Method and system for checking component symbol model based on table extraction OCR (Optical Character Recognition)

Abstract

The invention relates to the field of electronic design, and discloses a component symbol model checking method and system based on table extraction OCR (Optical Character Recognition). Along with the development of electronic technology, component symbol model design is crucial, but a traditional manual pin information inspection mode has the defects of high error rate, low efficiency and the like. Comprising a symbol model input and preprocessing module, a data manual pin function description table OCR extraction module, a pin data comparison module and the like. The method comprises the following steps: firstly, respectively preprocessing a symbol model and a data manual, extracting pin information, filling the pin information into a standard table, strictly comparing the two tables, verifying from multiple dimensions such as row number, pin number, name and the like, if any link is not consistent, outputting Fail, and if all links are matched, outputting Pass.

Description

Technical Field

[0001] The present invention belongs to the field of computer technology, and particularly relates to a method and system for inspecting a component symbol model based on table extraction OCR. Background Art

[0002] With the development of technology, in the process of electronic design, before manufacturers carry out conventional schematic design, PCB design, digital prototype design, mechatronic collaborative design, and simulation design, they first need to carry out component symbol model design (drawing). Therefore, in the electronic design stage, symbol model design (drawing) needs to be carried out in advance. The pin information required for symbol model design is usually obtained from the pin function description table in the component data sheet (Datasheet). As the basis of electronic design, the correctness of the symbol model determines the correctness of the electronic design result. Therefore, currently, manufacturers pay more and more attention to the inspection of the symbol models of electronic components. For manufacturers, it is a huge difficulty to prepare at least several or a dozen, or even hundreds or thousands of correct symbol models for the next step of design and development according to the products they need to design and develop. Especially with the increasing chip integration, more complex functions, and more pins, correct design symbol models and post-design inspections bring many challenges.

[0003] Currently, however, manufacturers use the method of manual inspection: manually obtain the pin information of components from the component data sheet (Datasheet), and manually obtain the pin information of the designed symbol model from the EDA modeling software. Then, manually compare the pin information of the components obtained from the component data sheet (Datasheet) with the pin information of the designed symbol model obtained from the EDA modeling software, and judge whether the symbol model design is correct by manually judging the comparison result. Therefore, the existing method of inspecting symbol models has the following defects: (1) It is easy to make mistakes when manually obtaining the pin information of components from the component data sheet (Datasheet); (2) It is easy to make mistakes when manually obtaining the pin information of the designed symbol model from the EDA modeling software; (3) It is easy to make mistakes when manually comparing the pin information of components; (4) Manual processing efficiency is low and time-consuming. Summary of the Invention

[0004] One embodiment of the present invention provides an inspection system for a component symbol model based on table extraction OCR, including: a symbol model input and preprocessing module, configured to accept and preprocess the symbol model, read the symbol model, and read out the information about the pins in the symbol model and fill it into a table A with 2 columns and N rows. One column A1 records the pin numbers, and the other column A2 records the pin names. N is equal to the number of pins AN of the symbol model. In the same row, the pin number and the pin name correspond one by one; a data sheet pin function description table OCR extraction module, configured to extract the pin information of the component from the pin function description table in the preprocessed data sheet through OCR and fill it into a table B with 2 columns and N rows. One column B1 records the pin numbers, and the other column B2 records the pin names. N is equal to the number of pins BN in the data sheet. In the same row, the pin number and the pin name correspond one by one, and table B is the comparison reference table; a pin data comparison module, configured to compare table A with table B.

[0005] In one embodiment, the data sheet input and preprocessing module is configured to accept and preprocess the data sheet, read and open the data sheet while keeping the organization form of the data sheet unchanged.

[0006] In one embodiment, the comparison method of the pin data comparison module includes: comparing the number of rows AN of table A and the number of rows BN of table B. If AN = BN, then proceed to the next comparison; otherwise, output Fail, indicating a symbol model design error.

[0007] In one embodiment, the comparison method of the pin data comparison module further includes: comparing the content of the cells in column A1. If there is no repetition in the content of any cell in column A1, then proceed to the next comparison. If there is a repetition, then output Fail, indicating a symbol model design error.

