A method and system for intelligently identifying chart data

Abstract

The application discloses a kind of method and system for intelligently identifying chart data, it is related to intelligent identification technical field, method includes: obtaining chart data;The chart data is transformed, and relational data is obtained;According to the number of label column and the number of data column of relational data;According to the number of label column, the number of data column and relational data, the data report of m label column n data column is obtained;The data report of m label column n data column is labeled and merged, and the data report after label merging is obtained;According to the set label subject inter-checking relationship formula, the data in the data report after label merging is labeled and checked between label subjects, and error data is obtained;Error data is corrected, and the data report after correction is obtained;The data report after correction is shown.The application can improve the efficiency of data report data collection work.

Description

Technical Field

[0001] This invention relates to the field of intelligent recognition technology, and in particular to a method and system for intelligently recognizing chart data. Background Technology

[0002] Currently, data analysis of data reports is still largely based on manual review and approval, which is inefficient and cannot automate the approval process. In corporate credit reviews, traditional manual review methods have severely hampered efficiency and business operations. In fact, credit reviews are based on core data in data reports. Improving the efficiency of data collection for data reports will significantly improve the efficiency of corporate credit reviews. Therefore, how to improve the efficiency of data collection for data reports has become a pressing issue for those skilled in the art. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for intelligently identifying chart data, thereby improving the efficiency of data collection for data reports.

[0004] To achieve the above objectives, the present invention provides the following solution:

[0005] A method for intelligently identifying chart data, the method comprising:

[0006] Acquire chart data; the chart data includes image data in PNG, JPG, JPEG, BMP, and TIFF formats, as well as data reports in WORD, PDF, and EXCEL formats.

[0007] The chart data is transformed to obtain relational data;

[0008] The number of label columns and the number of data columns are obtained based on the relational data.

[0009] Based on the number of label columns, the number of data columns, and the relational data, a data report with m label columns and n data columns is obtained; where m represents the number of label columns and n represents the number of data columns; the data report with m label columns and n data columns includes the relational data represented by labels and data.

[0010] The data report with m label columns and n data columns is merging labels to obtain a data report with merged labels;

[0011] According to the set formula for the reconciliation relationship between label items, the data in the data report after the label merging is calculated to obtain erroneous data; the calculation result of the erroneous data does not conform to the set formula for the reconciliation relationship between label items.

[0012] The erroneous data is corrected to obtain a corrected data report;

[0013] The revised data report is then displayed.

[0014] Optionally, the transformation of the chart data to obtain relational data specifically includes:

[0015] The chart data was transformed using Baidu's image processing technology to obtain relational data.

[0016] Optionally, obtaining a data report with m labels and n data columns based on the number of label columns, the number of data columns, and the relational data specifically includes:

[0017] Match a data report template with m label columns and n data columns from the data report template library;

[0018] If a data report template with m labels and n data columns is matched, the relational data is represented in the form of a data report template with m labels and n data columns, and a data report with m labels and n data columns is obtained.

[0019] If no matching data report template for column m and column n is found, then add a data report template for column m and column n to the data report template library, and represent the relational data in the form of a data report template for column m and column n to obtain a data report for column m and column n.

[0020] Optionally, after correcting the erroneous data to obtain a corrected data report, the process further includes:

[0021] Using graphic positioning technology, the data in the corrected data report is located in the chart data according to the row and column number of the subject.

[0022] The present invention also provides the following solutions:

[0023] A system for intelligently recognizing chart data, the system comprising:

[0024] The chart data acquisition module is used to acquire chart data; the chart data includes image data in PNG, JPG, JPEG, BMP, and TIFF formats, as well as data reports in WORD, PDF, and EXCEL formats.

[0025] The chart data conversion module is used to convert the chart data to obtain relational data;

[0026] The module for obtaining the number of label columns and data columns is used to obtain the number of label columns and the number of data columns based on the relational data.

[0027] The module for obtaining an m-label column and n-data column data report is used to obtain an m-label column and n-data column data report based on the number of label columns, the number of data columns, and the relational data; wherein m represents the number of label columns; n represents the number of data columns; and the m-label column and n-data column data report includes the relational data represented by labels and data.

[0028] The label merging module is used to merge labels on the data report with m label columns and n data columns to obtain a data report with merged labels.

[0029] The tag account reconciliation calculation module is used to calculate the tag account reconciliation relationship of the data in the data report after tag merging according to the set tag account reconciliation relationship formula, and obtain erroneous data; the calculation result of the erroneous data does not conform to the set tag account reconciliation relationship formula.

