Intelligent correction system and method for CAD graph

Abstract

The invention provides a CAD graph intelligent correction system and method, and relates to the field of correction systems. The CAD graph intelligent correction system comprises a preprocessing module which comprises intelligent denoising and enhancement, automatic labeling and segmentation and 3D drawing preprocessing, the intelligent denoising and enhancement is used for removing noise and irrelevant information in a drawing, then a color drawing is converted into a grey-scale map, binarization processing is carried out, and the grey-scale map is subjected to 3D drawing preprocessing; the automatic labeling and segmentation are used for identifying key elements in the drawing and labeling and segmenting the key elements to facilitate subsequent identification and correction, and the 3D drawing preprocessing is used for preprocessing the 3D drawing and comprises automatic generation of a profile map, a perspective map and the like. Through the arrangement of the preprocessing module, the recognition precision is higher, misjudgment and missed judgment are reduced, through the arrangement of the intelligent recognition and correction module, the system can understand terminologies and abbreviations in texts, the correction accuracy is higher, and through the arrangement of the data reconstruction and transportation module, the time and errors of manual report writing are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of correction systems, and in particular to a CAD drawing intelligent correction system and method. Background Art

[0002] Computer-aided design (CAD) refers to the use of computers and their graphics equipment to help designers perform design work. CAD drawings are drawings based on CAD. The machining center can process products based on the imported CAD drawings. The processing accuracy of its products is greatly affected by the quality of the CAD drawings. Therefore, improving the quality of CAD drawings, that is, reducing the error between the CAD drawings and the actual path dimensions of the product, has become a problem that must be faced.

[0003] Although the existing technology manually corrects errors in CAD drawings, the modification process is rather cumbersome, which increases the workload of staff. Therefore, those skilled in the art provide a CAD drawing intelligent correction system and method to solve the problems raised in the above background technology. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the deficiencies in the prior art, the present invention provides a CAD drawing intelligent correction system and method, which solves the problem that drawings need to be corrected manually and the modification process cannot be traced back.

[0006] (II) Technical solution

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a CAD drawing intelligent correction system, including: a preprocessing module, including intelligent denoising and enhancement, automatic marking and segmentation, and 3D drawing preprocessing, the intelligent denoising and enhancement is used to remove noise and irrelevant information in the drawing, and then convert the color drawing into a grayscale image and perform binarization processing, the automatic marking and segmentation is used to identify key elements in the drawing and mark and segment them to facilitate subsequent identification and correction, and the 3D drawing preprocessing is used to preprocess 3D drawings, including automatically generating cross-sectional views, perspective views, etc.;

[0008] Intelligent recognition and correction module, including geometric feature recognition, text recognition and correction, and bad pixel removal and fitting;

[0009] Data reconstruction and output module, including data reconstruction, file output and automatic report generation;

[0010] A review and compliance check module, which is used to perform intelligent review of drawings, and users can customize review rules to adapt to different specifications and standards;

[0011] The result viewing and saving module is used to view the revised drawings, perform necessary checks and confirmations, and save the revised drawing files.

[0012] Preferably, the geometric feature recognition can adapt to different specifications and standards through user-defined review rules. The text recognition and correction uses OCR technology to identify text information in drawings, and performs semantic analysis on the text in drawings to correct erroneous annotations and inconsistent text information. The bad pixel removal and fitting are used to identify and remove bad pixels that cause small grooves in curves or straight lines, and use orthogonal polynomial curve fitting or least squares method to interpolate and add points to form a complete line trace.

[0013] Preferably, the data reconstruction reconstructs the corrected CAD point information to form a high-quality CAD drawing, and processes dense points to avoid affecting the drawing quality. The file output selects the output format, the system generates the corrected drawing file, and provides a download link. The automatic report generation is used to automatically generate a correction report, which records in detail the changes before and after the drawing correction, correction suggestions and correction results, which is convenient for users to view and review.

[0014] Preferably, the preprocessing module uses Gaussian filtering to remove noise in the drawing to make the image smooth, and then converts the color drawing into a grayscale image and performs binarization processing, and the following formula is required:

[0015] Gaussian filtering:

[0016]

[0017] Where, is the standard deviation, which controls the smoothness of the filter;

[0018] Grayscale formula:

[0019] Gary(x,y)=0.299·R(x,y)+0.587·G(x,y)+0.114·B(x,y)

[0020] Among them, R, G, B are the pixel values ​​of the red, green and blue channels respectively;

[0021] Binarization formula:

[0022]

[0023] Among them, T(x,y) is an adaptive threshold, which is usually calculated based on the local area.