[0008] In one embodiment, the comparison method of the pin data comparison module further includes: comparing the cells in column A1 with the cells in column B1. If the content of a certain cell in column A1 is the same as the content of a certain cell in column B1, then compare the next cell in column A1 with the cells in column B1 until all the cells in column A1 and column B1 are compared. If all the cells in column A1 can find cells with the same content in column B1, then proceed to the next comparison. If one or more cells in column A1 cannot find cells with the same content in column B1, then output Fail, indicating a symbol model design error.

[0009] In one embodiment, the comparison method of the pin data comparison module further includes: comparing column A1 and column B1. If the content of a certain cell in column A1 is the same as the content of a certain cell in column B1, then compare the content of the cell in column A2 of the corresponding row of table A with the content of the cell in column B2 of the corresponding row of table B. If the content of the cell in column A2 of the corresponding row of table A is the same as the content of the cell in column B2 of the corresponding row of table B, then output Pass, indicating that the symbol model design is correct; otherwise, output Fail, indicating that the symbol model design is incorrect.

[0010] The present invention also provides an embodiment, a method for inspecting a component symbol model based on table extraction OCR, which is applied to any inspection system for a component symbol model based on table extraction OCR, and includes the following steps: accepting and preprocessing the symbol model, reading the symbol model, and reading out the information about the pins in the symbol model and filling it into a table A with 2 columns and N rows. One column A1 records the pin numbers, and the other column A2 records the pin names. N is equal to the number of pins AN of the symbol model. In the same row, the pin number and the pin name correspond one by one; extracting the pin information of the component from the pin function description table in the preprocessed data manual through OCR and filling it into a table B with 2 columns and N rows. One column B1 records the pin numbers, and the other column B2 records the pin names. N is equal to the number of pins BN in the data manual. In the same row, the pin number and the pin name correspond one by one, and table B is the comparison reference table; comparing table A with table B.

[0011] In one embodiment, the comparison method of the pin data includes: comparing the number of rows AN of table A with the number of rows BN of table B. If AN = BN, then proceed to the next comparison; otherwise, output Fail, indicating that the symbol model design is incorrect.

[0012] The present invention also includes an embodiment, an electronic device, one or more processors; a storage device, on which one or more programs are stored. When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of the above.

[0013] The present invention also includes an embodiment, a computer-readable medium, on which a computer program is stored. Wherein, when the computer program is executed by a processor, the method as described in any one of the above is implemented.

[0014] The inspection method and system for a component symbol model based on table extraction OCR provided by the above embodiments have the following beneficial effects:

[0015] The present invention completely abandons the traditional manual inspection method and adopts an automated process. In the past, during manual operation, when faced with complex data manuals and model information, staff had to spend a large amount of time comparing line by line, resulting in extremely low efficiency. Taking a complex component with hundreds of pins as an example, manual inspection might take several hours or even days, while the system of the present invention, with the help of advanced algorithms and OCR, can complete a comprehensive comparison within just a few minutes, greatly shortening the design cycle, enabling electronic design projects to progress rapidly and seize the market opportunity.

[0016] It is extremely easy to make mistakes when manually obtaining and comparing information. Whether it is the subtle parameters in the data manual or the pin markings in the model, a slight oversight will lead to the transmission of errors. The present invention, through rigorous module design and refined comparison processes, verifies the symbol model from multiple dimensions. When extracting information from the data manual, the OCR technology combined with intelligent recognition algorithms can accurately identify pin data in various formats and fill in Table B as an accurate comparison benchmark; during the comparison with Table A, it checks layer by layer, and any inconsistency can be promptly detected and reported, effectively avoiding symbol model errors caused by human errors, providing a highly reliable basic model for subsequent electronic design, and greatly reducing the design rework rate.

[0017] Traditional manual inspection methods require operators to concentrate highly for a long time. Facing a large amount of boring data, they are extremely prone to fatigue, which not only affects work efficiency but may also introduce more errors due to negligence. After the present invention is automated, operators only need to perform simple operations to import the data manual and symbol model into the system, and the subsequent complex comparison work is efficiently completed by the system. Operators can be freed from the cumbersome and repetitive labor and can devote their energy to more creative electronic design work, improving the work efficiency and innovation vitality of the entire team.

