Architectural drawing processing method and device and medium

Abstract

The invention discloses an architectural drawing processing method, which comprises the steps of extracting feature points of a vector object in an architectural drawing, and generating a feature vector comprising multi-dimensional features; setting DBSCAN clustering parameters, and performing DBSCAN clustering grouping on the feature vectors representing the feature points to obtain a plurality of clustering clusters; horizontal / vertical lines of each cluster are extracted, and effective building drawing elements are verified; on the basis of comprehensive judgment of geometric features and character features, identifying detailed drawing indexes and detailed drawing positions in each verified clustering cluster; and positioning a clustering cluster according to the graph number in the detail graph index, and associating the detail graph index with the detail graph data. According to the method, the DBSCAN clustering algorithm is combined with geometric feature and character feature analysis, each component and detail drawing index and position in the drawing can be segmented and recognized more accurately, the performance is excellent especially in complex detail parts, and the processing precision of the architectural drawing can be remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of image recognition and processing technology, and more specifically, to a method, device and medium for processing architectural drawings. Background Art

[0002] In the field of architectural design and drafting, architectural drawings are crucial tools, meticulously depicting a building's structure, layout, and details. With the advancement of computer technology and artificial intelligence, more and more architectural drawings are being digitized for easier storage, management, and modification. Furthermore, to improve efficiency, some architectural design software is incorporating intelligent features, such as automatic identification of components and nodes in drawings. The implementation of these intelligent features relies on the application of computer vision and artificial intelligence technologies.

[0003] Existing technical solutions primarily rely on image processing and pattern recognition to automatically segment and identify architectural drawings. For example, edge detection and connected domain analysis can be used to segment drawings into distinct components, such as walls, doors, windows, and stairs. These components are then subjected to feature extraction and classification to determine their type and location. Furthermore, deep learning can be used to train models to automatically identify components and nodes in drawings, accelerating drawing speed and accuracy.

[0004] However, the existing technical solutions still have the following problems in practical applications:

[0005] (1) Due to the complexity of architectural drawings, existing image processing and pattern recognition technologies often have difficulty accurately segmenting and identifying the various components in the drawings. In particular, when dealing with complex details, mis-segmentation and mis-identification are prone to occur.

[0006] (2) Existing technical solutions usually require a large amount of labeled data to train the model, which not only increases the workload but is also easily limited by data quality and quantity.

[0007] (3) Existing technical solutions often require different models and methods when processing different types of drawings, and lack versatility and flexibility.

[0008] (4) The existing solution is sensitive to the resolution of the drawing, and scaling can easily lead to feature distortion. It is also difficult to handle complex scenarios such as uncertain number of frames and closely connected frames. Summary of the Invention

[0009] In response to at least one defect or improvement need in the prior art, the present application provides a method, device and medium for processing architectural drawings, aiming to improve the accuracy and efficiency of intelligent processing of architectural drawings and achieve robust processing of complex drawing scenes.

[0010] To achieve the above objectives, in a first aspect, the present application provides a method for processing architectural drawings, comprising:

[0011] Extracting feature points of vector objects in architectural drawings and generating feature vectors including multi-dimensional features;

[0012] Set DBSCAN clustering parameters, perform DBSCAN clustering on the feature vectors representing the feature points, and obtain several clusters;

[0013] Extract the horizontal / vertical lines of each cluster to verify the valid architectural drawing elements;

[0014] Based on the comprehensive judgment of geometric features and text features, the detailed drawing index and detailed drawing location in each verified cluster are identified;

[0015] Locate clusters based on the drawing number in the detail index and associate the detail index with the detail data.

[0016] Furthermore, extracting feature points of vector objects in architectural drawings and generating feature vectors including multi-dimensional features includes:

[0017] Input architectural drawing vector data and extract the feature points of all vector objects; feature points include endpoints, midpoints, and intersections of lines, vertices of bounding boxes of block objects, text anchor points, and / or cell vertices and border intersections of tables;

[0018] A feature vector of a multi-dimensional feature including coordinates, direction angles and / or type labels is formed for each feature point.

