Method and device for identifying information in structural drawing, storage medium and electronic equipment

Abstract

The invention relates to an information identification method and device in a structural drawing, a storage medium and electronic equipment. A method for identifying information in a structural drawing belongs to the technical field of building manufacturing and comprises the following steps of: cutting a target drawing frame by adopting a sliding window to obtain a plurality of blocks with preset sizes; the blocks are input into a target detection model, component positioning is carried out in combination with component graphic features of the beam structure components, and bounding box positions and category labels of the components are output; in the detection area where the bounding box position is located, target line segments conforming to component construction features in the detection area are extracted by combining the line segment form and the construction rule of the component; based on the relative position relation between the component and the beam structure, in combination with the marking line direction and the text information, target line segment screening and matching and binding of the target line segment and the corresponding marking information are completed, and therefore the drawing recognition and auditing efficiency can be improved.

Description

Technical Field

[0001] The present invention relates to the field of building manufacturing technology, and in particular to a method, device, storage medium and electronic equipment for identifying information in structural drawings. Background Art

[0002] As construction projects become increasingly complex and larger in scale, drawing review, a crucial step in the design phase, has a direct impact on the safety and compliance of project implementation. During the architectural design process, structural drawings convey key information such as the building's load-bearing system, structural joints, and reinforcement arrangements, serving as the core basis for ensuring structural safety and constructibility.

[0003] In practice, structural drawing review is typically performed manually by experienced technicians. While this approach offers a degree of flexibility, the large number of drawings and the multitude of details can easily lead to omissions and misjudgments, which can create engineering risks. This is particularly true when calculating the layout and quantity of structural components like beams, columns, and walls. This requires not only precise identification of component information but also accurate interpretation of reinforcement text, node connection methods, and dimensioning, placing significant demands on human resources. Therefore, a better approach to structural drawing review is needed. Summary of the Invention

[0004] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a method, device, storage medium and electronic device for identifying information in structural drawings, which solve the technical problem of difficulty in manual review of structural drawings in the prior art.

[0005] In order to achieve the above objectives, the main technical solutions adopted by the present invention include: A first aspect of an embodiment of the present invention provides a method for identifying information in a structural drawing.

[0006] The method for identifying information in a structural drawing proposed in an embodiment of the present invention includes: Using a sliding window to cut the target frame to obtain a plurality of blocks of predetermined size; wherein the target frame has beam structure information; Input the image block into the target detection model, combine the component graphic features of the beam structure component to locate the component, and output the bounding box position and category label of the component; Extracting target line segments that meet the structural characteristics of the component within the detection area based on the line segment morphology and structural rules of the component within the detection area; Based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information, the target line segment screening and the matching and binding of the target line segment with the corresponding annotation information are completed.

[0007] In some examples, the category labels include at least stirrups and hanger bars; The block is input into the target detection model, and the component is located in combination with the component graphic features of the beam structure component, and the bounding box position and category label of the component are output, including: The block is input into the target detection model, and the component is located in combination with the component graphic features of the beam structure component, and the first bounding box position of the stirrup and the second bounding box position of the hanger bar are output.

[0008] In some examples, the stirrups have a straight line segment form and a structural pattern of multiple straight line segments in parallel; The suspension bar has a broken line segment shape and a structural rule in which the broken line segment forms a trapezoidal representation; Extracting target line segments that meet the structural characteristics of the component within the detection area based on the line segment morphology and structural rules of the component within the detection area includes: In the detection area where the first bounding box is located, the stirrup segments that meet the stirrup structural characteristics are extracted in the detection area based on the straight line segment morphology of the stirrups and the structural law of the parallel characteristics of multiple straight line segments. In the detection area where the second bounding box is located, the hanger bar segments that meet the hanger bar structural characteristics in the detection area are extracted based on the broken line segment morphology of the hanger bar and the structural rule that the broken line segments form a trapezoidal representation.

[0009] In some examples, the target line segment screening and the matching and binding of the target line segment with the corresponding annotation information based on the relative position relationship between the component and the beam structure and the annotation line direction and text information are completed, including: Determine the stirrup target line segments that meet the relative position relationship between the stirrups and the beam structure, and complete the stirrup target line segment screening and the matching and binding of the stirrup target line segments with the corresponding annotation information in combination with the annotation line direction and text information; Determine the hanger target line segment in the hanger line segment that meets the relative position relationship between the hanger and the beam structure, and combine the annotation line direction and text information to complete the hanger target line segment screening and the matching and binding of the hanger target line segment with the corresponding annotation information.

