System and method for assisting in identifying steel bar raw material quality assurance data

Abstract

The invention relates to the field of buildings, in particular to a system and method for assisting in recognizing quality assurance data of reinforcing steel bar raw materials, and the system comprises a database module, a picture processing module and a counterfeit checking rule module, the database module is used for storing and managing project information, personnel role information, a reinforcing steel bar authenticity verification ledger, an AI authenticity verification recognition model, algorithm rules and manufacturer information; the picture processing module is used for shooting and preprocessing a picture of a reinforcing steel bar raw material quality guarantee, and extracting information in the picture by applying an icon algorithm and a table recognition algorithm; and the authenticity verification rule module performs authenticity verification judgment on the extracted information according to a preset authenticity verification rule. The method can automatically and quickly recognize the character, icon and table information in the steel bar raw material quality assurance data, improves the recognition efficiency, and can achieve the precise recognition of the steel bar raw material quality assurance data through the training of a large number of samples through the deep learning algorithm, and reduces the error rate of manual auditing.

Description

Technical Field

[0001] The invention relates to the technical field of construction, and in particular to a system and method for assisting in identifying quality assurance data of steel bar raw materials. Background Art

[0002] In the construction industry, steel bars are important building materials, and their quality is directly related to the safety and stability of buildings. Therefore, it is crucial to ensure the quality of steel bar raw materials. The traditional way of reviewing the quality assurance data of steel bar raw materials mainly relies on manual inspection, which is not only time-consuming and labor-intensive, but also prone to errors. With the development of information technology, especially the continuous advancement of image processing and artificial intelligence technology, solutions have been provided for the automatic identification and verification of steel bar raw material quality assurance data.

[0003] However, the existing automatic identification system still has some shortcomings when processing the quality assurance information of steel bar raw materials. The quality assurance certificate of steel bar raw materials usually contains a large amount of text, icons and tabular information. The format of this information is diverse and the content is complex. It is difficult to accurately extract it through a single recognition algorithm. In addition, the formats and rules of the quality assurance certificates of steel bar raw materials produced by different manufacturers are also different, which brings great difficulties to the versatility and adaptability of the automatic identification system. Summary of the invention

[0004] Therefore, the technical problem to be solved by the present invention is to overcome the defects in the prior art, thereby providing a system and method for assisting in identifying the quality assurance data of steel bar raw materials, which combines multiple modules such as database management, image processing, artificial intelligence and verification rules, and can achieve comprehensive, accurate and efficient identification and verification of the quality assurance data of steel bar raw materials.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A system for assisting in identifying quality assurance data of steel bar raw materials, comprising a database module, an image processing module and a counterfeit detection rule module:

[0007] The database module is used to store and manage project information, personnel role information, steel bar verification records, AI verification and recognition models, algorithm rules and manufacturer information;

[0008] The image processing module is used to take and pre-process the image of the steel bar raw material quality guarantee, and extract information from the image using an icon algorithm and a table recognition algorithm;

[0009] The authentication rule module performs authentication judgment on the extracted information according to preset authentication rules.

[0010] Furthermore, the database module includes a project table, a personnel table, a steel bar ledger table, an AI model table, an algorithm rule table and a manufacturer table:

[0011] The project table is used to store project information;

[0012] The personnel table is used to store personnel and their role information;

[0013] The steel bar ledger table is used to record the detailed ledger of steel bar verification;

[0014] The AI ​​model table stores parameters and data of the AI ​​false recognition model;

[0015] The algorithm rule table stores algorithm rules for counterfeit verification;

[0016] The manufacturer table stores basic information and specific rules of the manufacturer.

[0017] Furthermore, the picture processing module includes a picture shooting processing unit, a picture preprocessing unit, an icon algorithm recognition unit and a picture table recognition unit:

[0018] The picture shooting and processing unit is used to assist in shooting pictures of the steel bar raw material quality guarantee certificate;

[0019] The image preprocessing unit performs preprocessing such as resizing, padding, cropping, and normalization on the captured image;

[0020] The icon algorithm recognition unit uses the YOLOv8 detection algorithm to recognize icons in the picture;

[0021] The picture table recognition unit uses the SLENET table recognition model to recognize the table data in the picture.