[0018] With the continuous progress of electronic technology, the types of components are increasing day by day, and the formats of data manuals are also ever-changing. The system of the present invention, based on flexible OCR and modular design, can quickly adapt to the data manual formats of different manufacturers and different types of components. Whether it is a scanned copy of a traditional paper manual or a new electronic document format, it can accurately extract the required pin information. At the same time, for symbol models of different complexities, from simple basic components to highly integrated chip models, it can effectively check according to a unified standard process, with strong versatility and scalability, providing strong support for the sustainable development of the electronic design industry. Brief Description of the Drawings

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on the structures shown in these drawings.

[0020] Figure 1 It is a block diagram of an inspection system for a component symbol model based on table extraction OCR provided by the first embodiment of the present invention;

[0021] Figure 2 It is a working step diagram of a method for inspecting a component symbol model based on table extraction OCR provided by the first embodiment of the present invention;

[0022] Figure 3 It is a working step diagram of a pin data comparison module for component symbols based on OCR screenshot extraction technology provided by the first embodiment of the present invention;

[0023] Figure 4 It is a block diagram of a system for component symbols based on OCR screenshot extraction technology provided by the first embodiment of the present invention;

[0024] Figure 5 It is a working step diagram of a method for component symbols based on OCR screenshot extraction technology provided by the first embodiment of the present invention Detailed implementation manners

[0025] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0026] It should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of the present invention, the directional indications are only used to explain the relative position relationship and movement conditions between components in a specific posture. If the specific posture changes, the directional indications will also change accordingly.

[0027] In addition, if there are descriptions such as "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, if "and / or" or "and / or" appears throughout the text, its meaning includes three parallel scenarios. Taking "A and / or B" as an example, it includes Scenario A, or Scenario B, or the scenario where both A and B are satisfied simultaneously. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions results in contradictions or is unable to be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0028] Embodiment 1

[0029] Reference Figures 1 - 3 The inspection method and system for the component symbol model based on table extraction OCR of the present invention mainly cover the following key modules:

[0030] Symbol model input and preprocessing module: Responsible for receiving and preprocessing the symbol model. Specifically, it accurately reads the symbol model, completely extracts the information about the pins therein, and fills it into a table A with 2 columns and N rows according to specific rules. One column A1 is specifically used to record the pin numbers, and the other column A2 is used to record the pin names. Here, N is equal to the number of pins AN of the symbol model, and within the same row, the pin number and the pin name strictly correspond one by one to ensure the orderly arrangement of information.

[0031] Data sheet pin function description table OCR extraction module: Using advanced OCR, it performs an efficient extraction operation on the pin function description table in the preprocessed data sheet (Datasheet). It can fill the extracted component pin information into table B in the same standard format of 2 columns and N rows. One column B1 records the pin numbers, and the other column B2 records the pin names. N is equal to the number of pins BN in the data sheet (Datasheet), and the pin number and the pin name maintain a one-to-one correspondence relationship in the same row. This table B will be used as the reference table for subsequent comparison, and its accuracy is crucial.

[0032] Pin data comparison module: Bears the important task of accurately comparing table A and table B. Its comparison process is rigorous and interlocking:

[0033] First, compare the number of rows AN in table A and the number of rows BN in table B. If AN=BN, it means that the two match in basic structure and we can proceed to the next step of comparison. If they are not equal, Fail is directly output, which clearly indicates that there is an error in the symbolic model design and controls the integrity of the model from a macro level.

[0034] Next, the contents of the cells in column A1 are compared, requiring that there should be no duplication in any cell in column A1. This step is to ensure the uniqueness of the pin number. If duplication occurs, Fail is output, and it is determined that the symbol model design is incorrect, avoiding subsequent problems caused by number confusion.

[0035] Further compare the contents of the cells in column A1 with the contents of the cells in column B1. If the contents of a cell in column A1 are the same as those of a cell in column B1, then compare the next cell in column A1 with the cells in column B1 in order, and continue in this way until all the cells in column A1 and column B1 have been compared. Only when all cells in column A1 can find cells with the same contents in column B1, will they be eligible to proceed to the next step of comparison; if one or more cells in column A1 cannot find corresponding cells with the same contents in column B1, Fail will be output immediately, accurately locating the problem of inconsistent pin numbers between the symbol model and the data sheet.