[0030] The error data correction module is used to correct the error data and obtain a corrected data report;

[0031] The revised data report display module is used to display the revised data report.

[0032] Optionally, the chart data conversion module specifically includes:

[0033] The chart data conversion unit is used to convert the chart data using Baidu image processing technology to obtain relational data.

[0034] Optionally, the module for obtaining the data report from the m-label column and n-data column specifically includes:

[0035] The data report template matching unit is used to match data report templates with m label columns and n data columns from the data report template library.

[0036] The first relational data representation unit is used to represent the relational data in the form of a data report template with m labels and n data columns if a data report template with m labels and n data columns is matched, thereby obtaining a data report with m labels and n data columns;

[0037] The second relational data representation unit is used to add a data report template with m labels and n data columns to the data report template library if no matching data report template with m labels and n data columns is found, and to represent the relational data in the form of a data report template with m labels and n data columns, thereby obtaining a data report with m labels and n data columns.

[0038] Optionally, the system further includes:

[0039] The data positioning module is used to locate the data in the corrected data report to the chart data based on the row and column number of the subject using graphic positioning technology.

[0040] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0041] This invention discloses a method and system for intelligently identifying chart data. It provides a process for automatically identifying chart data, which involves transforming the chart data to obtain relational data, determining the number of label columns and data columns based on the relational data, and generating a data report with m label columns and n data columns, representing the relational data. The data report with m label columns and n data columns is then merging labels to obtain a merged data report. Based on a predefined formula for the inter-label reconciliation relationship, the data in the merged data report is used to calculate the inter-label reconciliation relationship. Error data that does not conform to the predefined formula is identified, and this error data is corrected to obtain a corrected data report. Finally, the corrected data report is displayed. This method achieves automatic data collection, significantly improving the efficiency of data collection compared to manual collection. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart illustrating an embodiment of the method for intelligently recognizing chart data according to the present invention;

[0044] Figure 2 This is a structural diagram of a system embodiment for intelligent recognition of chart data according to the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] The purpose of this invention is to provide a method and system for intelligently identifying chart data, thereby improving the efficiency of data collection for data reports.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] Figure 1 This is a flowchart illustrating an embodiment of the method for intelligently recognizing chart data according to the present invention. See also... Figure 1 The method for intelligently recognizing chart data includes:

[0049] Step 101: Obtain chart data; chart data includes image data in PNG, JPG, JPEG, BMP, and TIFF formats, as well as data reports in WORD, PDF, and EXCEL formats.

[0050] Step 102: Transform the chart data to obtain relational data.

[0051] Step 102 specifically includes:

[0052] The chart data was transformed using Baidu's image processing technology to obtain relational data.

[0053] Step 103: Obtain the number of label columns and the number of data columns based on the relational data.

[0054] Step 104: Based on the number of label columns, the number of data columns, and the relational data, obtain a data report with m label columns and n data columns; where m represents the number of label columns and n represents the number of data columns; the data report with m label columns and n data columns includes relational data represented by labels and data.

[0055] Step 104 specifically includes:

[0056] Match the data report template with m label columns and n data columns in the data report template library.

[0057] If a data report template with m labels and n data columns is matched, then the relational data is represented in the form of a data report template with m labels and n data columns, resulting in a data report with m labels and n data columns.

[0058] If no matching data report template for column m and column n is found, add a data report template for column m and column n to the data report template library, and represent the relational data in the form of a data report template for column m and column n to obtain a data report for column m and column n.

[0059] Step 105: Perform label merging on the data report of m label column and n data column to obtain the data report after label merging.

[0060] Step 106: Based on the set formula for the reconciliation relationship between label items, calculate the reconciliation relationship between label items in the data report after label merging, and obtain erroneous data; the calculation result of the erroneous data does not conform to the set formula for the reconciliation relationship between label items.

[0061] Step 107: Correct the erroneous data to obtain the corrected data report.

[0062] Step 107 is followed by:

[0063] Using graphic positioning technology, the data in the corrected data report is located in the chart data according to the row and column number of the subject.

[0064] Step 108: Display the revised data report.