[0019] Furthermore, setting the DBSCAN clustering parameters includes:

[0020] Neighborhood radius = α × average frame margin + β × direction difference threshold; the first parameter α is adjusted according to the standardization of the drawing; the second parameter β is considered to be the same type when it is ≤ 10°;

[0021] Minimum points ≥ 4;

[0022] Distance = λ1 × Euclidean distance + λ2 × direction cosine distance; when recognizing frames, the first weight coefficient λ1 > the second weight coefficient λ2; when recognizing tables, λ2 > λ1.

[0023] Furthermore, validating the architectural drawing elements includes drawing frame verification; the drawing frame verification specifically includes:

[0024] If there is a closed rectangle with four sides in the cluster, the text coordinates in the cluster are located within the cluster bounding box and include the frame feature words of "drawing number" and / or "project name", then it is judged to be a valid frame; the conditions for forming a closed rectangle include the error of the length of the opposite side is ≤5%, and the error of the angle between the adjacent sides relative to 90° is ≤5%.

[0025] Furthermore, validating the architectural drawing elements includes table verification; the table verification specifically includes:

[0026] If the row / column spacing difference of the grid points formed by the intersection of horizontal / vertical lines is ≤10%, the number of grid points in the area is ≥3×3, and the cells in the area include table feature words such as "serial number", "name" and / or "specification", then it is judged to be a valid table.

[0027] Furthermore, based on the comprehensive judgment of geometric features and text features, identifying the detailed drawing index and detailed drawing location in each verified cluster includes:

[0028] Get the circle data in the cluster, extract the text inside the circle, determine whether the text meets the index feature, whether there is a horizontal line passing through the center of the circle, and the position of the text relative to the line;

[0029] If there is a line in the circle that passes through the center and contains text, it is determined to be a detailed drawing index. The index number of the detailed drawing and the drawing number of the sub-drawing to which it belongs are determined based on the upper / lower position of the text on the horizontal line. The coordinate points are then indexed based on the end-to-end connection features of the line connected to the circle.

[0030] If there is no straight line passing through the circle and it only contains text, it is determined to be the detailed drawing location; the text alone is the index number of the detailed drawing.

[0031] Furthermore, locating the cluster according to the drawing number in the detail index and associating the detail index with the detail data includes:

[0032] According to the drawing number in the detailed drawing index, the corresponding sub-drawing is searched using the frame information obtained by DBSCAN clustering. If there is no drawing number, it is assumed that the detailed drawing belongs to the current sub-drawing.

[0033] Find the detailed drawing data corresponding to the detailed drawing index in the sub-drawing, use the dictionary data structure to store the detailed drawing index and detailed drawing data, find the corresponding detailed drawing data by the index number, and establish the association between the detailed drawing index and the detailed drawing data.

[0034] Furthermore, adaptive parameter adjustment, multi-density region processing and / or error correction mechanisms are also included;

[0035] The adaptive parameter adjustment includes: counting the minimum bounding box side length of the valid frame in the historical data and calculating the average neighborhood distance, dynamically initializing the neighborhood radius; using the X-Means algorithm to determine the optimal number of clusters and assist in adjusting the minimum number of points;

[0036] The multi-density region processing includes: firstly performing a preliminary grouping of closely connected frames by hierarchical clustering, and then performing DBSCAN clustering grouping within the group;

[0037] The error correction mechanism includes: for small frames that may be omitted in the clustering results, the neighborhood range is expanded through the expansion and corrosion algorithm, and the edge points are re-detected to perform error correction.

[0038] In a second aspect, the present application provides an electronic device comprising at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program, and when the computer program is executed by the processing unit, the processing unit is enabled to perform the steps of any of the aforementioned architectural drawing processing methods.

[0039] In a third aspect, the present application provides a storage medium storing a computer program executable by an access authentication device. When the computer program runs on the access authentication device, the access authentication device is enabled to execute the steps of any of the aforementioned architectural drawing processing methods.