[0022] Furthermore, the icon algorithm recognition unit further includes:

[0023] The icon recognition result judgment subunit is used to judge whether the recognized icon is missing or has a large offset.

[0024] Furthermore, the picture table recognition unit also includes a table data extraction subunit and a table data preprocessing subunit:

[0025] The table data extraction subunit is used to extract the identified table data;

[0026] The table data preprocessing subunit is used to process the situation where the entire row of data is in one cell due to the absence of a table vertical line.

[0027] Furthermore, the verification rule module includes an icon verification unit, a basic information verification unit, a steel bar data verification unit and a QR code scanning data comparison verification unit:

[0028] The icon verification unit is used to determine the authenticity of the steel bar raw material quality guarantee certificate based on the icon recognition result;

[0029] The basic information verification unit is used to verify the integrity and consistency of the basic information;

[0030] The steel bar data verification unit is used to verify the steel bar data according to the rules;

[0031] The two-dimensional code scanning data comparison and verification unit is used to scan the two-dimensional code and compare the scanning result with the data in the steel bar raw material quality guarantee certificate.

[0032] A method for assisting in identifying quality assurance information of steel bar raw materials comprises the following steps:

[0033] Use the database module to store and manage project information, personnel role information, steel bar verification records, AI verification and recognition models, algorithm rules and manufacturer information;

[0034] Use the image processing module to take and pre-process the image of the steel bar raw material quality guarantee, and apply the icon algorithm and table recognition algorithm to extract the information in the image;

[0035] The authentication rule module is used to perform authentication judgment on the extracted information according to the preset authentication rules.

[0036] Furthermore, the pre-processing step includes:

[0037] Adjust image size, padding, cropping and normalization;

[0038] Use the YOLOv8 detection algorithm to identify icons in images;

[0039] Use the SLENET table recognition model to recognize tabular data in images.

[0040] Furthermore, the step of verifying the authenticity includes:

[0041] Determine whether the icon is missing or has a large offset;

[0042] Verify the integrity and consistency of basic information;

[0043] Verify the authenticity of the steel bar data according to the rules;

[0044] Scan the QR code and compare the scan result with the data on the steel bar raw material quality guarantee certificate.

[0045] Furthermore, it also includes:

[0046] Optimize performance and provide security protection for database modules;

[0047] Regularly update and optimize the image processing module and counterfeit verification rule module.

[0048] By means of the above technical solution, the present invention provides a system and method for assisting in identifying quality assurance data of steel bar raw materials, which has at least the following beneficial effects:

[0049] 1. The present invention can automatically and quickly identify the text, icons and table information in the quality assurance data of steel bar raw materials, thereby improving the recognition efficiency. At the same time, by using a deep learning algorithm to train a large number of samples, it can realize the accurate recognition of the quality assurance data of steel bar raw materials and reduce the error rate of manual review.

[0050] 2. The present invention fully considers the differences in formats and rules of quality assurance certificates for steel bar raw materials produced by different manufacturers. By constructing a flexible recognition model and a counterfeit verification rule library, it achieves comprehensive support for different formats and quality assurance materials, improves the versatility and adaptability of the system, and enables the system to be applied in a wider range of scenarios.

[0051] 3. The present invention greatly reduces the burden of manual review and reduces the manpower cost of enterprises through automated identification and verification. At the same time, the efficient operation of the system also shortens the review cycle of steel bar raw materials and improves the overall progress of the construction project.

[0052] 4. The present invention effectively prevents the influx of counterfeit and shoddy products by accurately identifying the quality assurance information of steel bar raw materials, provides a strong guarantee for the physical quality of construction projects, improves the overall safety of the construction industry, and reduces quality and safety accidents caused by steel bar quality problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 A block diagram of a system for assisting in identifying quality assurance data of steel bar raw materials;

[0055] Figure 2 A block diagram of a database module in a system for assisting in identifying quality assurance data of steel bar raw materials;

[0056] Figure 3 A block diagram of an image processing module in a system for assisting in identifying quality assurance data of steel bar raw materials;

[0057] Figure 4A block diagram of a counterfeit verification rule module in a system for assisting in identifying quality assurance data of steel bar raw materials;

[0058] Figure 5 A block diagram of a method for assisting in identifying quality assurance data of steel bar raw materials. DETAILED DESCRIPTION

[0059] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0060] The word "exemplary" is used here to mean "used as an example, embodiment or illustrative". Any embodiment described here as "exemplary" is not necessarily to be interpreted as superior or better than other embodiments. In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present disclosure.