[0036] Finally, compare columns A1 and B1. When the content of a cell in column A1 is the same as the content of a cell in column B1, quickly compare the content of the cell in column A2 of the corresponding row of table A with the content of the cell in column B2 of the corresponding row of table B. If the two are exactly the same, output Pass, indicating that the symbol model is designed correctly. Otherwise, output Fail. Strictly judge the matching degree of the pin names to ensure the accuracy of the details of the symbol model.

[0037] In addition, it also includes a datasheet input and preprocessing module, whose function is to properly accept and preprocess the datasheet. During the process of reading and opening the datasheet, it strives to keep the original organizational form of the datasheet unchanged, providing a stable and reliable data source for subsequent OCR extraction.

[0038] Select a common ordinary discrete resistor component. Its data sheet (Datasheet) is a common PDF format electronic document, which contains a clear table of pin function descriptions, listing in detail information such as the numbers and names of two pins. At the same time, the resistor symbol model that has been designed and completed is stored in the EDA modeling software and waiting to be checked. Import the resistor's data sheet (Datasheet) into the data sheet (Datasheet) input and preprocessing module. This module quickly reads the document and maintains its original format layout to prepare for subsequent OCR extraction. Export the resistor symbol model from the EDA modeling software to the symbol model input and preprocessing module. This module accurately reads the model and fills the two-pin information into a 2-column and 2-row table A in sequence. Column A1 records the pin numbers "1" and "2", and column A2 correspondingly records the pin names "Positive" and "Negative". The OCR extraction module for the pin function description table in the data sheet performs OCR on the preprocessed PDF data sheet and successfully extracts the pin information, which is also filled into a 2-column and 2-row table B. Column B1 records the pin numbers "1" and "2", and column B2 records the pin names "Positive" and "Negative". At this time, table B is established as an accurate comparison benchmark. The pin data comparison module starts the comparison process. First, compare the number of rows of tables A and B, both of which are 2 rows, meeting AN = BN, and proceed to the next step; then check that the cell contents in column A1 have no duplicates and continue; then compare the cell contents in column A1 and column B1, matching them one by one; finally, compare the cell contents in column A2 and column B2 of the corresponding rows, which are also exactly the same. Finally, output Pass, indicating that the design of the resistor symbol model is correct.

[0039] In some embodiments, it involves the inspection of an integrated circuit symbol model with medium complexity. Taking a medium-scale integrated circuit with dozens of pins as an example, its data sheet (Datasheet) is a paper scan, and after digital processing, an image format file is obtained. Although there are certain problems with the clarity of the text, the information in the pin function description table is complete. The corresponding integrated circuit symbol model has been preliminarily designed and completed in professional EDA software.

[0040] Import the data sheet in image format into the data sheet (Datasheet) input and preprocessing module. The module uses an image optimization algorithm to preprocess it to improve the clarity of the text and create good conditions for OCR extraction.

[0041] Import the integrated circuit symbol model into the symbol model input and preprocessing module. This module carefully analyzes the model and accurately fills the information of dozens of pins into a 2-column and N-row (N = number of pins) table A. Column A1 records the numbers of each pin, and column A2 records the corresponding pin names.

[0042] Data extraction: The data manual pin function description table OCR extraction module uses high-precision OCR to extract the preprocessed image data manual, overcomes difficulties such as blurred text, and successfully fills the pin information into Table B, which also has 2 columns and N rows. Column B1 records the pin numbers, and column B2 records the pin names, generating a reliable comparison benchmark.

[0043] Data comparison: The pin data comparison module operates strictly according to the preset process. First, it compares the number of rows. After they match, it checks that there are no duplicates in the cells of column A1. Then, it sequentially compares the cell contents of column A1 with column B1 and column A2 with column B2. During this process, although there are individual character deviations due to image recognition problems, the system uses an intelligent error correction algorithm and combines context information for accurate judgment, and finally accurately determines whether the symbol model design is correct and outputs the corresponding result.