[0040] In general, the above technical solutions conceived by this application can achieve the following beneficial effects compared with the existing technology:

[0041] (1) This application uses the DBSCAN clustering algorithm combined with geometric feature and text feature analysis to more accurately segment and identify the various components, detailed drawing indexes, and locations in the drawings. It performs particularly well in complex details and can significantly improve the processing accuracy of architectural drawings.

[0042] (2) This application utilizes the unsupervised clustering characteristics of DBSCAN to achieve drawing segmentation and feature recognition without the need for a large amount of labeled data, thereby reducing the cost of data labeling, being free from the limitations of data quality and quantity, and improving the generalization ability of the model.

[0043] (3) This application is based on density clustering and geometric feature fusion, and is suitable for drawings of different types and resolutions, such as civil engineering and non-civil engineering. It has stronger robustness for scenarios with uncertain number of frames and closely connected frames, and has greater versatility and flexibility.

[0044] (4) This application performs clustering directly based on vector coordinates, avoiding feature distortion caused by scaling and improving the consistency and accuracy of processing drawings with different resolutions.

[0045] (5) This application can quickly and accurately establish the association between the detailed drawing index and the detailed drawing data, allowing users to find the node practices of the components more quickly, improving work efficiency and reducing errors and troubles in work. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 A core flow chart of a method for processing architectural drawings provided in an embodiment of the present application;

[0048] Figure 2 A schematic diagram of a valid table provided in an embodiment of the present application;

[0049] Figure 3 A schematic diagram of a detailed index provided for an embodiment of the present application;

[0050] Figure 4 A schematic diagram of the detailed positions provided for the embodiments of this application;

[0051] Figure 5 A block diagram of a device suitable for implementing the architectural drawing processing method described above, provided in an embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely for the purpose of explaining this application and are not intended to limit this application. In addition, the technical features involved in the various embodiments of this application described below may be combined with each other as long as they do not conflict with each other.

[0053] The terms "first," "second," or "nth" in the specification, claims, or drawings of this application are used to distinguish different objects, not to describe a specific order. In addition, the terms "including," "having," and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to the process, method, product, or apparatus.

[0054] refer to Figure 1 One embodiment of the present application proposes a method for processing architectural drawings based on DBSCAN clustering and geometric feature recognition. The process outline of the architectural drawing processing method is as follows:

[0055] A[Drawing vector data] --> B[Feature point extraction and normalization]

[0056] B --> C [DBSCAN clustering]

[0057] C --> D{Is it a plot frame / table cluster?}

[0058] D -->|Yes| E[Generate envelope and take screenshot]

[0059] D -->|No|F[Mark as noise or other object]

[0060] E --> G [OCR extracts text features]

[0061] G --> H[Multimodal model validation (frame title / table structure)]

[0062] H --> I[Coordinate inverse calculation and drawing positioning]

[0063] I --> J[output sub-image results]

[0064] J --> K[Detailed drawing index and location identification]

[0065] K --> L[Establish association between detail index and detail data]

[0066] L --> M [Adaptive parameter adjustment]

[0067] M --> N [Multi-density area processing]

[0068] N --> O[Error Correction Mechanism]

[0069] O --> P[Output final processing result]

[0070] Specifically, the following steps may be included.

[0071] Step 1: Extract feature points of vector objects in architectural drawings and generate feature vectors including multi-dimensional features.

[0072] In some embodiments, specifically, architectural drawing vector data is input and feature points of all vector objects are extracted, including endpoints, midpoints, and intersections of lines, bounding box vertices of block objects, text anchor points, and cell vertices and border intersections of tables.

[0073] A feature vector is formed for each feature point, which contains multi-dimensional features such as coordinates (x, y), direction angle, type label (frame line / table line / other).

[0074] The scaling ratio is obtained through the drawing metadata, all coordinates are converted to a unified scale (such as millimeters), and noise points such as short line segments and isolated points with lengths less than the preset threshold are filtered out to complete the preprocessing of coordinate normalization.

[0075] This application performs clustering directly based on vector coordinates, avoiding feature distortion caused by scaling and improving the consistency and accuracy of processing drawings with different resolutions.

[0076] Step 2: Set the DBSCAN clustering parameters, perform DBSCAN clustering on the feature vectors representing the feature points, and obtain several clusters.