[0065] The technical solution of the present invention is illustrated below with a specific embodiment:

[0066] This invention provides a novel method for intelligently recognizing chart data, a novel technology for automatically recognizing data reports. This technology can help enterprises, tax authorities, and auditors improve efficiency and build automated credit review systems. The specific solution for this invention's intelligent chart data recognition method is as follows:

[0067] Step 1: Multi-format Support: Supports various image data types such as PNG, JPG, JPEG, BMP, and TIFF, as well as data reports from different document types such as WORD, PDF, and EXCEL. It intelligently identifies chart data types and recognizes reports of different formats. Users only need to import the files, and Baidu's image processing technology will convert the text, tables, and images on the images (videos) into specific data (relational data) and store it in the database. Data collection is then achieved through a computer program. Step 1 utilizes Baidu to identify and store data in relational data format.

[0068] Step 2: Template Matching: Report templates are continuously updated to create a data report template library. To address report differences caused by varying update details and industry variations, the intelligent chart data recognition method automatically matches templates for uploaded files of the same type (the user's previous file to be recognized) and also supports custom template models on the page. Examples include: 123 Data Report (i.e., a data report with 1 label column and 23 data columns), 134 Data Report (i.e., a data report with 1 label column and 34 data columns), 134578 Data Report (i.e., a data report with 1 label column and 34578 data columns), etc. Manual template replacement is also supported, allowing precise selection of label and data columns during data entry and specifying the number of chart columns. Once successful, template types are added. Template matching involves matching the corresponding template type from the database (data report template library) based on the user's selected type. The main function of template matching is to facilitate locating data areas and subject areas by defining the relationship between labels and values ​​and their specific display positions according to the template specifications.

[0069] Step 3: Refine Recognition Rules: The intelligent chart data recognition method supports manual modification of label relationships and label merging. New label matching rules can be generated for missing label rules. For example, different needs may require different data or different data names; mappings can be added to selected templates. After backend review, the system will automatically fill in the company's saved report templates according to accounting standards. The mapping will be automatically completed the next time a similar situation is encountered. Label merging means that label A and label B have the same data recognition. Label matching rules mean that label A can be precisely matched while also being configured with multiple fuzzy matching schemes. For example, labels B, C, and D can be treated as label A after being recognized. This step 3 mainly addresses data attribution issues. The data in step 1 obtains its specific meaning through the correspondence between labels and data. Simultaneously, label definition rules can be refined during use to achieve multiple labels belonging to the same definition and multiple labels having fuzzy belonging to the same definition.

[0070] Step 4: Trial Balance Error Alert: Unlike traditional image recognition software, the intelligent data report recognition method (intelligent chart data recognition method) is deeply optimized based on the data, accurate to every single value. It identifies each data point as its corresponding label category and uses the recognition results to assist in calculating the cross-correlation relationship between label categories. If the calculation result does not conform to the formula (the set formula for the cross-correlation relationship between label categories), such as A = B + C / D, it will prompt business personnel to perform manual verification, effectively ensuring data accuracy. For example, in the balancing formula, if an identification result is intentionally changed incorrectly, the data that cannot be balanced will receive a special prompt, helping to quickly check and correct it. This function is quite useful when a large number of tables need to be recognized and entered. Clicking on the recognition result row allows you to view the original image (chart data) of the corresponding area, facilitating result comparison. In addition, the recognition platform also has an automatic balancing function, with intelligent reminders for unbalanced areas. The purpose of Step 4 is to use a custom formula to calculate the desired verification result from the result of Step 1, and to ensure the correctness of data recognition based on the verification result. It also enables quick location of erroneous identification positions, thereby correcting the incorrect data and obtaining a corrected data report.

[0071] Step 5: Subject Matching and Original Image Location: Supports data location and comparison within the original image. Clicking on data (i.e., data already processed in Step 3, with its specific meaning ensured through label relationships) will locate its corresponding position in the original image. For example, data or fields can be located and centered in the original image. Intelligent reminders will also be provided for re-balancing areas, facilitating proofreading and review. This is a very user-friendly feature. Clicking on data uses graphical positioning technology, locating data based on the row and column number of the subject. The specific location is determined by the subject name corresponding to the number.

[0072] Step 6: Standardized Output: The recognition results, i.e., the final data set after processing in steps 3, 4, and 5 (the final data after template processing, label attribution confirmation, and intelligent recognition verification), support exporting to an Excel file, thus enabling the display of corrected data reports. Original Image Positioning Comparison: Supports data positioning comparison within the original image; clicking on the data will locate the corresponding position in the original image, facilitating verification. Manual Input: Supports manual input for individual data points that failed recognition. Output methods for recognition results include report display, page display, etc.