[0024] Preferably, when the bad pixel removal and fitting identify the small groove bad pixels appearing in the curve or straight line, it is necessary to use orthogonal polynomial curve fitting or least square method to interpolate and add points to ensure the formation of a complete trace, and the following formula will be used:

[0025] Least Squares Method:

[0026]

[0027] Among them, (x i ,y i ) are the data points, a and b are the slope and intercept of the line;

[0028] Orthogonal polynomial curve fitting is used to fit complex curves. The formula is:

[0029]

[0030] Among them, P n (x) is a polynomial, is an orthogonal polynomial basis function, c k is the coefficient.

[0031] Preferably, in addition to the common DWG, DXF, and PDF formats, the file output also supports other formats, such as SVG, OBJ, STEP, etc., to facilitate users to use it in different platforms and software, and supports custom output templates. Users can choose different output formats and templates according to their needs.

[0032] Preferably, the review and compliance check module adds AI technology to perform more advanced compliance checks and automatically detect whether the drawings comply with industry standards and specifications.

[0033] Preferably, a method of a CAD drawing intelligent correction system comprises the following steps:

[0034] S1. Upload the locally stored CAD drawing file from the local file and confirm the uploaded file path and file name to ensure that the correct file is selected. After uploading, the user can preview the drawing and use the annotation tool to mark areas or issues that require special attention:

[0035] S2. Use the system to perform advanced denoising and image enhancement to ensure the clarity and accuracy of drawings, automatically detect and repair blurred, broken or missing parts in drawings, and automatically identify key elements in drawings (such as walls, doors, windows, equipment, etc.) and mark and segment them to facilitate subsequent identification and correction:

[0036] S3. The system uses deep learning models to identify geometric features in drawings, such as straight lines, circles, arcs, etc., automatically corrects irregular lines and graphics, uses OCR technology to identify text information in drawings, and performs semantic analysis to correct incorrect annotations and inconsistent text information. At the same time, it identifies and removes bad points with small grooves in curves or straight lines, and uses orthogonal polynomial curve fitting or least squares method to interpolate and add points to form a complete trace:

[0037] S4. Reconstruct the corrected CAD point information to form a high-quality CAD drawing, and process dense points to avoid affecting the drawing quality. After processing, users can choose different output formats and templates as needed. At the same time, support for other formats such as SVG, OBJ, STEP, etc. is added to facilitate users to use in different platforms and software. After the image is saved, the system automatically generates a correction report, which records in detail the changes before and after the drawing is corrected, correction suggestions and correction results, which is convenient for users to view and review:

[0038] S5. Users can customize review rules in advance as needed to adapt to different specifications and standards. After the drawings are saved, the system automatically performs compliance checks and marks the parts that do not meet the specifications, while providing modification suggestions. Finally, users can view the revised drawings in the system interface, perform necessary checks and confirmations, and save the revised drawing files locally or in the cloud. At the same time, the system automatically saves historical records for users to view and trace.

[0039] Working principle: Upload the locally stored CAD drawing file from the local file, and confirm the uploaded file path and file name to ensure that the correct file is selected. After uploading, the user can preview the drawing and use the annotation tool to mark the areas or problems that need special attention. Use the system for advanced denoising and image enhancement to ensure the clarity and accuracy of the drawing, automatically detect and repair the blurred, broken or missing parts in the drawing, and the system automatically identifies the key elements in the drawing (such as walls, doors and windows, equipment, etc.) and annotates and segments them to facilitate subsequent identification and correction. The system uses a deep learning model to identify the geometric features in the drawing, such as straight lines, circles, arcs, etc., and automatically corrects irregular lines and graphics. It uses OCR technology to identify text information in the drawing and perform semantic analysis to correct incorrect annotations and inconsistent text information. At the same time, it identifies and removes bad points with small grooves in curves or straight lines, and uses orthogonal polynomial curve fitting or least squares method to interpolate and add points. Form a complete line trace, reconstruct the corrected CAD point information to form a high-quality CAD drawing, and process dense points to avoid affecting the quality of the drawing. After the processing is completed, the user can choose different output formats and templates as needed, and increase support for other formats, such as SVG, OBJ, STEP, etc., to facilitate users to use in different platforms and software. After the image is saved, the system automatically generates a correction report, which records in detail the changes before and after the drawing is corrected, correction suggestions and correction results, which is convenient for users to view and review. Users can customize review rules in advance as needed to adapt to different specifications and standards. After the drawing is saved, the system automatically performs a compliance check and marks the parts that do not meet the specifications, and provides modification suggestions. Finally, the user can view the corrected drawing in the system interface, perform necessary checks and confirmations, and save the corrected drawing file locally or in the cloud. At the same time, the system automatically saves historical records for users to view and trace.

