obtain the directed graph of the column outline; Based on the directed graph of the column outline, the column outline primitives and the text entities are traversed and combined to obtain multiple sets of target column outlines.

2. The method according to claim 1, characterized in that, When the constraint value is an elevation segment, the combination of multiple constraint values ​​to obtain multiple target constraint values ​​includes: Multiple target elevation segments are obtained by performing intersection and / or remainder operations on the multiple elevation segments.

3. The method according to claim 1, characterized in that, When the constraint value is a name, the combination of multiple constraint values ​​to obtain multiple target constraint values ​​includes: The target constraint name is determined from the plurality of names; The other names are combined with the target constraint name to obtain a combined name, which is consistent with the target constraint name.

4. The method according to claim 1, characterized in that, When the constraint value is marked, the combination of multiple constraint values ​​to obtain multiple target constraint values ​​includes: The different tags are classified to obtain the tag classification results; Based on the label classification results, the labels are associated and combined to generate target constraint values ​​corresponding to each label.

5. The method according to any one of claims 1-4, characterized in that, Before constraining the column detail primitives based on multiple target constraint values ​​to obtain the directed column detail graph, the process includes: Determine whether there are duplicate constraint values ​​among the multiple target constraint values; When duplicate constraint values ​​exist, the duplicate constraint values ​​are merged to obtain the target constraint value after merging.

6. The method according to claim 1, characterized in that, The process of traversing and combining the primitives and text entities based on the directed graph of the column layout yields multiple sets of target column layouts, including: Constrained and unconstrained nodes are determined based on the directed graph of the column outline. Starting from the constraint node that has no successor node, perform a reverse traversal to obtain the backtrack node; Based on the information represented by the backtracking nodes and the unconstrained nodes, a target column sample is generated for each constrained node.

7. The method according to claim 6, characterized in that, The generation of target column samples for each constrained node based on the information represented by the backtracking node and the unconstrained node includes: Obtain the display settings for the backtracking node and the unconstrained node; Based on the display settings, the information represented by the backtracking nodes and the unconstrained nodes is displayed and controlled to generate the target column sample.

8. A device for identifying column samples, characterized in that, include: The acquisition module is used to acquire the detailed drawings of the columns to be identified; The recognition module is used to recognize the text information in the column detail drawing and obtain multiple text entities; The association module is used to associate multiple text entities with column detail elements in the column detail drawing to generate a directed column detail graph. The generation module is used to traverse and combine the column pattern primitives and the text entities based on the column pattern directed graph to obtain multiple sets of target column patterns; The association module includes: a first judgment submodule, used to determine whether each text entity is associated with the column outline graphic element; a removal submodule, used to remove text entities that are not associated with the column outline graphic element when one or more text entities are not associated with the column outline, to obtain multiple target associated entities; and an association submodule, used to obtain constraint values ​​corresponding to each target associated entity; combine multiple constraint values ​​to obtain multiple target constraint values; and constrain the column outline graphic element based on the multiple target constraint values ​​to obtain the column outline directed graph.

9. An electronic device, characterized in that, include: A memory and a processor are interconnected, the memory stores computer instructions, and the processor executes the computer instructions to perform the column sample identification method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the column sample identification method according to any one of claims 1-7.

Images