[0010] In some examples, the steps of combining the annotation line direction with the text information to complete the selection of stirrup target line segments and the matching and binding of the stirrup target line segments with the corresponding annotation information include: Eliminate target line segments with incorrect annotation line directions and incorrect text information from the stirrup target line segments, complete stirrup target line segment screening, and match and bind stirrup target line segments with corresponding annotation information; The method of combining the direction of the annotation line with the text information to complete the selection of the target line segments of the hanger bar and the matching and binding of the target line segments of the hanger bar with the corresponding annotation information includes: Eliminate target line segments with incorrect annotation line directions and incorrect text information from the target line segments of the hanger bar, complete the screening of the target line segments of the hanger bar, and match and bind the target line segments of the hanger bar with the corresponding annotation information.

[0011] In some instances, the target detection model is a modified YOLOv5 model; The improved YOLOv5 model consists of a shallow feature layer and three deep feature layers, constructed based on the BiFPN structure; The three deep feature layers are arranged in ascending order from the second feature layer, the third feature layer and the fourth feature layer.

[0012] In some instances, the text information includes layer attribute information; The method of combining the annotation line direction and text information to complete the selection of stirrup target line segments and the matching and binding of stirrup target line segments with corresponding annotation information includes: Combined with the direction of the annotation line and the layer distribution of the stirrups, the target layer categories whose layer information accounts for more than a predetermined proportion are screened out, and the stirrup target line segments that do not belong to the target layer categories are removed. The stirrup target line segments are screened and matched and bound to the corresponding annotation information. The combination of the annotation line direction and text information to complete the selection of the hanger bar target line segments and the matching and binding of the stirrup target line segments with the corresponding annotation information includes: Combined with the direction of the annotation line and the layer distribution of the hanger bars, the target layer categories whose layer information accounts for more than a predetermined proportion are screened out, and the hanger bar target line segments that do not belong to the target layer categories are eliminated. The hanger bar target line segment screening and the matching and binding of the stirrup target line segments with the corresponding annotation information are completed.

[0013] A second aspect of an embodiment of the present invention provides a device for identifying information in a structural drawing, comprising: A frame cutting unit is used to cut the target frame using a sliding window to obtain a plurality of blocks of predetermined sizes; wherein the target frame has beam structure information; a component positioning unit, configured to input the image block into the target detection model, locate the component based on the component graphic features of the beam structure component, and output the bounding box position and category label of the component; A line segment extraction unit is configured to extract, within the detection area where the bounding box is located, target line segments that meet the structural characteristics of the component in the detection area based on the line segment morphology and structural rules of the component; The line segment matching unit is used to complete the target line segment screening and the matching and binding of the target line segment with the corresponding annotation information based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information.

[0014] A third aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when the program is executed by a processor.

[0015] A fourth aspect of an embodiment of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the first aspect when executing the program.

[0016] A method for identifying information in a structural drawing of the present invention comprises: using a sliding window to cut a target frame to obtain a plurality of blocks of predetermined sizes; wherein the target frame has beam structure information; inputting the blocks into a target detection model, locating the components in combination with component graphic features of the beam structure components, and outputting the bounding box position and category label of the components; within a detection area where the bounding box position is located, extracting target line segments that meet the structural features of the components in the detection area in combination with the line segment morphology and structural regularity of the components; and completing target line segment screening and matching and binding the target line segments with corresponding annotation information based on the relative positional relationship between the components and the beam structure and in combination with the direction of the annotation line and text information. In this application, component positioning is performed through a target detection model, and the bounding box position and category label of the component are output; within the detection area where the bounding box position is located, the target line segments that meet the structural characteristics of the component in the detection area are extracted in combination with the line segment morphology and construction rules of the component; based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information, the target line segments are screened and the target line segments are matched and bound with the corresponding annotation information. This is conducive to identifying the corresponding component line segments according to the line segment morphology and construction rules of different components, completing the target line segment screening and the matching and binding of the target line segments with the corresponding annotation information, thereby improving the efficiency of drawing recognition and review. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flow chart of a method for identifying information in a structural drawing provided by the prior art; Figure 2 A schematic diagram of the structure of an improved YOLOv5 model provided by an embodiment of the present invention; Figure 3 A flowchart of information identification in a structural drawing provided by an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of the information identification device in the structural drawing provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation methods in conjunction with the accompanying drawings.