[0044] In some embodiments, it involves the inspection of the symbol model of a high-complexity chip. A currently advanced high-integration chip is selected. Its data manual (Datasheet) is an electronic document containing nested multi-layer tables and complex graphic annotations. It has a large number of pins and detailed function descriptions but a complex format. The corresponding chip symbol model is also carefully designed on a cutting-edge EDA platform and carries a large amount of fine pin information. The data manual input and preprocessing module uses multi-layer parsing technology to open the electronic document, sorts out the clear pin function description table structure, maintains the integrity of the document, and provides a solid foundation for subsequent operations. The symbol model input and preprocessing module obtains the chip symbol model from a high-end EDA platform and uses a high-speed data reading and parsing algorithm to quickly and orderly fill the pin information of hundreds of pins into Table A to ensure the accuracy of the information. The data manual pin function description table OCR extraction module enables super-intelligent OCR, combines with a deep learning model, deeply mines the data manual with a complex format, and successfully extracts accurate pin information and fills it into Table B, becoming a reliable comparison benchmark.

[0045] The pin data comparison module starts a rigorous comparison process. Facing a large amount of data, it efficiently compares the number of rows, checks the uniqueness of the numbers, and compares the pin information row by row and column by column. When encountering complex annotations of special function pins, the system uses the built-in professional knowledge graph for intelligent understanding and comparison to comprehensively ensure the high consistency between the chip symbol model and the data manual, and finally outputs accurate inspection results to escort the design and development of high-end chips.

[0046] From the above-mentioned multiple embodiments of different types and complexities, it can be seen that the inspection method system of the component symbol model based on table extraction OCR of the present invention can stably, efficiently, and accurately complete the symbol model inspection task whether it is facing simple discrete components or complex integrated circuits and high-end chips, providing a highly valuable technical solution for the electronic design industry.

[0047] The present invention completely abandons the traditional manual inspection method and adopts an automated process. In the past, during manual operation, when faced with complex data manuals and model information, staff needed to spend a large amount of time comparing line by line, resulting in extremely low efficiency. Taking a complex component with hundreds of pins as an example, manual inspection might take several hours or even days, while the system of the present invention, with the help of advanced algorithms and OCR, can complete a comprehensive comparison in just a few minutes, greatly shortening the design cycle, enabling electronic design projects to progress rapidly and seize the market opportunity.

[0048] It is extremely easy to make mistakes in manually obtaining and comparing information. Whether it is the subtle parameters in the data manual or the pin markings in the model, a slight oversight will lead to the transmission of errors. The present invention conducts verification of the symbol model from multiple dimensions through rigorous module design and a refined comparison process. When extracting information from the data manual, the OCR technology combined with intelligent recognition algorithms can accurately identify pin data in various formats and fill in Table B as an accurate comparison benchmark; during the comparison with Table A, it checks layer by layer, and any inconsistency can be promptly detected and reported, effectively avoiding symbol model errors caused by human errors, providing a highly reliable basic model for subsequent electronic design, and greatly reducing the design rework rate.

[0049] Traditional manual inspection methods require operators to concentrate highly for a long time. Facing a large amount of boring data, they are extremely prone to fatigue, which not only affects work efficiency but may also introduce more errors due to negligence. After the present invention is automated, operators only need to perform simple operations to import the data manual and symbol model into the system, and the subsequent complex comparison work is efficiently completed by the system. Operators can be freed from the tedious and repetitive labor and can devote their energy to more creative electronic design work, improving the work efficiency and innovation vitality of the entire team.

[0050] With the continuous progress of electronic technology, the types of components are increasing day by day, and the formats of data manuals are also ever-changing. The system of the present invention, based on flexible OCR and modular design, can quickly adapt to the data manual formats of different manufacturers and different types of components. Whether it is a scanned copy of a traditional paper manual or a new electronic document format, it can accurately extract the required pin information. At the same time, for symbol models of different complexities, from simple basic components to highly integrated chip models, it can effectively check according to a unified standard process, with strong versatility and scalability, providing strong support for the sustainable development of the electronic design industry.

[0051] Example 2

[0052] Reference Figures 4 - 5, the present invention also provides an embodiment, including a component symbol inspection system based on OCR screenshot extraction technology. The system in this embodiment mainly consists of the following key modules:

[0053] Datasheet input and preprocessing module: This module is responsible for receiving datasheets in various formats and preprocessing them. During the process of reading and opening the datasheet, advanced file parsing technology and data cleaning algorithms are used to keep the original organization form of the datasheet unchanged, ensuring the accuracy and stability of subsequent processing. For example, for a PDF-format datasheet, it can accurately identify the text content and table structure, and remove possible noise interferences, such as page watermarks and irrelevant annotations, providing a clear and reliable data source for subsequent OCR extraction work.

