Electronic drawing verification method and system and storage medium

Abstract

The invention provides an electronic drawing verification method and system and a storage medium. The verification method comprises the following steps: importing a group of electronic drawings; wherein the group of electronic drawings comprises a directory page drawing and at least one other electronic drawing; using a first model obtained by training to identify directory page data of the directory page drawing, verifying the archiving condition of the group of electronic drawings based on the directory page data, and outputting a first verification result of the archiving condition of the group of electronic drawings; and outputting a second verification result of the archiving condition of the group of electronic drawings according to feedback input of the user to the first verification result. According to the verification method, the data integrity and accuracy of the electronic drawings in the conversion and transmission process can be ensured, the dependence on manual verification is reduced, the verification efficiency of the electronic drawings is improved, and the reliability and interactivity of the workflow can be enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic drawing processing, and in particular to a verification method, system and storage medium for electronic drawings. Background Art

[0002] At present, the whole process of design, construction, completion and archiving of pilot projects in many places adopts the closed-loop management of "one set of drawings", emphasizing the importance of digital drawings as the basis for supervision and inspection, acceptance inspection, quality and safety accident investigation, etc. of relevant departments. In the process of processing electronic drawings, it is necessary to check the basic information of the drawings (such as design parameters, annotations, drawing names, drawing numbers, drawing sizes, dates, etc.) to ensure the integrity, availability and consistency of the drawings. The traditional processing mode is manual proofreading, which is time-consuming, inefficient and low in accuracy. At the same time, there will be errors in the recording process, such as wrong data input, missing information or format errors, especially when processing a large amount of complex drawing data. These errors may directly affect the quality review and final acceptance of the project, and increase the cost and time of rework and modification.

[0003] In this context, the traditional mode of manually checking the integrity, consistency and recording of drawings is no longer suitable for the efficient and high-quality management requirements of digital drawings. Therefore, developing an efficient and accurate electronic drawing verification method is crucial to ensure the accuracy and reliability of design drawings and the archiving results of electronic drawings. Summary of the invention

[0004] Based on this, the present invention aims to provide an improvement to solve at least one of the above problems.

[0005] In a first aspect, the present application provides a method for verifying an electronic drawing, comprising: Importing a set of electronic drawings; wherein the set of electronic drawings includes a catalog page drawing and at least one other electronic drawing; Using the trained first model to identify the catalog page data of the catalog page drawing, and verifying the filing status of the group of electronic drawings based on the catalog page data, and outputting a first verification result of the filing status of the group of electronic drawings; Based on the user's feedback input on the first verification result, a second verification result of the filing status of the group of electronic drawings is output.

[0006] The above method completes the recognition of the catalog page data of the catalog page drawing through the first model obtained through training, and uses it to verify the filing status of the group of electronic drawings, outputs a first verification result, and then forms feedback input through the user's confirmation and editing of the first verification result to obtain a more accurate second verification result. This verification method can ensure the data integrity and accuracy of electronic drawings during the conversion and transmission process through automated verification technology, reduce dependence on manual verification, and improve the verification efficiency of electronic drawings; on the other hand, it can significantly reduce errors caused by manual entry and enhance the reliability of the workflow; in addition, it can also enhance interactivity, facilitate users to view verification results in real time and edit and modify them directly on the interface, improve user experience, and enable non-professionals to effectively verify electronic drawings.

[0007] In one of the embodiments, the method further includes: training and optimizing the first model based on the second verification result to obtain a second model, wherein the second model is configured to verify the next set of electronic drawings.

[0008] In one of the embodiments, the verifying of the filing status of the group of electronic drawings based on the catalog page data includes: verifying based on the catalog page data whether the group of electronic drawings is missing electronic drawings recorded in the catalog page data and / or has extra electronic drawings not recorded in the catalog page data; and, verifying based on the catalog page data whether the cataloging data of other electronic drawings is consistent with that recorded in the catalog page data.

[0009] In one embodiment, the first model is determined according to a first training process, and the first training process includes: obtaining a first training sample; wherein the first training sample includes a variety of verified electronic drawings and historical verification results, and the historical verification results include error examples and correct examples; based on the first training sample, an initial model is trained to obtain the first model; wherein the initial model is an OCR model that applies a deep learning model.