[0073] Step 7: Template Settings: This step displays the existing template list, allows editing of the selected template and addition of rules, shows the recognition history, and displays entries for the recognized data reports (output results). It continuously improves the templates to expand the application scenarios of this technology. The main task of Step 7 is no longer focused on the recognition result itself; its primary function is to record the template rules used, refine the template types, and improve the label definitions.

[0074] This invention provides a method for intelligently recognizing chart data. It leverages efficient data processing capabilities to analyze data from specified reports and automatically generate relevant analytical charts. It also enables multi-dimensional analysis, dynamically and flexibly sets mapping relationships, reduces manual costs, and improves company management efficiency. The final presentation of this invention involves continuously improving the template library during use and defining tag relationships and tag merging functions. The intelligent recognition results are then displayed on a webpage in the form of a data table.

[0075] This invention provides an intelligent method for recognizing chart data, utilizing optical character recognition (OCR) technology. This involves converting text, tables, and images in a picture into electronic data through image processing techniques, and then rapidly collecting this data via a computer program. Since data reports often contain large amounts of data that require manual entry, this data report recognition technology will significantly improve the efficiency and accuracy of data collection.

[0076] Figure 2 This is a structural diagram of a system embodiment for intelligently recognizing chart data according to the present invention. See also... Figure 2 The system for intelligently recognizing chart data includes:

[0077] The chart data acquisition module 201 is used to acquire chart data; the chart data includes image data in PNG, JPG, JPEG, BMP, and TIFF formats, as well as data reports in WORD, PDF, and EXCEL formats.

[0078] The chart data conversion module 202 is used to convert chart data into relational data.

[0079] The chart data transformation module 202 specifically includes:

[0080] The chart data transformation unit is used to transform chart data using Baidu's image processing technology to obtain relational data.

[0081] The module 203, which obtains the number of label columns and data columns, is used to obtain the number of label columns and data columns based on relational data.

[0082] The module 204, which generates a data report with m labels and n data columns, is used to generate a data report with m labels and n data columns based on the number of label columns, the number of data columns, and relational data. Here, m represents the number of label columns, and n represents the number of data columns. The data report with m labels and n data columns includes relational data represented by labels and data.

[0083] The data report obtained from the m-label column and n-data column in module 204 specifically includes:

[0084] The data report template matching unit is used to match data report templates with m label columns and n data columns from the data report template library.

[0085] The first relational data representation unit is used to represent relational data in the form of a data report template with m labels and n data columns if a data report template with m labels and n data columns is matched, thereby obtaining a data report with m labels and n data columns.

[0086] The second relational data representation unit is used to add a data report template with m labels and n data columns to the data report template library if no matching data report template with m labels and n data columns is found, and to represent relational data in the form of a data report template with m labels and n data columns, thereby obtaining a data report with m labels and n data columns.

[0087] The label merging module 205 is used to merge labels on a data report with m label columns and n data columns to obtain a data report with merged labels.

[0088] The tag account reconciliation calculation module 206 is used to calculate the tag account reconciliation relationship of the data in the data report after tag merging according to the set tag account reconciliation relationship formula, and obtain erroneous data; the calculation result of the erroneous data does not conform to the set tag account reconciliation relationship formula.

[0089] Error data correction module 207 is used to correct error data and obtain a corrected data report.

[0090] The revised data report display module 208 is used to display the revised data reports.

[0091] Specifically, the system for intelligently recognizing chart data also includes:

[0092] The data positioning module is used to locate the data in the corrected data report into the chart data based on the row and column number of the subject, using graphic positioning technology.