[0019] The information recognition method in structural drawings proposed in an embodiment of the present invention is used to solve the technical problem of difficulty in manual review of structural drawings. Component positioning is performed through a target detection model, and the bounding box position and category label of the component are output; within the detection area where the bounding box position is located, the target line segments that meet the component structural characteristics in the detection area are extracted in combination with the line segment morphology and construction rules of the component; based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information, the target line segments are screened and the target line segments are matched and bounded with the corresponding annotation information. This is conducive to identifying the corresponding component line segments according to different component line segment morphologies and construction rules, completing the target line segment screening and the matching and binding of the target line segments with the corresponding annotation information, thereby improving the efficiency of drawing recognition and review.

[0020] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. Instead, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0021] Figure 1 The following is a flow chart of a method for identifying information in a structural drawing provided by the prior art. Figure 1 As shown, the method for identifying information in a structural drawing proposed in an embodiment of the present invention includes: Step 100: Using a sliding window to cut the target frame into multiple blocks of predetermined sizes; wherein the target frame has beam structure information; Step 110: Input the image block into the target detection model, locate the component based on the component graphic features of the beam structure component, and output the bounding box position and category label of the component; Step 120: extracting target line segments that meet the structural characteristics of the component within the detection area, combining the line segment morphology and structural rules of the component within the detection area; Step 130 : Based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information, target line segment screening and matching and binding of the target line segment with the corresponding annotation information are completed.

[0022] In this exemplary embodiment, a sliding window is used to perform frame cutting on the target frame to obtain a plurality of blocks of predetermined sizes, including: When using a sliding window to cut the target frame, the sliding window size is adjusted according to the distribution density of the components in the target frame; The target frame is cut based on the sliding window adjusted in real time to obtain a plurality of blocks of predetermined sizes.

[0023] The target frame is cut based on a sliding window adjusted in real time to obtain multiple blocks of predetermined sizes, including: Determine the priority of the area to be segmented based on the distribution density of the components in the target frame; Based on the priorities of the regions to be segmented, the target frame is segmented in descending order of density to obtain a plurality of blocks of predetermined sizes.

[0024] In this exemplary embodiment, the structural drawing can be a CAD drawing, and the drawing frame can be a sub-drawing frame of the CAD drawing. The sub-drawing frame is cut using a sliding window method so that the model can effectively adapt to the drawing frame input of different sizes. Secondly, the target detection method based on yolov5 is used to detect and recognize the main components of the beam quantity calculation and drawing (additional stirrups and hangers). Then, based on the existing drawing recognition results, the line segments of the entire sub-drawing frame are screened, and the existing beam results are used to determine the position of the additional stirrups and hangers. Finally, after obtaining the specific line segments and positions of the additional stirrups and hangers, the hangers are marked and matched.

[0025] This application can not only accurately identify additional stirrups, hanger bars and other components, but also match them with the corresponding annotation information, greatly improving work efficiency and accuracy.

[0026] In this exemplary embodiment, since the drawing size of the beam sub-frame is large, directly inputting it into the deep learning model for processing has problems of insufficient accuracy and high computational load. Therefore, this application uses a sliding window method to cut the sub-frame image and generate several predetermined-size blocks so that they can be more efficiently fed into the target detection model for processing.

[0027] Second, the preliminarily located additional stirrups and hanger bars in the cut-out images is performed based on the trained deep learning object detection model. This stage utilizes the typical graphical features of additional stirrups and hanger bars in the structural diagram to accurately identify their location areas.

[0028] Subsequently, based on the above-mentioned detection area, the line segment elements in the area are finely screened, and at the same time, other linear elements with similar characteristics in the entire image are analyzed to further extract possible additional stirrups and hanger bar line segments.

[0029] Then, through matching and judgment of graphic features such as line segment shape, positional relationship, and structural rules, the target line segments that truly meet the structural characteristics of additional stirrups and hanger bars are screened out.

[0030] Finally, for the identified hanger bar area, the direction of the annotation line and the text position in the figure are further combined to complete the matching and binding between the hanger bar and its corresponding annotation information.