[0010] In one embodiment, the initial model is trained based on the first training sample to obtain the first model, including: training the initial model to recognize the catalog page data of the electronic drawing and the frame information data of other electronic drawings in the first training sample; training the initial model to distinguish between erroneous examples and correct examples of historical verification results in the first training sample.

[0011] In one embodiment, the training of the initial model to identify the catalog page data of the electronic drawing and the frame information data of other electronic drawings in the first training sample includes: training the initial model to extract one or more of the text feature data, border feature data, table line feature data, symbol feature data, drawing index feature data, serial number data, drawing number data, format feature data, and spatial relationship feature data of the electronic drawing and text in the first training sample; and training the initial model to identify the catalog page data of the electronic drawing and the frame information data of other electronic drawings in the first training sample based on the extracted feature data.

[0012] In one embodiment, the training of the initial model based on the first training sample to obtain the first model also includes: training the initial model to post-process the identified data; wherein the post-processing step includes one or more steps of semantic correction, format standardization, noise removal, checking the continuity of drawing numbers, and checking the progressive relationship of drawing versions.

[0013] In one of the embodiments, the training of the initial model based on the first training sample to obtain the first model also includes: training the initial model to provide correction and / or correction suggestions based on the type of error example after identifying an abnormal situation; wherein, when the error example is a regular error, training the initial model to correct the error example according to predefined correction rules; when the error example is an irregular error, training the initial model to provide one or more solution options for correcting the error example.

[0014] In one of the embodiments, the method further includes: preprocessing the group of electronic drawings before verifying the group of electronic drawings using the first model; wherein the preprocessing step includes one or more steps of format conversion of the electronic drawing files, identification of directory files, list output of the electronic drawing files, and acquisition of the structure of sub-folders.

[0015] In a second aspect, the present application provides an electronic drawing verification system, comprising: at least one processor and at least one memory; wherein at least one of the memories is used to store computer instructions, and at least one of the processors is used to execute at least part of the computer instructions to implement the steps of the method described in any of the preceding embodiments; and an interactive component, coupled to the processor, configured to display verification results and interact with a user.

[0016] The above system can ensure the data integrity and accuracy of electronic drawings during the conversion and transmission process through automated verification, reduce reliance on manual verification, and improve the verification efficiency of electronic drawings; and can significantly reduce errors caused by manual entry and enhance the reliability of the workflow; in addition, interactive components can be used to facilitate users to view verification results in real time and edit and modify them directly on the interface, thereby improving user experience and enabling non-professionals to effectively verify electronic drawings.

[0017] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements the steps of the method described in any of the foregoing embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the implementation methods of this specification or the technical solutions in the prior art, the drawings required for use in the implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some implementation methods recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0019] Figure 1 A flowchart of the steps of a verification method according to an embodiment of the present application; Figure 2 A schematic diagram of an interactive interface of a verification system according to an embodiment of the present application; Figure 3 A schematic diagram showing the contents of a catalog page drawing of an embodiment of the present application; Figure 4 A schematic diagram showing the contents of other electronic drawings according to an embodiment of the present application; Figure 5 A schematic diagram showing a first verification result of an embodiment of the present application; Figure 6 A flowchart of the steps of a verification method according to an embodiment of the present application; Figure 7 A flowchart of the steps of a verification method according to an embodiment of the present application; Figure 8 A flowchart of the steps of a first training process according to an embodiment of the present application; Fig. 9 A flowchart of the steps of a first training process according to an embodiment of the present application; Fig.10 A flowchart of the steps of a first training process according to an embodiment of the present application; Fig.11 A flowchart of the steps of a first training process according to an embodiment of the present application; Fig.12 A flowchart of the steps of a verification method according to an embodiment of the present application; Fig.13 Schematic diagram of the composition of a verification system according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific embodiments disclosed below.

[0021] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0022] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0023] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "upper", "lower", "left", "right" and similar expressions used herein are for illustrative purposes only and are not intended to be the only implementation method.

[0024] The traditional method of archiving urban construction drawings mainly has the following problems: 1. Poor data accuracy: In the traditional method, when converting or not converting the original CAD drawings, the step of manually checking the drawing catalog is still required, which will inevitably lead to problems such as data errors and omissions.

[0025] 2. Low execution efficiency and high labor costs: Traditional methods inevitably require a lot of manual input and verification, cannot perform automated operations, and have high labor costs. In some solutions, directory input can only be performed through batch files or small text processing programs, and automatic extraction and proofreading cannot be achieved, so the execution efficiency is always low.