[0054] Symbol model input and preprocessing module: This module is used to receive symbol model files in different formats. Through a specially designed model parsing engine, it deeply reads the pin information in the symbol model and fills it into a two-column N-row table A according to strict rules. One column A1 is specifically used to record the pin numbers, and the other column A2 records the pin names. Here, N is equal to the number of pins AN in the symbol model, and within the same row, the pin number and the pin name always maintain a one-to-one correspondence. Taking the common kicad_sym format symbol model as an example, the module can accurately identify the pin definition area in the model, accurately extract the number and name of each pin, and fill them into table A in an orderly manner to ensure the integrity and accuracy of the information.

[0055] Pin configuration and function diagram OCR extraction module: This module first uses intelligent screenshot technology to accurately capture the pin configuration and function diagram in the datasheet according to preset rules and image recognition algorithms. Then, it uses advanced OCR (Optical Character Recognition) technology to process the screenshot, extracts the pin information of the component from the image, and fills this information into a two-column N-row table B. During this process, one column B1 records the pin numbers, and the other column B2 records the pin names. N is equal to the number of pins BN in the datasheet, and in the same row, the pin number and the pin name are in one-to-one correspondence. In the process of processing, this module adopts high-precision OCR algorithms and image preprocessing technology, which can effectively handle problems such as blurring and noise in the image, ensuring the accuracy of the extracted pin information. For example, for some datasheet images with poor scanning quality, through preprocessing operations such as image enhancement, denoising, and binarization, the accuracy of OCR recognition is improved, and the extracted information is accurately filled into table B to make it a reliable comparison reference table.

[0056] Pin data comparison module: This is the core judgment module of the entire system, responsible for comprehensively and meticulously comparing the A table output by the symbol model input and preprocessing module with the B table output by the OCR extraction module of the pin function description table in the data manual. The comparison process follows a rigorous process:

[0057] First, perform a row count comparison, that is, compare the number of rows AN in the A table with the number of rows BN in the B table. If AN = BN, it indicates that the two tables are consistent in the basic structure, meeting the prerequisite for further comparison, and then the next comparison step can be entered; if the two are not equal, directly output Fail, indicating that there is an error in the symbol model design, and there may be problems such as mismatched pin numbers, quickly screening out possible incorrect models from a macroscopic level.

[0058] Next, perform a self-comparison of the cell contents in column A1 to check whether there are duplicate cell contents in column A1. If the content of any cell in column A1 is not repeated, it indicates that the pin numbers are unique and conform to the normal design specifications, and the subsequent comparison can continue; on the contrary, if duplicates are found, output Fail, indicating an error in the symbol model design, and there may be serious problems such as duplicate definition of pin numbers, ensuring the uniqueness and accuracy of each pin number.

[0059] Then, perform a pin number comparison, comparing the cell contents in column A1 with the cell contents in column B1 in detail. When the content of a cell in column A1 is the same as the content of a cell in column B1, for example, the content of cell A13 is the same as the content of cell B12, the system will automatically compare the next cell in column A1 with the cells in column B1 according to the preset program, and so on, until all cells in column A1 and column B1 are compared. During this process, if all cells in column A1 can find cells with the same content in column B1, it indicates that the pin numbers are basically matched in the two tables, and the next comparison step can be entered; if one or more cells in column A1 cannot find cells with the same content in column B1, output Fail, indicating an error in the symbol model design, and there may be problems such as incorrect or missing pin numbers, ensuring that each pin number can accurately correspond in the data manual and the symbol model.

[0060] Finally, the pin names are compared. When the content of a certain cell in column A1 is the same as that in column B1, the system will immediately compare the content of the cell in column A2 of the corresponding row in Table A with the content of the cell in column B2 of the corresponding row in Table B. For example, when the content of cell A12 in column A1 is the same as the content of cell B15 in column B1, the system will automatically compare the content of cell A22 with that of cell B25, and so on until all cells in columns A2 and B2 are compared. If the content of the cell in column A2 of the corresponding row in Table A is the same as the content of the cell in column B2 of the corresponding row in Table B, "Pass" is output, indicating that the symbol model design is correct, which means that the pin numbers and names are exactly matched in the data sheet and the symbol model; otherwise, "Fail" is output, indicating that the symbol model design is incorrect, and there may be problems such as inconsistent pin name definitions, ensuring that the name of each pin can also be accurately corresponding in the two data sources.