[0040] (III) Beneficial effects

[0041] The present invention provides a CAD drawing intelligent correction system and method, which has the following beneficial effects:

[0042] 1. By setting up a preprocessing module and performing preliminary denoising and image enhancement on the drawings, the quality of the drawings is higher, the accuracy of recognition and correction is also higher, and the system automatically identifies and marks key elements, with higher recognition accuracy and reduced misjudgments and missed judgments.

[0043] 2. By setting up an intelligent recognition and correction module, text recognition and correction are used to support multi-language recognition and translation, which is convenient for use in multinational projects and effectively improves the scope of application. By analyzing semantics, the system can understand professional terms and abbreviations in the text, making corrections more accurate.

[0044] 3. By setting up the data reconstruction and transportation module and utilizing data reconstruction, the structure and layout of the drawings can be optimized to ensure the integrity and readability of the drawings. By utilizing file output, multiple output formats are supported to meet the needs of different users and platforms, making it more applicable. Users can choose different output formats and templates as needed to improve the flexibility of the system and user satisfaction. At the same time, the system automatically generates a correction report, which records in detail the changes before and after the drawing correction, correction suggestions and correction results, reducing the time and errors in manual report writing. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the process of the present invention;

[0046] Figure 2 It is a schematic diagram of the process of the pre-processing module of the present invention;

[0047] Figure 3 It is a flow chart of the intelligent identification and correction module of the present invention;

[0048] Figure 4 It is a schematic diagram of the process structure of the data reconstruction and output module of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Embodiment 1:

[0051] like Figure 1-4 As shown, an embodiment of the present invention provides a CAD drawing intelligent correction system, including: a preprocessing module, including intelligent denoising and enhancement, automatic labeling and segmentation, and 3D drawing preprocessing, the intelligent denoising and enhancement is used to remove noise and irrelevant information in the drawing, and then convert the color drawing into a grayscale image and perform binarization processing, the automatic labeling and segmentation is used to identify key elements in the drawing and perform labeling and segmentation to facilitate subsequent recognition and correction, the 3D drawing preprocessing is used to preprocess 3D drawings, including automatically generating cross-sectional views, perspective views, etc., by setting the preprocessing module, by performing preliminary denoising and image enhancement on the drawings, the drawing quality is higher, the accuracy of recognition and correction is also higher, and the system automatically identifies and labels key elements, the recognition accuracy is higher, and misjudgment and missed judgment are reduced.

[0052] Intelligent recognition and correction module, including geometric feature recognition, text recognition and correction, and bad pixel removal and fitting.

[0053] Data reconstruction and output module, including data reconstruction, file output and automatic report generation;

[0054] The review and compliance check module is used to perform intelligent review on drawings, and users can customize review rules to adapt to different specifications and standards.

[0055] The result viewing and saving module is used to view the revised drawings, perform necessary checks and confirmations, and save the revised drawing files.

[0056] The geometric feature recognition allows users to customize review rules to adapt to different specifications and standards. The text recognition and correction uses OCR technology to recognize text information in drawings, and performs semantic analysis on the text in drawings to correct erroneous annotations and inconsistent text information. The bad point removal and fitting are used to identify and remove bad points that cause small grooves in curves or straight lines, and use orthogonal polynomial curve fitting or least squares method to interpolate and add points to form complete line traces. By setting up an intelligent recognition and correction module, text recognition and correction are used to support multi-language recognition and translation, which is convenient for use in multinational projects and effectively improves the scope of application. By analyzing semantics, the system can understand professional terms and abbreviations in the text, and the correction accuracy is higher.

[0057] The data reconstruction reconstructs the corrected CAD point information to form a high-quality CAD drawing, and processes dense points to avoid affecting the drawing quality. The file output selects the output format, the system generates the corrected drawing file, and provides a download link. The automatic report generation is used to automatically generate a correction report, which records in detail the changes before and after the drawing is corrected, correction suggestions and correction results, which is convenient for users to view and review. By setting the data reconstruction and transportation module, the drawing structure and layout can be optimized by using data reconstruction to ensure the integrity and readability of the drawing. File output supports multiple output formats to meet the needs of different users and platforms, and has stronger applicability. Users can choose different output formats and templates according to their needs to improve the flexibility of the system and user satisfaction. At the same time, the system automatically generates a correction report, which records in detail the changes before and after the drawing is corrected, correction suggestions and correction results, reducing the time and errors of manual report writing.