[0026] Error detection and correction: Traditional methods require manual verification, so for such problems, the automatic processing mode of the present invention is needed to perform error detection and correction.

[0027] Based on the above problems, this application provides a verification method for electronic drawings, which uses deep learning and OCR (Optical Character Recognition) technology to perform deep verification on the converted drawing files to ensure accurate matching of various types of data and advanced error detection. The application of this system can greatly improve the work efficiency and accuracy of users in various stages such as approval, supervision, construction and archiving, and at the same time provide strong technical support for the digital management of urban construction archives, filling the technical gap in the market.

[0028] The embodiment of the present application provides a method for verifying electronic drawings, which completes the recognition of the catalog page data of the catalog page drawing through the first model obtained through training, and uses it to verify the filing status of the electronic drawing, outputs a first verification result, and then forms feedback input through the user's confirmation and editing of the first verification result to obtain a more accurate second verification result. In this way, through automated verification technology, on the one hand, the data integrity and accuracy of the electronic drawings during the conversion and transmission process can be ensured, reducing the dependence on manual verification and improving the verification efficiency of electronic drawings; on the other hand, it can significantly reduce the errors caused by manual entry and enhance the reliability of the workflow; in addition, it can also enhance interactivity, facilitate users to view the verification results in real time and edit and modify them directly on the interface, improve user experience, and enable non-professionals to effectively verify electronic drawings.

[0029] In some embodiments, Figure 2 As shown, the above verification method includes the following steps: S100, importing a group of electronic drawings; wherein the group of electronic drawings includes a catalog page drawing and at least one other electronic drawing.

[0030] For example, Figure 2As shown, the imported electronic drawings include a catalog page drawing 100 and at least one other drawing 200 (such as a floor plan drawing, a staircase drawing, a workshop drawing, etc.). It can be seen that the verification method of this embodiment can verify multiple electronic drawings in batches, greatly improving the efficiency of the drawing management process.

[0031] S300: using the trained first model to identify the catalog page data of the group of electronic drawings, verifying the filing status of the group of electronic drawings based on the catalog page data, and outputting a first verification result of the filing status of the group of electronic drawings.

[0032] For example, the catalog page data may include the name data, serial number data, date data, drawing number data, drawing name data, symbol data, designer data, supervisor data, drawing size data, etc. of each electronic drawing. Figure 2 For example, "0100&&A-01A-0000K&&Drawing Catalog" is the name data, "0020", "0101" and other data are the sequence data, "A-01A-0000K" and other data are the drawing number data, "Drawing Catalog", "First Floor Plan" and other data are the drawing name data. Figure 3 Taking the example shown in the figure, data such as "2024-04-17" in the directory file is date data. In addition, there are design unit data, plot number data, project number data, and archiving data of drawings under the drawing directory column (such as number data, name data, drawing data, design time data, etc.). After the directory page data is identified by the first model, the archiving data of other electronic drawings in the group of electronic drawings can be verified with reference to the identified directory page data to determine whether errors occurred in the previous archiving process of other electronic drawings, and the first verification result is output after the verification is completed, wherein the archiving data of other electronic drawings includes the name data of other electronic drawings (such as Figure 2 as shown) and other frame information data in electronic drawings (such as Figure 4 (data selected by the black dashed box in the figure).

[0033] S400: Output a second verification result of the filing status of the group of electronic drawings according to the user's feedback input on the first verification result.

[0034] For example, since the verification algorithm of the first model may still be imperfect, the verification result can be improved by interaction with the user. Figure 5As shown, an interactive component can be set to display the first verification result (including detailed information of errors and warnings) to allow the user to view the verification result in real time; at the same time, the user can edit or confirm the first verification result on the interface to form a second verification result with higher accuracy. This design not only improves the user experience, but also enables non-professionals to effectively use the system for electronic drawing verification. It can be understood that after the user confirms that there is no problem with the first verification result, he can click "Confirm" or "Export Verification Results (such as Excel)", at which time it can also be understood as the user's feedback input on the first verification result, and the second verification result is output accordingly (at this time, the second verification result is consistent with the content of the first verification result).

[0035] Exemplarily, the user can also make modifications directly on the interface according to the second verification result, thereby completing the error revision of the electronic drawing in a timely manner and saving modification time.

[0036] In some embodiments, Figure 6 As shown, the above verification method also includes: S500. Based on the second verification result, the first model is trained and optimized to obtain a second model, and the second model is configured to verify the next set of electronic drawings.