[0031] In this application, component positioning is performed through a target detection model, and the bounding box position and category label of the component are output; within the detection area where the bounding box position is located, the target line segments that meet the structural characteristics of the component in the detection area are extracted in combination with the line segment morphology and construction rules of the component; based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information, the target line segments are screened and the target line segments are matched and bound with the corresponding annotation information. This is conducive to identifying the corresponding component line segments according to the line segment morphology and construction rules of different components, completing the target line segment screening and the matching and binding of the target line segments with the corresponding annotation information, thereby improving the efficiency of drawing recognition and review.

[0032] In some examples, the category labels include at least stirrups and hanger bars; The block is input into the target detection model, and the component is located in combination with the component graphic features of the beam structure component, and the bounding box position and category label of the component are output, including: The block is input into the target detection model, and the component is located in combination with the component graphic features of the beam structure component, and the first bounding box position of the stirrup and the second bounding box position of the hanger bar are output.

[0033] In this exemplary embodiment, stirrups and hanger bars are both beam structural components. The component graphic features of stirrups and hanger bars are different. During frame recognition, the image block can be input into the object detection model. Combined with the component graphic features of the beam structural components, the components are located, and the first bounding box position of the stirrups and the second bounding box position of the hanger bars are output, thereby achieving preliminary positioning of the additional stirrups and hanger bars in the cut-block image.

[0034] In some examples, the stirrups have a straight line segment form and a structural pattern of multiple straight line segments in parallel; The suspension bar has a broken line segment shape and a structural rule in which the broken line segment forms a trapezoidal representation; Extracting target line segments that meet the structural characteristics of the component within the detection area based on the line segment morphology and structural rules of the component within the detection area includes: In the detection area where the first bounding box is located, the stirrup segments that meet the stirrup structural characteristics are extracted in the detection area based on the straight line segment morphology of the stirrups and the structural law of the parallel characteristics of multiple straight line segments. In the detection area where the second bounding box is located, the hanger bar segments that meet the hanger bar structural characteristics in the detection area are extracted based on the broken line segment morphology of the hanger bar and the structural rule that the broken line segments form a trapezoidal representation.

[0035] In this exemplary embodiment, since the line segment morphology and structural patterns of stirrups and hanger bars are different, stirrup line segments that match the structural characteristics of stirrups and hanger bar segments within the inspection area can be extracted based on these line segment morphology and structural patterns. This allows the extraction of stirrup line segments and hanger bar line segments.

[0036] In some examples, the target line segment screening and the matching and binding of the target line segment with the corresponding annotation information based on the relative position relationship between the component and the beam structure and the annotation line direction and text information are completed, including: Determine the stirrup target line segments that meet the relative position relationship between the stirrups and the beam structure, and complete the stirrup target line segment screening and the matching and binding of the stirrup target line segments with the corresponding annotation information in combination with the annotation line direction and text information; Determine the hanger target line segment in the hanger line segment that meets the relative position relationship between the hanger and the beam structure, and combine the annotation line direction and text information to complete the hanger target line segment screening and the matching and binding of the hanger target line segment with the corresponding annotation information.

[0037] In this exemplary embodiment, the text information may include component number, reinforcement description, dimensioning, text notes, structural details and other information.

[0038] The method of combining the annotation line direction and text information to complete the selection of stirrup target line segments and the matching and binding of stirrup target line segments with corresponding annotation information includes: Obtain the size and proportion of the stirrup target line segment in the drawing frame and match them with the dimension annotation in the text information to complete the stirrup target line segment screening and the matching and binding of the stirrup target line segment with the corresponding annotation information.

[0039] The method of combining the direction of the annotation line with the text information to complete the selection of the target line segments of the hanger bar and the matching and binding of the target line segments of the hanger bar with the corresponding annotation information includes: The size and proportion of the target line segment of the hanger bar in the drawing frame are obtained and matched with the dimension annotation in the text information to complete the selection of the target line segment of the hanger bar and the matching and binding of the target line segment of the hanger bar with the corresponding annotation information.

[0040] In this exemplary embodiment, since the stirrups, hanger bars, and beam structures all have relative positions. These stirrups and hanger bars are located at predetermined positions near the beam structure. Therefore, target stirrup and hanger bar line segments that conform to the relative positions of the stirrups and beam structures are determined, and line segments that do not conform to the relative positions are eliminated. Line segments that conform to the relative positions are then combined with the annotation line direction and text information to complete the selection of hanger bar target line segments and the matching and binding of the hanger bar target line segments with the corresponding annotation information.