[0061] Figure 1 A block diagram of a system for assisting in identifying quality assurance information of steel bar raw materials according to an embodiment of the present invention is shown. Figure 1 As shown, the system for assisting in identifying quality assurance data of steel bar raw materials includes a database module, an image processing module and a counterfeit detection rule module.

[0062] Database module, used to store and manage project information, personnel role information, steel bar verification records, AI verification and recognition models, algorithm rules and manufacturer information;

[0063] Specifically, project information covers basic information such as project name, project location, and construction unit, which facilitates the system to track and manage projects. Personnel role information records in detail the identity information, responsibilities and authorities of project participants to ensure the security of information and compliance of operations. The steel bar counterfeit verification ledger records the information of all quality assurance certificates of steel bar raw materials identified by the system, including identification results, verification status, etc. AI verification and recognition models and algorithm rules are the core of the system's intelligent recognition, and through continuous learning and optimization, the accuracy and efficiency of recognition are improved. In addition, the system also includes information on various manufacturers, including manufacturer name, product specifications, production batches, etc., to provide important reference for verification and judgment. The database module ensures the integrity, security and traceability of data through efficient data management and storage technology.

[0064] The image processing module is used to take and pre-process images of the steel bar raw material quality assurance certificate, and apply icon algorithm and table recognition algorithm to extract information from the image.

[0065] The image processing module has the functions of shooting, preprocessing images, and applying icon algorithms and table recognition algorithms to extract image information. In the shooting stage, high-definition shooting is supported to ensure that the image quality meets the recognition requirements. The preprocessing stage includes operations such as denoising, enhancing contrast, and correcting deformation of the image to improve recognition accuracy.

[0066] In the information extraction stage, the icon algorithm is first used to identify the icon information in the image, such as the manufacturer's logo, quality certification mark, etc. This information is of great significance for the preliminary judgment of the authenticity of the steel bar raw materials. Subsequently, the table recognition algorithm is used to extract the table information in the image, including key data such as the specifications, batches, and production dates of the steel bars. The image processing module uses advanced image processing technology and algorithms to achieve rapid and accurate extraction of the quality guarantee information of the steel bar raw materials, providing strong support for subsequent counterfeit verification.

[0067] The authentication rule module performs authentication judgment on the extracted information according to the preset authentication rules.

[0068] The verification rule module compares and verifies the extracted information one by one according to the preset verification rules. The verification rules cover multiple aspects, such as the authenticity of manufacturer information, compliance with product specifications, and rationality of production date.

[0069] First, the extracted manufacturer information is matched with the manufacturer information in the database to verify the authenticity of the manufacturer information in the picture. Then, the specifications, batches and other information of the steel bars are verified according to the algorithm rules to ensure that the products meet the relevant standards and regulations. At the same time, the rationality of the production date will be checked to prevent products with abnormal dates or forged products from entering the project. During the verification process, the AI ​​verification recognition model will be used to conduct in-depth analysis of the text, icons, etc. in the picture to further improve the accuracy of the verification. The verification rules module uses strict rules and algorithms to achieve comprehensive and accurate verification of the quality guarantee information of steel raw materials.

[0070] In this embodiment, the text, icons and table information in the quality assurance data of steel bar raw materials can be automatically and quickly identified, thereby improving the recognition efficiency. At the same time, by training a large number of samples using a deep learning algorithm, accurate recognition of the quality assurance data of steel bar raw materials can be achieved, thereby reducing the error rate of manual review.

[0071] In one possible implementation, Figure 2As shown in the figure, the database module includes: project table, which is used to store project information, including but not limited to project name, project location, construction unit, design unit, supervision unit, project start and end date, project scale, budget, etc. The design of the project table fully considers the complexity and diversity of the project to ensure that all key information of the project can be recorded comprehensively and accurately. Through the project table, it is easy to track and manage the project. At the same time, the project table also supports associated queries with other tables (such as personnel table, steel bar ledger table, etc.), which is convenient for understanding the overall situation of the project from multiple angles.