[0037] Exemplarily, the first model can learn according to the second verification result, adjust and optimize the data recognition verification algorithm, improve the overall accuracy and response speed of the model, make the model more intelligent, and also help to adapt to changing verification needs.

[0038] In some embodiments, Figure 7 As shown, step S300 may include: S310. Check, based on the catalog page data, whether the group of electronic drawings is missing electronic drawings recorded in the catalog page data and / or contains extra electronic drawings not recorded in the catalog page data.

[0039] Exemplarily, the sequence number data of the electronic drawings in the catalog page data may be used to check whether there are any missing or extra electronic drawings in the group of electronic drawings. For example, when the first model recognizes that there are no electronic drawings with sequence numbers recorded in the catalog page data in the group of electronic drawings, it may be considered that there are any missing corresponding electronic drawings in the group of electronic drawings; and / or, when the first model recognizes that there are any electronic drawings with sequence numbers not recorded in the catalog page data in the group of electronic drawings, it may be considered that there are any extra corresponding electronic drawings in the group of electronic drawings.

[0040] S320. Based on the catalog page data, verify whether the catalog data of other electronic drawings is consistent with the catalog page data.

[0041] Exemplarily, the description data of other electronic drawings may include symbol data, drawing number data, drawing name data, etc. in the names of other electronic drawings, as well as date data, designer data, supervisor data, drawing size data, etc. in other electronic drawings, wherein the description data of other drawings may be partially recorded in the drawing frame information area. By identifying and comparing the catalog page data with the description data of other electronic drawings, the description data of the group of electronic drawings can be effectively verified.

[0042] The existence of the electronic drawing can be verified through step S310, and the electronic drawing catalog data can be verified through step S320. The above two-fold verification can effectively ensure the integrity and accuracy of the electronic drawing data during the archiving process.

[0043] In some embodiments, Figure 8 As shown, the first model can be determined according to a first training process, and the first training process may include: SA, obtaining a first training sample; wherein the first training sample includes a plurality of verified electronic drawings and historical verification results, and the historical verification results include error examples and correct examples; SB. Train the initial model based on the first training sample to obtain a first model; wherein the initial model is an OCR model that applies a deep learning model.

[0044] Exemplarily, model training for catalog page table recognition can be performed based on the basic OCR model by applying convolutional neural network (CNN) and recurrent neural network (RNN) or Transformer model. The first training sample can include a variety of verified electronic drawings, and the initial model is trained to extract text features and graphic features from these verified electronic drawings. Extracting text features mainly includes extracting complete text lines in electronic drawings, including continuous character recognition and spatial relationship recognition, so as to correctly parse the data in the table; extracting graphic features mainly includes border detection, marking symbol recognition (such as symbols), and recognition of graphics related to text (such as table lines, dividing lines, etc.).

[0045] For example, in order to more accurately identify specific types of drawings (such as urban construction drawings), a training data set containing specific feature data can be prepared in advance, and the OCR model using deep learning technology can be trained through the training data set to enable it to more intelligently identify the catalog page data in urban construction drawings. Optionally, the specific feature data can include drawing index feature data, serial number data, symbol data, drawing number data, format feature data, spatial relationship feature data between graphics and text, etc.

[0046] Optional, such as Fig. 9 As shown, step SB may include: SB1, training an initial model to recognize the catalog page data of the electronic drawings and the frame information data of other electronic drawings in the first training sample; SB3. Train the initial model to distinguish between incorrect examples and correct examples of historical verification results in the first training sample.

[0047] Exemplarily, the text information of each electronic drawing in the first training sample may appear in different fonts, sizes, and formats, thereby ensuring that the initial model can adapt to a variety of inputs.

[0048] Exemplarily, the correct text form can be learned by training the initial model to recognize correct examples, and the error recognition capability can be improved by recognizing incorrect examples. Optionally, the incorrect examples can include incorrect spelling, mismatch between drawing number and description, incorrect drawing size data, etc.

[0049] Optionally, step SB1 may include: SB11, training the initial model to extract one or more of text feature data, border feature data, table line feature data, symbol feature data, drawing index feature data, sequence number data, drawing number data, format feature data, and spatial relationship feature data of the electronic drawing in the first training sample; SB12. Train the initial model to identify the catalog page data of the electronic drawing and the frame information data of other electronic drawings in the first training sample according to the extracted feature data.