[0041] In some examples, the steps of combining the annotation line direction with the text information to complete the selection of stirrup target line segments and the matching and binding of the stirrup target line segments with the corresponding annotation information include: Eliminate target line segments with incorrect annotation line directions and incorrect text information from the stirrup target line segments, complete stirrup target line segment screening, and match and bind stirrup target line segments with corresponding annotation information; The method of combining the direction of the annotation line with the text information to complete the selection of the target line segments of the hanger bar and the matching and binding of the target line segments of the hanger bar with the corresponding annotation information includes: Eliminate target line segments with incorrect annotation line directions and incorrect text information from the target line segments of the hanger bar, complete the screening of the target line segments of the hanger bar, and match and bind the target line segments of the hanger bar with the corresponding annotation information.

[0042] In this exemplary embodiment, the target line segments with incorrect marking line directions and incorrect text information can be eliminated from the stirrup target line segments, and then the text information correction can be performed based on the remaining stirrup target line segments and hanger target line segments to complete the matching and binding of the corresponding marking information of the stirrup target line segments and hanger target line segments.

[0043] In some instances, the target detection model is a modified YOLOv5 model; The improved YOLOv5 model consists of a shallow feature layer and three deep feature layers, constructed based on the BiFPN structure; The three deep feature layers are arranged in ascending order from the second feature layer, the third feature layer and the fourth feature layer.

[0044] In this application, the improved YOLOv5 model includes: A new shallow feature extraction channel is added, and the feature fusion between shallow and deep layers is enhanced through the Bidirectional Feature Pyramid Network (BiFPN) structure.

[0045] Based on the YOLOv5 network structure, this application adds a shallow feature extraction channel (P2 layer) and enhances the feature fusion mechanism between shallow and deep layers (introducing the BiFPN structure) to preserve fine line texture information in the image, thereby improving the detection ability of small linear objects such as hanger bars and additional stirrups. This structural optimization not only improves the recall rate of small objects, but also maintains detection efficiency.

[0046] Combined with the size distribution of small objects in the training set, the K-means clustering method is used to reconfigure the default anchor settings of YOLOv5 to make it more suitable for the geometric scale of the additional stirrups and hanger symbols.

[0047] In this exemplary embodiment, Figure 2 Schematic diagram of an improved YOLOv5 model structure provided by an embodiment of the present invention. Figure 2 As shown in the figure, in the YOLOv5 + BiFPN architecture, the improved YOLOv5 model includes: the first feature layer P2, the second feature layer P3, the third feature layer P4, and the fourth feature layer P5; wherein the first feature layer P2, the second feature layer P3, the third feature layer P4, and the fourth feature layer P5 are feature maps representing different resolutions and semantic levels respectively; P2, P3, P4, and P5 represent feature maps of different resolutions and semantic levels respectively, supporting the detection of multi-scale objects from small to large targets. The details are as follows: P2 is a shallow, high-resolution feature map with a downsampling ratio of 4 (input ÷ 4), typically sized 160×160 (when the input is 640×640). It preserves the most spatial detail and is well-suited for small object detection. By introducing BiFPN, this feature map is fused with other layers in the neck, enhancing the representation of small objects.

[0048] P3 is a feature map with a downsampling rate of 8 (640 ÷ 8 = 80). As the first default detection layer, it integrates semantic and partial spatial information and is an important node in small and medium scale detection.

[0049] P4 corresponds to a downsampling rate of 16 (output is 40×40). It is more powerful in semantic expression, but has a lower spatial resolution than P3. It is a key layer for medium-scale object recognition.

[0050] P5 is a feature map with a downsampling rate of 32 (output is 20×20), which has the strongest semantic information but less spatial details and is suitable for detecting larger targets.

[0051] Through the above optimization measures, the model significantly improves the detection accuracy and robustness of small target components such as additional stirrups and hanger bars while maintaining a high detection speed.