[0061] For the pin configuration and functional diagram OCR extraction module, data sheet files in various formats, such as common PDF, JPEG, etc., are input into this module. The module first identifies and parses the format of the data sheet, and then, according to the preset image capture rules, uses intelligent image recognition technology to accurately capture the pin configuration and functional diagram therein. For example, for a data sheet in PDF format, by analyzing the document structure and page layout, the page area containing pin information is automatically located and the relevant image is accurately captured. Then, advanced OCR technology is used to process the captured image. During the OCR processing, the image is first preprocessed, including operations such as grayscale conversion, noise reduction, and binarization, to improve the image quality and text clarity. Then, a well-trained OCR recognition model is used to accurately identify the text information in the image, and the information is extracted and classified according to the characteristics of the pin numbers and names. Finally, the extracted information is filled into Table B with 2 columns and N rows, where N is equal to the number of pins BN in the data sheet, ensuring that the pin numbers and names in the same row correspond one by one, and Table B is established as the reference table for subsequent comparison.

[0062] For the symbol model input and preprocessing module, symbol model files in different formats, such as kicad_sym, Altium Designer, etc., are input into this module. The module uses a dedicated model parsing algorithm to deeply analyze the structure and data storage method of the symbol model, and accurately reads the pin numbers and pin name information therein. During the reading process, it strictly identifies according to the pin definition and numbering rules to avoid incorrect reading. For example, for a complex multi-layer symbol model structure, through recursive parsing, it ensures that all pin information can be completely extracted. Then, the read information is output to a table A with 2 columns and N rows in a specified format, where one column A1 records the pin numbers and the other column A2 records the pin names. The value of N is equal to the number of pins AN of the symbol model, and the pin numbers and pin names in the same row maintain a one-to-one correspondence relationship.

[0063] The pin data comparison module receives table A and table B from the above two modules and operates according to a strict comparison process. First, compare the number of rows AN of table A and the number of rows BN of table B. If AN = BN, then continue with the next comparison; otherwise, immediately output Fail, indicating a symbol model design error. At this time, the system will record the error information and prompt the user that there may be a problem with the pin numbers not matching. Next, compare the content of the cells in column A1 to check for duplicate cell content. If the content of any cell in column A1 has no duplicates, then proceed to the next comparison; if duplicates are found, output Fail, indicating a symbol model design error, and the system will specifically point out the position of the duplicate cells to facilitate the user to find the root cause of the problem. Then, compare the content of the cells in column A1 with the content of the cells in column B1. When the content of a cell in column A1 is the same as the content of a cell in column B1, the system will automatically compare the next cell in column A1 with the cells in column B1 in sequence until all cells in column A1 and column B1 are compared. If all cells in column A1 can find cells with the same content in column B1, then continue with the next comparison; if one or more cells in column A1 cannot find cells with the same content in column B1, output Fail, indicating a symbol model design error, and the system will list the information of the unmatched cells in detail to help the user locate the problem. Finally, perform the pin name comparison. When the content of a cell in column A1 and column B1 is the same, the system will compare the content of the cell in column A2 of the corresponding row in table A and the content of the cell in column B2 of the corresponding row in table B.

[0064] With the continuous progress of electronic technology, the types of components are increasing day by day, and the data sheet formats are also ever-changing. The system of the present invention is based on flexible OCR technology and modular design, and can quickly adapt to the data sheet formats of different manufacturers and different types of components. Whether it is a scanned copy of a traditional paper manual or an emerging electronic document format, it can accurately extract the required pin information. At the same time, for symbol models of different complexities, from simple basic components to highly integrated chip models, effective inspections can be carried out according to a unified standard process, with strong versatility and scalability, providing strong support for the continuous development of the electronic design industry.

[0065] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structural transformation made under the inventive concept of the present invention by using the content of the specification and drawings of the present invention, or direct / indirect application in other related technical fields, is included in the patent protection scope of the present invention.