[0050] Exemplarily, by training the initial model to extract the above-mentioned feature data and identifying the catalog page data and the frame information data of other electronic drawings based on the extracted feature data, the initial model can gradually master the ability to recognize the text information of specific drawings (such as urban construction drawings).

[0051] In some embodiments, Fig.10 As shown, step SB may also include: SB2. Train the initial model to post-process the identified data; wherein the post-processing steps include one or more steps of semantic correction, format standardization, noise removal, checking the continuity of drawing numbers, and checking the progressive relationship of drawing versions.

[0052] Exemplarily, the initial model can be trained to perform semantic correction on the recognized data, such as correcting spelling errors in OCR recognition; the initial model can be trained to standardize the format of the recognized data, such as unifying the data format (including font and size); the initial model can be trained to remove noise from the recognized data, such as removing non-data text (including footer and header information); the initial model can be trained to check the continuity of drawing numbers and the progressive relationship of drawing versions on the recognized data, for example, by introducing natural language processing (NLP) technology for training.

[0053] In some embodiments, Fig.11 As shown, step SB may also include: SB4. After identifying an abnormal situation, the training initial model provides corrections and / or correction suggestions based on the type of error example; when the error example is a regular error, the training initial model corrects the error example according to predefined correction rules; when the error example is an irregular error, the training initial model provides one or more options for correcting the error example.

[0054] For example, when some common and rule-clear errors are identified, such as incorrect date format, missing units, etc., an automatic correction process can be triggered, and the initial model can be trained to apply a predefined automatic correction algorithm for direct correction; when non-regular errors are identified, such as date errors, personnel information errors, etc., a semi-automatic correction process can be triggered. This process will show the user where the problem lies and provide several possible correction options. The user can choose the most appropriate correction plan from the options provided, and the system can learn from the user's choice to optimize future recommendation strategies.

[0055] In some embodiments, Fig.12 As shown, the above verification method also includes: S200. Before verifying the group of electronic drawings using the first model, preprocess the group of electronic drawings; wherein the preprocessing step includes one or more steps of format conversion of the electronic drawing files, identification of directory files, output of a list of electronic drawing files, and acquisition of a sub-folder structure.

[0056] For example, it is difficult for existing OCR models to directly recognize text information from CAD drawing files. Therefore, the uploaded file type can be identified first and necessary format conversion can be performed. For example, DWG and DXF drawing files can be converted into PDF files, and then directory files can be identified (such as files with "directory" in the name), other drawing files can be displayed in a list (for users to click to view and edit), and the subfolder structure can be obtained.

[0057] The present application also provides a verification system for electronic drawings, which can receive CAD files uploaded by users or directly connect to the drawing database through a user-friendly interface. Once the drawing is uploaded or selected, the system will automatically start the verification process. This process can include three main steps: file format identification and conversion, content extraction and verification. In the file format identification and conversion stage, the system can automatically determine the file type and apply the corresponding parsing rules to determine whether to convert the native CAD drawing file format to a PDF file without user intervention; in the content extraction stage, the system can use OCR technology to extract necessary information from the drawing, such as drawing size data, layer information and annotations; then, the system can compare the extracted information with the uploaded drawings, automatically identify any differences that do not meet the standards, and check the accuracy of the archived information; finally, all verification results will be automatically recorded by the system and a detailed verification report will be generated. The report lists all the problems found and their locations, and the user can directly access the corresponding drawing part from the report for review or modification. In addition, in order to further improve the efficiency of the process, the system also integrates machine learning models, which can continuously optimize parameters based on historical verification data, so that the system becomes more accurate and efficient as the use time increases. In addition, the system also provides an intuitive and highly interactive user interface that allows users to easily manage drawing files, view verification results in real time, and edit and modify them directly on the interface. The system will also adjust itself based on user feedback to continuously improve verification quality.

[0058] In some embodiments, Fig.13 As shown, the verification system 100 may include at least one processor and at least one memory; wherein, at least one of the memories is used to store computer instructions, and at least one of the processors is used to execute at least part of the computer instructions to implement the steps of the method described in any of the above embodiments; and, an interactive component, coupled to the processor, is configured to display the verification results and interact with the user.