[0052] In some instances, the text information includes layer attribute information; The method of combining the annotation line direction and text information to complete the selection of stirrup target line segments and the matching and binding of stirrup target line segments with corresponding annotation information includes: Combined with the direction of the annotation line and the layer distribution of the stirrups, the target layer categories whose layer information accounts for more than a predetermined proportion are screened out, and the stirrup target line segments that do not belong to the target layer categories are removed. The stirrup target line segments are screened and matched and bound to the corresponding annotation information. The combination of the annotation line direction and text information to complete the selection of the hanger bar target line segments and the matching and binding of the stirrup target line segments with the corresponding annotation information includes: Combined with the direction of the annotation line and the layer distribution of the hanger bars, the target layer categories whose layer information accounts for more than a predetermined proportion are screened out, and the hanger bar target line segments that do not belong to the target layer categories are eliminated. The hanger bar target line segment screening and the matching and binding of the stirrup target line segments with the corresponding annotation information are completed.

[0053] In this exemplary embodiment, each line segment has layer attributes. In this application, some or most line segments with different layer attributes may be eliminated. Based on the remaining line segments, the target line segments for the hanger bars are screened, and the target line segments for the stirrup bars are matched and bound to the corresponding annotation information. The predetermined ratio may be a percentage less than 20%.

[0054] In some examples, the annotation line includes an extended annotation line; and combining the annotation line direction with the text information to complete the stirrup target line segment screening and the matching and binding of the stirrup target line segment with the corresponding annotation information includes: Determine whether there is an extended annotation line based on geometric relationships; If there is an extension dimension line, analyze whether the extension dimension line is associated with the text element of the stirrup category or specification; If the extended annotation line is associated with the text element of the stirrup category or specification, the stirrup target line segment screening and the matching and binding of the stirrup target line segment with the corresponding annotation information are completed based on the association relationship between the extended annotation line and the text element of the stirrup category or specification.

[0055] In this exemplary embodiment, the annotation line includes an extended annotation line; and the steps of combining the annotation line direction with the text information to complete the selection of the hanger bar target line segments and the matching and binding of the hanger bar target line segments with the corresponding annotation information include: Determine whether there is an extended annotation line based on geometric relationships; If there is an extended dimension line, analyze whether the extended dimension line is associated with the text element of the hanger bar category or specification; If the extended annotation line is associated with the text element of the hanger bar category or specification, the hanger bar target line segment screening and the matching and binding of the hanger bar target line segment with the corresponding annotation information are completed based on the association relationship between the extended annotation line and the text element of the hanger bar category or specification.

[0056] In this exemplary embodiment, whether the extended dimension line is associated with a text element of the stirrup category or specification is analyzed, including: Analyze whether the extended dimension line is marked with text elements indicating the type or specification of stirrups; If the extended dimension line contains a text element indicating the stirrup category or specification, the extended dimension line is determined to be associated with the text element indicating the stirrup category or specification. Based on this association, the target stirrup segments are selected and matched and bound to the corresponding annotation information.

[0057] Figure 3 This is a flow chart of information identification in the structural drawings provided by the embodiment of the present invention. Figure 3 As shown in the figure, the information identification process in the structural drawing includes: Step 30: Start; Step 31: Cutting the sub-frame sliding window and detecting additional stirrups / hanger bars; Step 32: screening of detection results, screening of original image line segments, and screening of image features; Step 33: Match the anchor bar marking lines; Step 34, end.

[0058] An embodiment of the present invention provides a device for identifying information in a structural drawing. Figure 4 This is a schematic diagram of the information identification device structure in the structural drawing provided by the embodiment of the present invention. Figure 4 As shown, the information identification device in the structural drawing includes: A frame cutting unit 40 is configured to cut a target frame using a sliding window to obtain a plurality of blocks of predetermined sizes; wherein the target frame has beam structure information; A component positioning unit 41 is used to input the image block into the target detection model, locate the component based on the component graphic features of the beam structure component, and output the bounding box position and category label of the component; A line segment extraction unit 42 is configured to extract, within the detection area where the bounding box is located, target line segments that meet the structural characteristics of the component in the detection area based on the line segment morphology and structural rules of the component; The line segment matching unit 43 is used to complete the target line segment screening and the matching and binding of the target line segment with the corresponding annotation information based on the relative position relationship between the component and the beam structure and in combination with the annotation line direction and text information.

[0059] In this exemplary embodiment, a sliding window approach is used to cut into subframes, allowing the model to effectively adapt to input frames of varying sizes. Next, a target detection method based on YOLOv5 is used to detect and identify the main components of the beam quantity calculation (additional stirrups and hanger bars). Then, based on the existing identification results, the line segments of the entire subframe are screened, and the positions of the additional stirrups and hanger bars are determined using the existing beam results. Finally, after obtaining the specific line segments and positions of the additional stirrups and hanger bars, the hanger bar annotation information is matched.