[0077] DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is a density-based spatial clustering algorithm that can classify data points into core points, boundary points, and noise points, thereby identifying clusters of arbitrary shapes. It can also handle noise points and discover clusters of arbitrary shapes. In DBSCAN, each data point is a feature vector, which form the basis of the dataset.

[0078] In some embodiments, specifically, based on the pre-processed feature point data, the DBSCAN clustering parameters are set:

[0079] EPS (neighborhood radius): Combined with geometric significance, it considers both coordinate distance and directional similarity, and is adjusted according to the drawing type. EPS = α × average drawing frame margin + β × directional difference threshold. α is adjusted based on the drawing standardization (non-civil engineering drawings can be appropriately increased). Lines with a directional difference of β ≤ 10° are considered to be of the same type.

[0080] min_samples (minimum number of points): Considering the structural characteristics of the frame / table, at least 4 vertices are required, so the minimum number of points is set to ≥ 4.

[0081] Distance metric: d(p,q) = λ1 × Euclidean distance + λ2 × direction cosine distance. Frame recognition prioritizes coordinate distance (λ1 has a higher weight), while table recognition prioritizes directional consistency (λ2 has a higher weight).

[0082] After the DBSCAN clustering parameters are set, the DBSCAN clustering algorithm is executed to group the drawing vector data into different clusters to obtain several clusters.

[0083] This application utilizes the unsupervised clustering characteristics of DBSCAN to achieve drawing segmentation and feature recognition without the need for a large amount of labeled data, thereby reducing the cost of data annotation, being free from restrictions on data quality and quantity, and improving the generalization ability of the model.

[0084] Step 3: Extracting horizontal / vertical lines from each cluster to verify valid architectural drawing elements. In some embodiments, this specifically includes the following.

[0085] Each cluster was verified by extracting horizontal and vertical lines (angles: 0°, 90°, and 180°) and checking whether they formed a four-sided enclosed rectangle (opposite sides were parallel and equal, adjacent sides were perpendicular, and the error was ≤5%). Optical Character Recognition (OCR) was also performed on the text within the cluster area to detect the presence of keywords such as "drawing title," "drawing number," and "project name." If the text coordinates fell within the cluster's bounding box and the keywords matched, the bounding box was considered valid.

[0086] Perform table verification on clusters suspected of being tables: filter horizontal / vertical lines, count the grid points formed by the intersection of horizontal / vertical lines (row / column spacing difference ≤ 10%), and if the number of grid points in the area is ≥ 3×3, it is determined to be a table candidate; check whether the cells in the area contain table feature words such as "serial number", "name", and "specification", and verify the row and column alignment relationship based on the OCR results to confirm the table. Figure 2 .

[0087] Step 4: Based on the comprehensive judgment of geometric features and text features, identify the detailed drawing index and detailed drawing location in each verified cluster.

[0088] In some embodiments, pattern recognition technology is used to identify detailed image indexes and locations based on the geometric and text features of the sub-images segmented by DBSCAN clustering. Circle data is obtained, and the text within the circle is extracted to determine whether the text meets the index features, whether there is a horizontal line passing through the center of the circle, and the position of the text relative to the line.

[0089] If there is a horizontal line in the circle that passes through the center and contains text, it is determined to be a detailed drawing index. According to whether the text is above or below the horizontal line, the index number of the detailed drawing and the drawing number of the sub-drawing where it is located are determined. Then, according to the feature index coordinate points connected to the beginning and end of the line connected to the circle, refer to Figure 3 .

[0090] If there is no straight line passing through the circle and the circle only contains text, it is determined to be the location of the detailed drawing. The text is the index number of the detailed drawing. Figure 4 .

[0091] Step 5: Locate the cluster according to the drawing number in the detail index, and associate the detail index with the detail data.

[0092] In some embodiments, specifically, according to the drawing number in the detailed drawing index, the corresponding sub-drawing is searched using the drawing frame information obtained by DBSCAN clustering and segmentation. If there is no drawing number, it is assumed that the detailed drawing belongs to the current sub-drawing.