[0058] The preprocessing module uses Gaussian filtering to remove noise from the drawing to make the image smooth, and then converts the color drawing into a grayscale image and performs binarization processing, which requires the following formula:

[0059] Gaussian filtering:

[0060]

[0061] Where, is the standard deviation, which controls the smoothness of the filter.

[0062] Grayscale formula:

[0063] Gary(x,y)=0.299·R(x,y)+0.587·G(x,y)+0.114·B(x,y)

[0064] Among them, R, G, B are the pixel values ​​of the red, green and blue channels respectively.

[0065] Binarization formula:

[0066]

[0067] Among them, T(x,y) is an adaptive threshold, which is usually calculated based on the local area.

[0068] When the bad point removal and fitting are used to identify the small groove bad points in the curve or straight line, it is necessary to use orthogonal polynomial curve fitting or least square method to interpolate and add points to ensure the formation of a complete trace. The following formula is used:

[0069] Least Squares Method:

[0070]

[0071] Among them, (x i ,y i ) are the data points, a and b are the slope and intercept of the line.

[0072] Orthogonal polynomial curve fitting is used to fit complex curves. The formula is:

[0073]

[0074] Among them, P n (x) is a polynomial, is an orthogonal polynomial basis function, c k is the coefficient.

[0075] In addition to the common DWG, DXF, and PDF formats, the file output also supports other formats, such as SVG, OBJ, STEP, etc., which is convenient for users to use in different platforms and software, and supports custom output templates. Users can choose different output formats and templates according to their needs. According to a CAD drawing intelligent correction system according to claim 1, it is characterized in that: the review and compliance check module adds AI technology to perform more advanced compliance checks, and automatically detects whether the drawings comply with industry standards and specifications.

[0076] A method for a CAD drawing intelligent correction system, comprising the following steps:

[0077] S1. Upload the locally stored CAD drawing file from the local file and confirm the uploaded file path and file name to ensure that the correct file is selected. After uploading, the user can preview the drawing and use the annotation tool to mark areas or issues that require special attention:

[0078] S2. Use the system to perform advanced denoising and image enhancement to ensure the clarity and accuracy of drawings, automatically detect and repair blurred, broken or missing parts in drawings, and automatically identify key elements in drawings (such as walls, doors, windows, equipment, etc.) and mark and segment them to facilitate subsequent identification and correction:

[0079] S3. The system uses deep learning models to identify geometric features in drawings, such as straight lines, circles, arcs, etc., automatically corrects irregular lines and graphics, uses OCR technology to identify text information in drawings, and performs semantic analysis to correct incorrect annotations and inconsistent text information. At the same time, it identifies and removes bad points with small grooves in curves or straight lines, and uses orthogonal polynomial curve fitting or least squares method to interpolate and add points to form a complete trace:

[0080] S4. Reconstruct the corrected CAD point information to form a high-quality CAD drawing, and process dense points to avoid affecting the drawing quality. After processing, users can choose different output formats and templates as needed. At the same time, support for other formats such as SVG, OBJ, STEP, etc. is added to facilitate users to use in different platforms and software. After the image is saved, the system automatically generates a correction report, which records in detail the changes before and after the drawing is corrected, correction suggestions and correction results, which is convenient for users to view and review:

[0081] S5. Users can customize review rules in advance as needed to adapt to different specifications and standards. After the drawing is saved, the system automatically performs compliance checks and marks the parts that do not meet the specifications, while providing modification suggestions. Finally, users can view the revised drawings in the system interface, perform necessary checks and confirmations, and save the revised drawing files locally or in the cloud. At the same time, the system automatically saves historical records for users to view and trace.

[0082] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A CAD drawing intelligent correction system, characterized by: include: Preprocessing module, including intelligent denoising and enhancement, automatic labeling and segmentation, and 3D drawing preprocessing. The intelligent denoising and enhancement is used to remove noise and irrelevant information in the drawings, and then convert the color drawings into grayscale images and perform binarization processing. The automatic labeling and segmentation is used to identify key elements in the drawings and perform labeling and segmentation to facilitate subsequent identification and correction. The 3D drawing preprocessing is used to preprocess 3D drawings, including automatically generating cross-section drawings, perspective drawings, etc.; Intelligent recognition and correction module, including geometric feature recognition, text recognition and correction, and bad pixel removal and fitting; Data reconstruction and output module, including data reconstruction, file output and automatic report generation; A review and compliance check module, which is used to perform intelligent review of drawings, and users can customize review rules to adapt to different specifications and standards; The result viewing and saving module is used to view the revised drawings, perform necessary checks and confirmations, and save the revised drawing files.