[0072] The personnel table is used to store personnel and role information. It records the basic information of personnel, such as name, gender, age, contact information, position, department, role authority, etc. The personnel table can be used to manage project participants comprehensively, including the entry, modification, query and deletion of personnel information. At the same time, the personnel table also supports associated query with the project table, which makes it easier for users to understand the personnel configuration and division of responsibilities for each project. In addition, the personnel table also implements the authority management of project participants, ensuring that only personnel with corresponding authority can access and operate relevant data.

[0073] The steel bar ledger table is used to record the detailed ledger of steel bar verification; it records in detail the identification results, verification status, manufacturer, specification model, batch number, production date, quantity and other information of each batch of steel bars. Through the steel bar ledger table, it is possible to fully track and manage the steel bar raw materials, providing a basis for project quality traceability. At the same time, the steel bar ledger table also supports associated queries with project tables, personnel tables and other tables, which is convenient for users to understand the use of steel bars and verification results from multiple angles. In addition, the steel bar ledger table also realizes the inventory management and early warning functions of steel bar raw materials to ensure the timely supply and reasonable use of steel bars required for the project.

[0074] The AI ​​model table stores the parameters and data of the AI ​​false recognition model; the AI ​​model table records key information such as the name, version, training data, model structure, and parameter settings of the AI ​​model. Through the AI ​​model table, comprehensive management and optimization of the AI ​​model can be achieved. The system supports operations such as training, verification, testing, and deployment of AI models to ensure that the AI ​​model can accurately and efficiently identify the information in the quality guarantee certificate of steel bar raw materials. At the same time, the AI ​​model table also supports the monitoring and evaluation of model performance, so that users can understand the recognition effect and stability of the model. In addition, the backup and recovery functions of model data are also implemented through the AI ​​model table to ensure the security and reliability of model data.

[0075] The algorithm rule table stores the algorithm rules used for counterfeit verification; the algorithm rule table records key information such as the name, description, parameter settings, and scope of application of the algorithm rules. Through the algorithm rule table, the system can achieve comprehensive management and optimization of the algorithm rules. The system supports operations such as adding, modifying, deleting, and querying algorithm rules to ensure that the algorithm rules can be accurately and efficiently applied to the counterfeit verification judgment of the steel bar raw material quality guarantee. At the same time, the algorithm rule table also supports the testing and verification functions of the algorithm rules, which is convenient for users to understand the accuracy and stability of the algorithm rules.

[0076] The manufacturer table stores the basic information and specific rules of the manufacturer. The manufacturer table records the manufacturer's name, address, contact information, production qualifications, product specifications, batch number rules and other key information. Through the manufacturer table, the system can achieve comprehensive management and verification of manufacturers. The system supports operations such as adding, modifying, deleting and querying manufacturers to ensure the accuracy and completeness of manufacturer information. At the same time, the manufacturer table also supports the storage and management of manufacturer-specific rules, such as product identification rules and production date coding rules for specific manufacturers.

[0077] In one possible implementation, Figure 3 As shown, the picture processing module includes: a picture shooting processing unit, which is used to assist in shooting pictures of the steel bar raw material quality guarantee; the picture shooting processing unit is responsible for assisting users in shooting pictures of the steel bar raw material quality guarantee. The picture shooting processing unit integrates image capture technology and optimization algorithms to ensure that the captured pictures are clear and accurate. In actual operation, the user only needs to place the steel bar raw material quality guarantee in the shooting area, and the system can automatically adjust parameters such as focus and exposure to capture the best image. In addition, the picture shooting processing unit also supports multiple shooting modes, such as single shooting, continuous shooting, etc., to meet the shooting needs in different scenes. In order to ensure the quality of the captured pictures, the unit also has an image quality assessment function, which can detect and feedback key indicators such as picture clarity and brightness in real time, helping users to adjust shooting conditions in time.