[0093] Find the detailed drawing data corresponding to the detailed drawing index in the sub-drawing, use the dictionary data structure to store the detailed drawing index and detailed drawing data, find the corresponding detailed drawing data by the index number, and establish the association between the detailed drawing index and the detailed drawing data.

[0094] This application can quickly and accurately establish the association between the detailed drawing index and the detailed drawing data, allowing users to find the node practices of the components more quickly, improving work efficiency and reducing errors and troubles in the work.

[0095] Preferably, in some embodiments, the architectural drawing processing method may further include adaptive parameter adjustment, multi-density area processing and / or error correction mechanism.

[0096] Adaptive parameter adjustment: For different types of drawings (civil engineering / non-civil engineering), the minimum bounding box side length of the valid drawing frame in the historical data is counted, the average neighborhood distance is calculated, the neighborhood radius eps is dynamically initialized, and algorithms such as X-Means are used to automatically determine the optimal number of clusters, assist in adjusting min_samples, and achieve adaptive parameter adjustment.

[0097] Multi-density area processing: For closely connected frames, hierarchical clustering is first performed for preliminary grouping, and then DBSCAN clustering is performed on the data within each group, thereby avoiding cross-frame misclustering.

[0098] Error correction mechanism: For small frames that may be missed in the clustering results, the neighborhood range is expanded through the expansion and corrosion algorithm, and the edge points are re-detected to perform error correction.

[0099] The above steps can be completed on a computer with regular computing power, and the required parameters such as algorithm parameters and model parameters can be adjusted according to specific circumstances.

[0100] DBSCAN-based drawing segmentation, detail recognition, association establishment, and algorithm optimization technologies can be integrated into a single system to form an intelligent architectural drawing processing system. Based on the input architectural drawings, this system can automatically complete operations such as vector feature extraction, DBSCAN clustering, frame / table verification, detail index and location recognition, association establishment, and provide reference for component node practices.

[0101] This application is based on density clustering and geometric feature fusion, and is suitable for drawings of different types and resolutions, such as civil engineering and non-civil engineering. It has stronger robustness, versatility and flexibility for scenarios such as uncertain number of frames and closely connected frames.

[0102] This application uses the DBSCAN clustering algorithm combined with geometric feature and text feature analysis to more accurately segment and identify the various components, detailed drawing indexes, and locations in the drawings. It performs particularly well in complex details and can significantly improve the processing accuracy of architectural drawings.

[0103] This technical solution applies the DBSCAN clustering algorithm to architectural drawing processing, representing a novel application of image processing and pattern recognition technologies. Through intelligent segmentation and feature recognition of architectural drawings, it further promotes the application and development of computer vision technology in complex engineering drawing scenarios. Furthermore, the techniques involved, such as density-based clustering and vector feature processing, are universal and can be applied and promoted in other fields, such as medical image processing and natural scene recognition.

[0104] This technical solution, a specific application of AI in the construction field, promotes the application of AI in architectural design, construction, and management through intelligent processing of architectural drawings, thereby improving the intelligence and automation level of the construction industry. The AI ​​technologies involved, such as unsupervised learning and adaptive parameter adjustment, offer flexibility and can be expanded to other fields such as transportation, logistics, and retail.

[0105] In general, with the development of computer technology and artificial intelligence, the demand for intelligent processing of architectural drawings is increasing. This technical solution has broad application prospects and market demand in the fields of architectural design and drawing, computer vision technology, and artificial intelligence applications.

[0106] Figure 5 The following schematically shows a block diagram of a device suitable for implementing the architectural drawing processing method described above according to an embodiment of the present application. Figure 5 The device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0107] like Figure 5 As shown, the device 1000 described in this embodiment includes: a processor 1001, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 1002 or the program loaded from the storage part 1008 to the random access memory (RAM) 1003. The processor 1001 may include, for example, a general-purpose microprocessor (such as a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (such as an application-specific integrated circuit (ASIC)), etc. The processor 1001 may also include on-board memory for caching purposes. The processor 1001 may include a single processing unit or multiple processing units for executing different actions of the architectural drawing processing method according to the embodiment of the present application.