2. A CAD drawing intelligent correction system according to claim 1, characterized in that: The geometric feature recognition can adapt to different specifications and standards through user-defined review rules. The text recognition and correction uses OCR technology to identify text information in drawings, and performs semantic analysis on the text in drawings to correct erroneous annotations and inconsistent text information. The bad point removal and fitting is used to identify and remove bad points that cause small grooves in curves or straight lines, and use orthogonal polynomial curve fitting or least squares method to interpolate and add points to form a complete line trace.

3. The CAD drawing intelligent correction system according to claim 1, characterized in that: The data reconstruction reconstructs the corrected CAD point information to form a high-quality CAD drawing, and processes dense points to avoid affecting the drawing quality. The file output selects the output format, the system generates the corrected drawing file, and provides a download link. The automatic report generation is used to automatically generate a correction report, which records in detail the changes before and after the drawing correction, correction suggestions and correction results, which is convenient for users to view and review.

4. The CAD drawing intelligent correction system according to claim 1, characterized in that: The preprocessing module uses Gaussian filtering to remove noise from the drawing to make the image smooth, and then converts the color drawing into a grayscale image and performs binarization processing, which requires the following formula: Gaussian filtering: Where, is the standard deviation, which controls the smoothness of the filter; Grayscale formula: Gary(x,y)=0.299·R(x,y)+0.587·G(x,y)+0.114·B(x,y) Among them, R, G, B are the pixel values ​​of the red, green and blue channels respectively; Binarization formula: Among them, T(x,y) is an adaptive threshold, which is usually calculated based on the local area.

5. The CAD drawing intelligent correction system according to claim 1, characterized in that: When the bad point removal and fitting are used to identify the small groove bad points in the curve or straight line, it is necessary to use orthogonal polynomial curve fitting or least square method to interpolate and add points to ensure the formation of a complete trace. The following formula is used: Least Squares Method: Among them, (x i ,y i ) are the data points, a and b are the slope and intercept of the line; Orthogonal polynomial curve fitting is used to fit complex curves. The formula is: Among them, P n (x) is a polynomial, is an orthogonal polynomial basis function, c k is the coefficient.

6. The CAD drawing intelligent correction system according to claim 3, characterized in that: In addition to the common DWG, DXF, and PDF formats, the file output also supports other formats, such as SVG, OBJ, STEP, etc., which is convenient for users to use in different platforms and software, and supports custom output templates. Users can choose different output formats and templates according to their needs.

7. The CAD drawing intelligent correction system according to claim 1, characterized in that: The review and compliance check module adds AI technology to perform more advanced compliance checks and automatically detect whether the drawings comply with industry standards and specifications.

8. The method of the CAD drawing intelligent correction system according to claim 1, characterized in that: The following steps are involved: S1. Upload the locally stored CAD drawing file from the local file and confirm the uploaded file path and file name to ensure that the correct file is selected. After uploading, the user can preview the drawing and use the annotation tool to mark areas or issues that require special attention: S2. Use the system to perform advanced denoising and image enhancement to ensure the clarity and accuracy of drawings, automatically detect and repair blurred, broken or missing parts in drawings, and automatically identify key elements in drawings (such as walls, doors, windows, equipment, etc.) and mark and segment them to facilitate subsequent identification and correction: S3. The system uses deep learning models to identify geometric features in drawings, such as straight lines, circles, arcs, etc., automatically corrects irregular lines and graphics, uses OCR technology to identify text information in drawings, and performs semantic analysis to correct incorrect annotations and inconsistent text information. At the same time, it identifies and removes bad points with small grooves in curves or straight lines, and uses orthogonal polynomial curve fitting or least squares method to interpolate and add points to form a complete trace: S4. Reconstruct the corrected CAD point information to form a high-quality CAD drawing, and process dense points to avoid affecting the drawing quality. After processing, users can choose different output formats and templates as needed. At the same time, support for other formats such as SVG, OBJ, STEP, etc. is added to facilitate users to use in different platforms and software. After the image is saved, the system automatically generates a correction report, which records in detail the changes before and after the drawing is corrected, correction suggestions and correction results, which is convenient for users to view and review: S5. Users can customize review rules in advance as needed to adapt to different specifications and standards. After the drawings are saved, the system automatically performs compliance checks and marks the parts that do not meet the specifications, while providing modification suggestions. Finally, users can view the revised drawings in the system interface, perform necessary checks and confirmations, and save the revised drawing files locally or in the cloud. At the same time, the system automatically saves historical records for users to view and trace.

Images