[0059] The above system can ensure the data integrity and accuracy of electronic drawings during the conversion and transmission process through automated verification, reduce reliance on manual verification, and improve the verification efficiency of electronic drawings; and can significantly reduce errors caused by manual entry and enhance the reliability of the workflow; in addition, interactive components can be used to facilitate users to view verification results in real time and edit and modify them directly on the interface, thereby improving user experience and enabling non-professionals to effectively verify electronic drawings.

[0060] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the steps of the method described in any of the above embodiments are implemented.

[0061] It should be noted that the numbers representing quantities or properties used to describe and claim certain embodiments of the present application should be understood to be modified by the terms "roughly", "about", "approximately" or "substantially" in some cases. For example, unless otherwise specified, "roughly", "about", "approximately" or "substantially" can indicate a ±20% variation of the value described. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of the present application are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

[0062] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0063] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for verifying an electronic drawing, characterized in that: include: Importing a set of electronic drawings; wherein the set of electronic drawings includes a catalog page drawing and at least one other electronic drawing; The first model obtained through training is used to identify the catalog page data of the catalog page drawing, and the filing status of the group of electronic drawings is verified based on the catalog page data, and a first verification result of the filing status of the group of electronic drawings is output; wherein the verification of the filing status of the group of electronic drawings based on the catalog page data includes: verifying whether the group of electronic drawings lacks electronic drawings recorded by the catalog page data and / or has extra electronic drawings not recorded by the catalog page data based on the catalog page data; and verifying whether the bibliographic data of other electronic drawings is consistent with that recorded in the catalog page data based on the catalog page data; Based on the user's feedback input on the first verification result, a second verification result of the filing status of the group of electronic drawings is output.

2. The method according to claim 1, characterized in that Also includes: The first model is trained and optimized based on the second verification result to obtain a second model, and the second model is configured to verify the next set of electronic drawings.

3. The method according to claim 1 or 2, characterized in that: The first model is determined according to a first training process, and the first training process includes: Acquire a first training sample; wherein the first training sample includes a plurality of verified electronic drawings and historical verification results, and the historical verification results include error examples and correct examples; The initial model is trained based on the first training sample to obtain the first model; wherein the initial model is an OCR model that applies a deep learning model.

4. The method according to claim 3, characterized in that The step of training the initial model based on the first training sample to obtain the first model includes: Training the initial model to recognize the catalog page data of the electronic drawing and the frame information data of other electronic drawings in the first training sample; The initial model is trained to distinguish between erroneous examples and correct examples of historical verification results in the first training sample.

5. The method according to claim 4, characterized in that The training of the initial model to recognize the catalog page data of the electronic drawing and the frame information data of other electronic drawings in the first training sample includes: Training the initial model to extract one or more of text feature data, border feature data, table line feature data, symbol feature data, drawing index feature data, sequence number data, drawing number data, format feature data, and spatial relationship feature data of the graphic and text of the electronic drawing in the first training sample; The initial model is trained to identify the catalog page data of the electronic drawing and the frame information data of other electronic drawings in the first training sample according to the extracted feature data.

6. The method according to claim 4, characterized in that The training of the initial model based on the first training sample to obtain the first model further includes: The initial model is trained to perform post-processing on the identified data; wherein the post-processing steps include one or more steps of semantic correction, format standardization, noise removal, checking the continuity of drawing numbers, and checking the progressive relationship of drawing versions.

7. The method according to claim 4, characterized in that The training of the initial model based on the first training sample to obtain the first model further includes: The initial model is trained to provide corrections and / or correction suggestions based on the type of error examples after identifying abnormal situations; wherein, when the error example is a regular error, the initial model is trained to correct the error example according to predefined correction rules; when the error example is an irregular error, the initial model is trained to provide one or more solution options for correcting the error example.

8. The method according to claim 1 or 2, characterized in that: The method further comprises: Before using the first model to verify the group of electronic drawings, the group of electronic drawings is preprocessed; wherein the preprocessing step includes one or more steps of format conversion of the electronic drawing files, identification of directory files, list output of the electronic drawing files, and acquisition of the structure of subfolders.

9. An electronic drawing verification system, characterized in that: include: At least one processor and at least one memory; wherein at least one of the memories is used to store computer instructions, and at least one of the processors is used to execute at least part of the computer instructions to implement the steps of the method according to any one of claims 1 to 8; as well as, The interactive component is coupled to the processor and is configured to display the verification result and interact with the user.

10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, which implements the steps of the method according to any one of claims 1 to 8 when executed.

Images