[0078] The image preprocessing unit performs preprocessing such as resizing, padding, cropping, and normalization on the captured images; the image preprocessing unit is responsible for performing a series of preprocessing operations on the captured images to improve the accuracy and efficiency of the subsequent recognition algorithm. These preprocessing operations include but are not limited to resizing, padding, cropping, and normalization. The image resizing is to ensure that the image meets the input requirements of the subsequent recognition algorithm; the padding operation is to eliminate blank areas in the image and improve recognition accuracy; the cropping operation is to remove irrelevant information in the image and reduce the amount of calculation; the normalization operation is to convert the image data into a unified format and range for subsequent processing. Through these preprocessing operations, the image preprocessing unit can significantly improve the quality and recognition effect of the image, and provide data support for the subsequent icon algorithm recognition unit and the image table recognition unit.

[0079] The icon algorithm recognition unit uses the YOLOv8 detection algorithm to identify icons in pictures; it uses the advanced YOLOv8 detection algorithm to identify icons in pictures. YOLOv8 is a target detection algorithm based on deep learning, which has the advantages of fast detection speed and high accuracy. In the icon algorithm recognition unit, the system first uses the YOLOv8 algorithm to perform preliminary detection on the picture and extract possible icon areas. Then, the system further analyzes and processes these areas to identify specific icon types and information. Through the icon algorithm recognition unit, the system can accurately identify key icons such as the manufacturer's logo and quality certification mark in the quality assurance book of steel bar raw materials, providing strong support for subsequent information extraction and counterfeit verification.

[0080] The picture table recognition unit uses the SLENET table recognition model to recognize the table data in the picture. The picture table recognition unit uses the SLENET table recognition model to recognize the table data in the picture. SLENET is a table recognition model based on deep learning, which has the advantages of high recognition accuracy and strong adaptability. In the picture table recognition unit, the system first uses the SLENET model to locate and segment the table area in the picture. Then, the system recognizes each cell in the table one by one and extracts the text and digital information. Through the picture table recognition unit, the system can accurately identify key data such as specifications, batch numbers, and production dates in the quality assurance book of steel bar raw materials, providing an important basis for subsequent information integration and counterfeit verification. At the same time, the unit also supports the formatting and sorting of table data, which is convenient for users to conduct subsequent data analysis and processing.

[0081] In one possible implementation, Figure 3As shown, the icon algorithm recognition unit also includes: an icon recognition result judgment subunit, which is used to judge whether the recognized icon is missing or has a large offset. The icon algorithm recognition unit is responsible for further verification and judgment of the recognized icon to ensure the accuracy and integrity of the icon. The functions of this subunit mainly revolve around two aspects: one is to judge whether the recognized icon is missing, and the other is to judge whether the recognized icon has a large offset.

[0082] First, for the judgment of missing icons, the icon recognition result judgment subunit will compare the actual recognized icons with the expected icons based on the preset icon library and the results of the recognition algorithm. If an expected icon does not appear in the recognition result, or its confidence is lower than the set threshold, the subunit will determine that the icon is missing. This judgment is crucial to ensure the integrity and authenticity of the quality guarantee of steel bar raw materials, because the absence of certain key icons (such as manufacturer logos, quality certification marks, etc.) may mean that the guarantee is at risk of forgery or tampering.

[0083] Secondly, for the judgment of icon offset, the icon recognition result judgment subunit will use the positioning technology in image processing to perform precise position analysis on the identified icons. By calculating the deviation between the actual position of the icon and the theoretical position (i.e., the standard position in the preset icon library), the subunit can judge the degree of icon offset. If the offset exceeds the set threshold, the subunit will determine that the icon has a large offset. This judgment is of certain significance for improving the accuracy and stability of icon recognition, because the offset of the icon will be caused by factors such as shooting angle and lighting conditions, and these factors will affect the subsequent information extraction and verification judgment.

[0084] In order to ensure the effectiveness and accuracy of the icon recognition result judgment subunit, the system also uses a variety of technical means for optimization. For example, the deep learning algorithm is used to extract and classify icons more finely to improve the accuracy and robustness of recognition; image enhancement technology is introduced to pre-process the captured images to reduce the impact of factors such as light and noise on icon recognition; and reasonable threshold settings and verification rules are designed to ensure the reliability and consistency of the judgment results. Through the optimization of these technical means, the icon recognition result judgment subunit can more accurately judge the accuracy and completeness of the recognized icons.