[0060] This application can not only accurately identify additional stirrups, hanger bars and other components, but also match them with the corresponding annotation information, greatly improving work efficiency and accuracy.

[0061] In this exemplary embodiment, since the drawing size of the beam sub-frame is large, directly inputting it into the deep learning model for processing has problems of insufficient accuracy and high computational load. Therefore, this application uses a sliding window method to cut the sub-frame image and generate several predetermined-size blocks so that they can be more efficiently fed into the target detection model for processing.

[0062] Second, the preliminarily located additional stirrups and hanger bars in the cut-out images is performed based on the trained deep learning object detection model. This stage utilizes the typical graphical features of additional stirrups and hanger bars in the structural diagram to accurately identify their location areas.

[0063] Subsequently, based on the above-mentioned detection area, the line segment elements in the area are finely screened, and at the same time, other linear elements with similar characteristics in the entire image are analyzed to further extract possible additional stirrups and hanger bar line segments.

[0064] Then, through matching and judgment of graphic features such as line segment shape, positional relationship, and structural rules, the target line segments that truly meet the structural characteristics of additional stirrups and hanger bars are screened out.

[0065] Finally, for the identified hanger bar area, the direction of the annotation line and the text position in the figure are further combined to complete the matching and binding between the hanger bar and its corresponding annotation information.

[0066] In this application, component positioning is performed through a target detection model, and the bounding box position and category label of the component are output; within the detection area where the bounding box position is located, the target line segments that meet the structural characteristics of the component in the detection area are extracted in combination with the line segment morphology and construction rules of the component; based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information, the target line segments are screened and the target line segments are matched and bound with the corresponding annotation information. This is conducive to identifying the corresponding component line segments according to the line segment morphology and construction rules of different components, completing the target line segment screening and the matching and binding of the target line segments with the corresponding annotation information, thereby improving the efficiency of drawing recognition and review.

[0067] Since the systems / devices described in the above embodiments of the present invention are systems / devices used to implement the methods of the above embodiments of the present invention, those skilled in the art will be able to understand the specific structures and variations of these systems / devices based on the methods described in the above embodiments of the present invention, and thus will not be described in detail here. All systems / devices used in the methods of the above embodiments of the present invention are within the scope of protection of the present invention.

[0068] A third aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when the program is executed by a processor.

[0069] A fourth aspect of an embodiment of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the first aspect when executing the program.

[0070] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0071] In the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," "connect," "fixed," etc. should be understood broadly. For example, they may refer to fixed connection, detachable connection, or integration; mechanical connection or electrical connection; direct connection or indirect connection through an intermediate medium; and internal communication between two components or interaction between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0072] In the present invention, unless otherwise expressly specified or limited, when a first feature is "above" or "below" a second feature, it may mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. Furthermore, when a first feature is "above," "above," or "above" a second feature, it may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is at a higher level than the second feature. When a first feature is "below," "below," or "below" a second feature, it may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is at a lower level than the second feature.

[0073] In the description of this specification, the terms "one embodiment", "some embodiments", "embodiments", "examples", "specific examples" or "some examples" refer to the specific features, structures, materials or characteristics described in conjunction with the embodiment or example and included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and features of different embodiments or examples, unless they are mutually inconsistent.

[0074] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for identifying information in a structural drawing, characterized in that: include: Using a sliding window to cut the target frame to obtain a plurality of blocks of predetermined size; wherein the target frame has beam structure information; Input the image block into the target detection model, combine the component graphic features of the beam structure component to locate the component, and output the bounding box position and category label of the component; Extracting target line segments that meet the structural characteristics of the component within the detection area based on the line segment morphology and structural rules of the component within the detection area; Based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information, the target line segment screening and the matching and binding of the target line segment with the corresponding annotation information are completed.

2. The method according to claim 1, characterized in that The category label at least includes stirrups and hanger bars; The block is input into the target detection model, and the component is located in combination with the component graphic features of the beam structure component, and the bounding box position and category label of the component are output, including: The block is input into the target detection model, and the component is located in combination with the component graphic features of the beam structure component, and the first bounding box position of the stirrup and the second bounding box position of the hanger bar are output.