[0108] RAM 1003 stores various programs and data required for the operation of device 1000. Processor 1001, ROM 1002, and RAM 1003 are interconnected via bus 1004. Processor 1001 executes the programs stored in ROM 1002 and / or RAM 1003 to perform the various operations of the architectural drawing processing method according to the embodiment of the present application. It should be noted that the programs may also be stored in one or more memories other than ROM 1002 and RAM 1003. Processor 1001 may also execute the programs stored in the one or more memories to perform the various operations of the architectural drawing processing method according to the embodiment of the present application.

[0109] According to an embodiment of the present application, device 1000 may further include an input / output (I / O) interface 1005, which is also connected to bus 1004. Device 1000 may also include one or more of the following components connected to I / O interface 1005: an input section 1006 including a keyboard, mouse, etc.; an output section 1007 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and speakers; a storage section 1008 including a hard disk; and a communication section 1009 including a network interface card such as a LAN card or a modem. Communication section 1009 performs communication processing via a network such as the Internet. Drive 1010 is also connected to I / O interface 1005 as needed. Removable media 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed in drive 1010 as needed, so that computer programs read from the removable media can be installed into storage section 1008 as needed.

[0110] The architectural drawing processing method process according to an embodiment of the present application can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program contains program code for executing the architectural drawing processing method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1009, and / or installed from the removable medium 1011. When the computer program is executed by the processor 1001, the above-mentioned functions defined in the system of the embodiment of the present application are performed. According to an embodiment of the present application, the systems, devices, means, modules and / or units described above can be implemented by computer program modules.

[0111] Embodiments of the present application also provide a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments, or may exist independently and not be incorporated into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the steps of the architectural drawing processing method according to the embodiments of the present application.

[0112] According to an embodiment of the present application, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In an embodiment of the present application, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present application, the computer-readable storage medium may include one or more memories other than the ROM 1002 and / or RAM 1003 described above.

[0113] It should be noted that the functional modules in the various embodiments of the present application can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product.

[0114] The flowcharts and / or block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flowchart and / or block diagram can represent a module, a program segment or a part of code, and the part of the above-mentioned module, program segment or code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. It should also be noted that each box in the block diagram or flowchart, and the combination of the boxes in the block diagram or flowchart, can be implemented by a special hardware-based system that performs the specified function or operation, or can be implemented by a combination of special hardware and computer instructions.

[0115] Those skilled in the art will appreciate that the features described in the various embodiments and / or claims of this application may be combined and / or coupled in various ways, even if such combinations and / or couplings are not explicitly described in this application. In particular, without departing from the spirit and teachings of this application, the technical features described in the various embodiments and / or claims of this application may be combined and / or coupled in various ways, and all such combinations and / or couplings fall within the scope of this application.

[0116] Although the present application has been shown and described with reference to certain exemplary embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made to the present application without departing from the spirit and scope of the present application as defined by the appended claims and their equivalents. Therefore, the scope of the present application should not be limited to the above-described embodiments, but should be determined not only by the appended claims but also by the equivalents of the appended claims.

Claims

1. A method for processing architectural drawings, characterized in that: include: Extracting feature points of vector objects in architectural drawings and generating feature vectors including multi-dimensional features; Set DBSCAN clustering parameters, perform DBSCAN clustering on the feature vectors representing the feature points, and obtain several clusters; Extract the horizontal / vertical lines of each cluster to verify the valid architectural drawing elements; Based on the comprehensive judgment of geometric features and text features, the detailed drawing index and detailed drawing location in each verified cluster are identified; Locate clusters based on the drawing number in the detail index and associate the detail index with the detail data.

2. The architectural drawing processing method according to claim 1, wherein: Extracting feature points of vector objects in architectural drawings and generating feature vectors containing multi-dimensional features include: Input architectural drawing vector data and extract the feature points of all vector objects; feature points include endpoints, midpoints, and intersections of lines, vertices of bounding boxes of block objects, text anchor points, and / or cell vertices and border intersections of tables; A feature vector of a multi-dimensional feature including coordinates, direction angles and / or type labels is formed for each feature point.