[0085] In one possible implementation, Figure 3As shown, the image table recognition unit also includes: a table data extraction subunit for extracting the recognized table data; the table data extraction subunit uses the advanced SLENET table recognition model to accurately identify and extract the table data in the image. The table data extraction subunit first accurately locates the table area in the image through a deep learning algorithm, and then parses each cell in the table one by one. Whether it is text or digital information, it can be accurately captured and extracted. In this process, the subunit is also highly robust and can effectively deal with interference factors such as blur, distortion, and noise in the image to ensure the integrity and accuracy of the data. The extracted data will be structured and stored according to its original layout in the table, which provides great convenience for subsequent data analysis and processing.

[0086] The table data preprocessing subunit is used to handle the situation where the entire row of data is in one cell due to the absence of vertical lines in the table. In practical applications, due to the influence of various factors such as shooting conditions and document formats, the vertical lines in the table are sometimes missing, resulting in the entire row of data being mistakenly identified as the content of a cell. To address this problem, the image table recognition unit is specially equipped with a table data preprocessing subunit to handle such special cases. Through image processing technology and algorithms, the table data preprocessing subunit can intelligently identify and segment cells that were originally merged incorrectly, and restore the data therein to the correct row and column structure. In this process, the table data preprocessing subunit will comprehensively consider multiple dimensions such as the text content, font size, spacing, etc. of the cell to ensure the accuracy and rationality of the segmentation. At the same time, for data fragmentation or missing data caused by segmentation, the table data preprocessing subunit can also repair and supplement it through intelligent interpolation, data matching and other technologies to ensure that the final output table data is both accurate and complete.

[0087] In one possible implementation, Figure 4 As shown, the verification rule module includes: an icon verification unit, which is used to judge the authenticity of the steel bar raw material quality guarantee based on the icon recognition result; the icon verification unit is based on the output result of the icon algorithm recognition unit, and judges the authenticity of the key icons (such as manufacturer logo, quality certification mark, etc.) in the steel bar raw material quality guarantee. The icon verification unit compares the recognized icon with the standard icon in the preset icon database, and comprehensively evaluates the authenticity of the guarantee by combining the icon's shape, color, size and other features, as well as the icon's arrangement order and position relationship. Any mismatch or abnormality of the icon may be regarded as a potential sign of counterfeiting, triggering further verification process.

[0088] The basic information verification unit is used to verify the integrity and consistency of the basic information; the basic information verification unit is used to verify the integrity and consistency of the basic information in the steel bar raw material quality guarantee. This information includes but is not limited to the manufacturer's name, address, contact information, and the model, specification, batch number, etc. of the steel bar. The basic information verification unit compares the identified information with the known correct information in the database to check whether there are omissions, errors or inconsistencies, thereby judging the authenticity of the guarantee.

[0089] The steel bar data verification unit is used to verify the authenticity of the steel bar data according to the rules of the steel bar data; the steel bar data verification unit uses the steel bar data rules to verify the authenticity of the steel bar data. These rules involve indicators such as the chemical composition, mechanical properties, and dimensional tolerance of the steel bar. The steel bar data verification unit compares the identified steel bar data with the standard data to check whether it meets the specified range and standards, so as to evaluate the authenticity of the steel bar data in the guarantee.

[0090] The QR code scanning data comparison and verification unit is used to scan the QR code and compare the scanning result with the data on the steel bar raw material quality guarantee. The QR code scanning data comparison and verification unit obtains the electronic data contained in the QR code on the steel bar raw material quality guarantee by scanning the QR code, and compares it with the data on the paper guarantee. This step not only verifies the authenticity of the QR code, but also ensures the consistency of the paper and electronic data, providing double protection for the authenticity of the steel bar raw material quality guarantee.

[0091] In one possible implementation, Figure 5 As shown, a method for assisting in identifying quality assurance information of steel bar raw materials comprises the following steps:

[0092] Use the database module to store and manage project information, personnel role information, steel bar verification records, AI verification and recognition models, algorithm rules and manufacturer information;

[0093] An information management platform is built using the database module. The database module not only stores detailed project information and personnel role information, but also records the steel bar verification ledger for easy tracking and backtracking. The database module integrates AI verification and identification models and algorithm rules. These rules and models are based on a large amount of historical data and professional knowledge, and can accurately identify the authenticity of steel bar raw material quality assurance information. At the same time, the database module also includes detailed information on each manufacturer, providing data support for verification work.