3. The method according to claim 2, characterized in that The stirrups have a straight line segment shape and a structural rule of multiple straight line segments in parallel; The suspension bar has a broken line segment shape and a structural rule in which the broken line segment forms a trapezoidal representation; Extracting target line segments that meet the structural characteristics of the component within the detection area based on the line segment morphology and structural rules of the component within the detection area includes: In the detection area where the first bounding box is located, the stirrup segments that meet the stirrup structural characteristics are extracted in the detection area based on the straight line segment morphology of the stirrups and the structural law of the parallel characteristics of multiple straight line segments. In the detection area where the second bounding box is located, the hanger bar segments that meet the hanger bar structural characteristics in the detection area are extracted based on the broken line segment morphology of the hanger bar and the structural rule that the broken line segments form a trapezoidal representation.

4. The method according to claim 3, characterized in that The target line segment screening and the matching and binding of the target line segment with the corresponding annotation information are completed based on the relative position relationship between the component and the beam structure and combined with the annotation line direction and text information, including: Determine the stirrup target line segments that meet the relative position relationship between the stirrups and the beam structure, and complete the stirrup target line segment screening and the matching and binding of the stirrup target line segments with the corresponding annotation information in combination with the annotation line direction and text information; Determine the hanger target line segment in the hanger line segment that meets the relative position relationship between the hanger and the beam structure, and combine the annotation line direction and text information to complete the hanger target line segment screening and the matching and binding of the hanger target line segment with the corresponding annotation information.

5. The method according to claim 4, characterized in that The method of combining the annotation line direction and text information to complete the selection of stirrup target line segments and the matching and binding of stirrup target line segments with corresponding annotation information includes: Eliminate target line segments with incorrect annotation line directions and incorrect text information from the stirrup target line segments, complete stirrup target line segment screening, and match and bind stirrup target line segments with corresponding annotation information; The method of combining the direction of the annotation line with the text information to complete the selection of the target line segments of the hanger bar and the matching and binding of the target line segments of the hanger bar with the corresponding annotation information includes: Eliminate target line segments with incorrect annotation line directions and incorrect text information from the target line segments of the hanger bar, complete the screening of the target line segments of the hanger bar, and match and bind the target line segments of the hanger bar with the corresponding annotation information.

6. The method according to any one of claims 1 to 5, characterized in that The target detection model is an improved YOLOv5 model; The improved YOLOv5 model consists of a shallow feature layer and three deep feature layers, constructed based on the BiFPN structure; The three deep feature layers are arranged in ascending order from the second feature layer, the third feature layer and the fourth feature layer.

7. The method according to claim 4, characterized in that The text information includes layer attribute information; The method of combining the annotation line direction and text information to complete the selection of stirrup target line segments and the matching and binding of stirrup target line segments with corresponding annotation information includes: Combined with the direction of the annotation line and the layer distribution of the stirrups, the target layer categories whose layer information accounts for more than a predetermined proportion are screened out, and the stirrup target line segments that do not belong to the target layer categories are removed. The stirrup target line segments are screened and matched and bound to the corresponding annotation information. The combination of the annotation line direction and text information to complete the selection of the hanger bar target line segments and the matching and binding of the stirrup target line segments with the corresponding annotation information includes: Combined with the direction of the annotation line and the layer distribution of the hanger bars, the target layer categories whose layer information accounts for more than a predetermined proportion are screened out, and the hanger bar target line segments that do not belong to the target layer categories are eliminated. The hanger bar target line segment screening and the matching and binding of the stirrup target line segments with the corresponding annotation information are completed.

8. A device for identifying information in a structural drawing, characterized in that: include: A frame cutting unit is used to cut the target frame using a sliding window to obtain a plurality of blocks of predetermined sizes; wherein the target frame has beam structure information; a component positioning unit, configured to input the image block into the target detection model, locate the component based on the component graphic features of the beam structure component, and output the bounding box position and category label of the component; A line segment extraction unit is configured to extract, within the detection area where the bounding box is located, target line segments that meet the structural characteristics of the component in the detection area based on the line segment morphology and structural rules of the component; The line segment matching unit is used to complete the target line segment screening and the matching and binding of the target line segment with the corresponding annotation information based on the relative position relationship between the component and the beam structure, combined with the annotation line direction and text information.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the method according to any one of claims 1 to 7 are implemented.

Images