3. The architectural drawing processing method according to claim 1, wherein: Setting DBSCAN clustering parameters includes: Neighborhood radius = α × average frame margin + β × direction difference threshold; the first parameter α is adjusted according to the standardization of the drawing; the second parameter β is considered to be the same type when it is ≤ 10°; Minimum points ≥ 4; Distance = λ1 × Euclidean distance + λ2 × direction cosine distance; when recognizing frames, the first weight coefficient λ1 > the second weight coefficient λ2; when recognizing tables, λ2 > λ1.

4. The architectural drawing processing method according to claim 1, wherein: Verification of valid architectural drawing elements includes drawing frame verification; the drawing frame verification specifically includes: A valid drawing frame is determined if a four-sided enclosed rectangle exists within the cluster, the text coordinates within the cluster are within the cluster bounding box, and the drawing frame feature words "drawing number" and / or "project name" are included. Conditions for forming an enclosed rectangle include the error in the length of opposite sides being ≤5% and the error in the angle between adjacent sides relative to 90° being ≤5%.

5. The architectural drawing processing method according to claim 1, wherein: Verification of valid architectural drawing elements includes form verification; the form verification specifically includes: A table is considered valid if the row / column spacing difference of the grid points formed by the intersection of horizontal / vertical lines is ≤10%, the number of grid points in the area is ≥3×3, and the cells in the area include table feature words such as "serial number", "name" and / or "specification".

6. The architectural drawing processing method according to claim 1, wherein: Based on the comprehensive judgment of geometric features and text features, the identification of the detailed drawing index and detailed drawing location in each verified cluster includes: Get the circle data in the cluster, extract the text inside the circle, determine whether the text meets the index feature, whether there is a horizontal line passing through the center of the circle, and the position of the text relative to the line; If there is a line in the circle that passes through the center and contains text, it is determined to be a detailed drawing index. The index number of the detailed drawing and the drawing number of the sub-drawing to which it belongs are determined based on the upper / lower position of the text on the horizontal line. The coordinate points are then indexed based on the end-to-end connection features of the line connected to the circle. If there is no straight line passing through the circle and it only contains text, it is determined to be the detailed drawing location; the text alone is the index number of the detailed drawing.

7. The architectural drawing processing method according to claim 1, wherein: Locate clusters based on the drawing number in the detail callout. Associate the detail callout with the detail data, including: According to the drawing number in the detailed drawing index, the corresponding sub-drawing is searched using the frame information obtained by DBSCAN clustering. If there is no drawing number, it is assumed that the detailed drawing belongs to the current sub-drawing. Find the detailed drawing data corresponding to the detailed drawing index in the sub-drawing, use the dictionary data structure to store the detailed drawing index and detailed drawing data, find the corresponding detailed drawing data by the index number, and establish the association between the detailed drawing index and the detailed drawing data.

8. The architectural drawing processing method according to claim 1, wherein: Also included are adaptive parameter adjustments, multi-density region processing, and / or error correction mechanisms; The adaptive parameter adjustment includes: counting the minimum bounding box side length of the valid frame in the historical data and calculating the average neighborhood distance, dynamically initializing the neighborhood radius; using the X-Means algorithm to determine the optimal number of clusters and assist in adjusting the minimum number of points; The multi-density region processing includes: firstly performing a preliminary grouping of closely connected frames by hierarchical clustering, and then performing DBSCAN clustering grouping within the group; The error correction mechanism includes: for small frames that may be omitted in the clustering results, the neighborhood range is expanded through the expansion and corrosion algorithm, and the edge points are re-detected to perform error correction.

9. An electronic device, characterized in that: The method comprises at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program, and when the computer program is executed by the processing unit, the processing unit is enabled to perform the steps of the architectural drawing processing method according to any one of claims 1 to 8.

10. A storage medium, characterized in that: It stores a computer program executable by an access authentication device. When the computer program runs on the access authentication device, the access authentication device is enabled to execute the steps of the architectural drawing processing method according to any one of claims 1 to 8.

Images