[0094] Use the image processing module to take and pre-process the image of the steel bar raw material quality guarantee, and apply the icon algorithm and table recognition algorithm to extract the information in the image;

[0095] The image processing module can not only easily take pictures of the steel bar raw material quality guarantee, but also has pre-processing capabilities, and can resize and enhance the clarity of the pictures to ensure accurate information extraction. The image processing module also applies advanced icon algorithms and table recognition algorithms, which can automatically extract key information from the pictures, such as manufacturer logo, steel bar model, production date, etc., greatly improving work efficiency and accuracy.

[0096] Use the verification rule module to verify the extracted information according to the preset verification rules;

[0097] The verification rule module conducts strict verification and judgment on the information extracted by the image processing module according to the preset verification rules. These rules cover multiple dimensions, such as the integrity, consistency, and compliance of information, to ensure the authenticity and reliability of each piece of quality assurance information for steel bar raw materials.

[0098] In one possible implementation, the steps of image preprocessing include: adjusting image size, padding, cropping and normalization. Resizing can ensure that the image adapts to the requirements of different recognition algorithms, padding and cropping can remove irrelevant background information, and normalization can further improve the contrast and clarity of the image.

[0099] Use the YOLOv8 detection algorithm to identify icons in images; YOLOv8 can quickly and accurately identify key icons such as manufacturer logos and quality certification marks in complex image environments;

[0100] Use the SLENET table recognition model to identify the table data in the image. The SLENET model is a deep learning model designed for table recognition. The SLENET model has strong context understanding capabilities and fine pixel-level segmentation capabilities. Through the SLENET model, the table data in the image can be accurately identified and extracted, including key information such as steel bar model, specification, batch number, etc.

[0101] In a possible implementation, the steps of verification include: determining whether the icon is missing or has a large offset. On the one hand, it will check whether the icon is missing. The missing of any key icon may mean that there is a problem with the quality guarantee. On the other hand, it will also determine whether the icon has a large offset, because the offset will cause errors in the picture shooting or processing, thereby affecting the accuracy of the information.

[0102] Verify the integrity and consistency of basic information; this includes checking whether key information such as manufacturer, steel bar model, production date, etc. is complete, and whether there are any logical contradictions or inconsistencies between these information.

[0103] Verify the steel bar data according to the rules; this usually involves comparing the key indicators of the steel bar such as chemical composition and mechanical properties to ensure that they meet the relevant standards and regulations. Any data outside the prescribed range will be regarded as a sign of forgery or tampering.

[0104] Scan the QR code and compare the scan result with the data on the steel raw material quality guarantee. The QR code on the quality guarantee will be scanned and the scan result will be compared with the data on the paper guarantee. This step not only verifies the authenticity of the QR code, but also ensures the consistency of paper and electronic data, providing double protection for the authenticity of the steel raw material quality guarantee.

[0105] In a possible implementation, the method of assisting in identifying the quality assurance data of steel bar raw materials also includes: optimizing the performance and security protection of the database module; performance optimization means improving the data processing speed and response time by adjusting the database structure, optimizing query statements, adding cache mechanisms, etc., so as to meet the growing data storage and query needs. At the same time, security protection requires the use of encryption technology, access control mechanisms, and regular data backup and recovery strategies to ensure that the data in the database is not illegally accessed, tampered with, or lost.

[0106] The image processing module and the verification rule module are regularly updated and optimized. With the continuous advancement of technology and the increasing variety of application scenarios, new image processing algorithms and verification rules continue to emerge. In order to ensure the advancement and accuracy of the method, these two modules need to be updated regularly to introduce the latest technologies and algorithms. At the same time, optimization means fine-tuning the existing algorithms and rules to improve recognition efficiency and verification accuracy. These updates and optimizations can not only improve the overall performance of the system, but also better adapt to the ever-changing needs of identifying quality assurance materials for steel bar raw materials.

[0107] The present invention can automatically and quickly identify text, icons and table information in the quality assurance data of steel bar raw materials, thereby improving the recognition efficiency. At the same time, by using a deep learning algorithm to train a large number of samples, it can achieve accurate recognition of the quality assurance data of steel bar raw materials and reduce the error rate of manual review.