Claims

1. A component symbol model inspection system based on table extraction OCR, characterized in that: include: The symbol model input and preprocessing module is used to receive and preprocess the symbol model, read the symbol model, and read out the information about the pins in the symbol model, and fill it into a table A with 2 columns and N rows, wherein one column A1 records the pin number, and the other column A2 records the pin name, N is equal to the number of pins AN of the symbol model, and the pin number and the pin name in the same row correspond to each other one by one; The data sheet pin function description table OCR extraction module is used to extract the pin information of components from the pin function description table in the pre-processed data sheet through OCR, and fill it into a 2-column N-row table B, where one column B1 records the pin number and the other column B2 records the pin name, N is equal to the number of pins BN in the data sheet, the pin number and the pin name in the same row correspond one to one, and table B is the comparison reference table; Pin data comparison module, used to compare Table A with Table B.

2. The component symbol model inspection system based on table extraction OCR according to claim 1 is characterized in that: The data manual input and preprocessing module is used to receive and preprocess the data manual, read and open the data manual and keep the organizational form of the data manual unchanged.

3. The component symbol model inspection system based on table extraction OCR according to claim 1 is characterized in that: The comparison method of the pin data comparison module includes: Compare the number of rows AN in table A with the number of rows BN in table B. If AN=BN, proceed to the next step of comparison; otherwise, output Fail, indicating that the symbol model design is wrong.

4. The inspection system for component symbol models based on table extraction OCR according to claim 3 is characterized in that: The comparison method of the pin data comparison module also includes: comparing the contents of the cells in the A1 column, and if there is no duplication in any cell content in the A1 column, the next step of comparison is performed, and if there is duplication, Fail is output, indicating that the symbol model design is wrong.

5. The component symbol model inspection system based on table extraction OCR according to claim 4 is characterized in that: The comparison method of the pin data comparison module further includes: comparing the contents of cells in column A1 and cells in column B1; if the contents of a cell in column A1 are the same as the contents of a cell in column B1, then comparing the next cell in column A1 with the cells in column B1, until all cells in column A1 and column B1 are compared; if cells in column A1 can find cells with the same contents in column B1, then proceed to the next step of comparison; if one or more cells in column A1 cannot find cells with the same contents in column B1, then output Fail, indicating that the symbol model design is wrong.

6. The component symbol model inspection system based on table extraction OCR according to claim 5 is characterized in that: The comparison method of the pin data comparison module also includes: comparing the A1 column and the B1 column, if the content of a cell in the A1 column is the same as the content of a cell in the B1 column, then comparing the content of the cell in the A2 column of the corresponding row of the A table and the cell in the B2 column of the corresponding row of the B table, if the content of the cell in the A2 column of the corresponding row of the A table and the cell in the B2 column of the corresponding row of the B table are the same, then outputting Pass, indicating that the symbol model design is correct, otherwise, outputting Fail, indicating that the symbol model design is wrong.

7. A method for inspecting component symbol models based on table extraction OCR, applied to any inspection system for component symbol models based on table extraction OCR, characterized in that: The following steps are included: Accept and preprocess the symbol model, read the symbol model, and read out the information about the pins in the symbol model, and fill it into a table A with 2 columns and N rows, where one column A1 records the pin number, and the other column A2 records the pin name, N is equal to the number of pins AN of the symbol model, and the pin number and the pin name in the same row correspond to each other one by one; Extract the pin information of components from the pin function description table in the pre-processed data sheet through OCR, and fill it into a 2-column N-row table B, where one column B1 records the pin number and the other column B2 records the pin name. N is equal to the number of pins BN in the data sheet. The pin number and the pin name in the same row correspond one to one, and table B is the comparison reference table. Compare Table A with Table B.

8. The component symbol model inspection method based on table extraction OCR according to claim 7 is characterized in that: The pin data comparison methods include: Compare the number of rows AN in table A with the number of rows BN in table B. If AN=BN, proceed to the next step of comparison; otherwise, output Fail, indicating that the symbol model design is wrong.

9. An electronic device, characterized in that: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the component symbol model inspection method based on table extraction OCR as described in any one of claims 7-8.

10. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method for checking component symbol models based on table extraction and OCR as described in any one of claims 7 to 8 is implemented.

Images