[0108] The present invention fully considers the differences in formats and rules of quality assurance certificates for steel bar raw materials produced by different manufacturers. By constructing a flexible recognition model and a counterfeit verification rule library, it achieves comprehensive support for different formats and quality assurance materials, improves the versatility and adaptability of the system, and enables the system to be applied in a wider range of scenarios.

[0109] It should be noted that the terms "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or apparatus.

[0110] 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 system for assisting in identifying quality assurance data of steel bar raw materials, characterized in that: include: Database module, used to store and manage project information, personnel role information, steel bar verification records, AI verification and recognition models, algorithm rules and manufacturer information; The image processing module is used to take and pre-process the image of the steel bar raw material quality guarantee, and apply the icon algorithm and table recognition algorithm to extract the information in the image; The authentication rule module performs authentication judgment on the extracted information according to the preset authentication rules.

2. The system for assisting in identifying quality assurance information of steel bar raw materials according to claim 1, characterized in that: The database module includes: Project table, used to store project information; Personnel table, used to store information about personnel and their roles; Steel bar ledger table, used to record the detailed ledger of steel bar verification; AI model table, which stores the parameters and data of the AI ​​false recognition model; An algorithm rule table, storing algorithm rules used for authentication; Manufacturer table, which stores basic information and specific rules of manufacturers.

3. The system for assisting in identifying quality assurance information of steel bar raw materials according to claim 1, characterized in that: The image processing module includes: A picture shooting and processing unit, used to assist in shooting pictures of the steel bar raw material quality guarantee certificate; The image preprocessing unit performs preprocessing such as resizing, padding, cropping, and normalization on the captured images; Icon algorithm recognition unit, which uses YOLOv8 detection algorithm to recognize icons in pictures; The picture table recognition unit uses the SLENET table recognition model to recognize the table data in the picture.

4. The system for assisting in identifying quality assurance information of steel bar raw materials according to claim 3, characterized in that: The icon algorithm recognition unit also includes: The icon recognition result judgment subunit is used to judge whether the recognized icon is missing or has a large offset.

5. The system for assisting in identifying quality assurance information of steel bar raw materials according to claim 3, characterized in that: The picture table recognition unit also includes: a table data extraction subunit, for extracting the identified table data; The table data preprocessing subunit is used to process the situation where the entire row of data is in one cell due to the absence of table vertical lines.

6. The system for assisting in identifying quality assurance information of steel bar raw materials according to claim 1, characterized in that: The verification rule module comprises: An icon verification unit is used to determine the authenticity of the steel bar raw material quality guarantee certificate based on the icon recognition result; Basic information verification unit, used to verify the integrity and consistency of basic information; A steel bar data verification unit, used to verify the steel bar data according to the rules; The QR code scanning data comparison and verification unit is used to scan the QR code and compare the scanning result with the data in the steel bar raw material quality guarantee certificate.

7. A method for assisting in identifying quality assurance information of steel bar raw materials, characterized in that: The following steps are involved: Use the database module to store and manage project information, personnel role information, steel bar verification records, AI verification and recognition models, algorithm rules and manufacturer information; Use the image processing module to take and pre-process the image of the steel bar raw material quality guarantee, and apply the icon algorithm and table recognition algorithm to extract the information in the image; The authentication rule module is used to perform authentication judgment on the extracted information according to the preset authentication rules.

8. The method for assisting in identifying quality assurance information of steel bar raw materials according to claim 7, characterized in that: The pre-processing steps include: Adjust image size, padding, cropping and normalization; Use the YOLOv8 detection algorithm to identify icons in images; Use the SLENET table recognition model to recognize tabular data in images.

9. The method for assisting in identifying quality assurance information of steel bar raw materials according to claim 7, characterized in that: The steps of the forgery verification include: Determine whether the icon is missing or has a large offset; Verify the integrity and consistency of basic information; Verify the authenticity of the steel bar data according to the rules; Scan the QR code and compare the scan result with the data on the steel bar raw material quality guarantee certificate.

10. The method for assisting in identifying quality assurance data of steel bar raw materials according to any one of claims 7 to 9, characterized in that: Also includes: Optimize performance and provide security protection for database modules; Regularly update and optimize the image processing module and counterfeit verification rule module.

Images