[0093] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0094] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for intelligently recognizing chart data, characterized in that, The method includes: Acquire chart data; the chart data includes image data in PNG, JPG, JPEG, BMP, and TIFF formats, as well as data reports in WORD, PDF, and EXCEL formats. The chart data is transformed to obtain relational data; The number of label columns and the number of data columns are obtained based on the relational data. A data report with m labels and n data columns is obtained based on the number of label columns, the number of data columns, and the relational data; where m represents the number of label columns and n represents the number of data columns; the data report with m labels and n data columns includes the relational data represented by labels and data; specifically, obtaining the data report with m labels and n data columns based on the number of label columns, the number of data columns, and the relational data includes: matching a data report template for m labels and n data columns in a data report template library; if a data report template for m labels and n data columns is matched, the relational data is represented in the form of the data report template for m labels and n data columns to obtain the data report with m labels and n data columns; if no data report template for m labels and n data columns is matched, m labels and n data columns are added to the data report template library. Based on the data report template, the relational data is represented in the form of a data report template with m label columns and n data columns, resulting in a data report with m label columns and n data columns. The report templates are continuously improved to obtain a data report template library that includes these continuously improved templates. To address report differences caused by update details and industry variations, the intelligent chart data identification method automatically matches templates for uploaded files of the same type and also supports custom template models on the page. Manual template replacement is also supported, i.e., precisely selecting label columns and data columns and specifying the number of chart columns during data entry. Once successful, template types are added. Template matching involves matching the corresponding template type from the data report template library based on the user-selected type. Template matching is used to locate data areas and subject areas based on the relationship between labels and values ​​and their specific display positions according to the template specifications. The data report with m labels and n data columns is merging labels to obtain a data report with merged labels. The intelligent chart data recognition method supports manual modification of label relationships and label merging. New label matching rules are generated for missing label rules. Different needs require different data or different data names, so mappings are added to the selected template. After background review, the report template saved by the company is automatically filled in according to accounting standards. The mapping is automatically completed the next time the same situation is encountered. When the same situation is encountered again, the mapping matching is performed directly. Among them, label merging means that label A and label B have the same data recognition. Label matching rules mean that label A is configured with multiple fuzzy matching schemes while matching precisely. That is, after labels B, C, D, etc. are identified, they are processed as label A. Based on the established formula for the reconciliation relationship between label items, the data in the data report after label merging is used to calculate the reconciliation relationship between label items, resulting in erroneous data. The calculation result of the erroneous data does not conform to the established formula for the reconciliation relationship between label items. The method for intelligently recognizing chart data is based on data that has undergone deep optimization, accurate to each value, identifying each data as a corresponding label item, and using the recognition result data to perform auxiliary calculation of the reconciliation relationship between label items. If the calculation result does not conform to the established formula for the reconciliation relationship between label items, it will prompt business personnel to perform manual verification. Clicking on the recognition result row will display the original image of the corresponding area, i.e., the chart data, facilitating result comparison. In addition, the recognition platform also has an automatic balancing function, and there will be intelligent reminders for uneven areas. The erroneous data is corrected to obtain a corrected data report; The revised data report is then displayed.

2. The method for intelligently recognizing chart data according to claim 1, characterized in that, The process of transforming the chart data to obtain relational data specifically includes: The chart data was transformed using Baidu's image processing technology to obtain relational data.

3. The method for intelligently recognizing chart data according to claim 1, characterized in that, The step of correcting the erroneous data to obtain a corrected data report further includes: Using graphic positioning technology, the data in the corrected data report is located in the chart data according to the row and column number of the subject.

4. A system for intelligently recognizing chart data, characterized in that, The system includes: The chart data acquisition module is used to acquire chart data; the chart data includes image data in PNG, JPG, JPEG, BMP, and TIFF formats, as well as data reports in WORD, PDF, and EXCEL formats. The chart data conversion module is used to convert the chart data to obtain relational data; The module for obtaining the number of label columns and data columns is used to obtain the number of label columns and the number of data columns based on the relational data. The module for obtaining an m-label column and n-data column data report is used to obtain a data report of m-label column and n-data column based on the number of label columns, the number of data columns, and the relational data; where m represents the number of label columns and n represents the number of data columns; the data report of m-label column and n-data column includes the relational data represented by labels and data; the module for obtaining an m-label column and n-data column data report specifically includes: a data report template matching unit, used to match a data report template of m-label column and n-data column in a data report template library; a first relational data representation unit, used to represent the relational data in the form of the data report template of m-label column and n-data column if a data report template of m-label column and n-data column is matched, thereby obtaining a data report of m-label column and n-data column; and a second relational data representation unit, used to represent the relational data in the form of the data report template of m-label column and n-data column if no data report template of m-label column and n-data column is matched. Add data report templates with m label columns and n data columns to the data report template library, and represent the relational data in the form of data report templates with m label columns and n data columns to obtain data reports with m label columns and n data columns; the report templates are improved periodically to obtain a data report template library that includes periodically improved report templates; to address the differences in reports caused by update details and industry differences, the intelligent identification method for chart data will automatically match templates for uploaded files of the same type and also support page-customized template models; it also supports manual template replacement, i.e., accurately selecting label columns and data columns and specifying the number of chart columns during data entry; after successful implementation, template types are added; template matching is used to match the corresponding template type in the data report template library according to the type selected by the user; template matching is used to locate the data area and subject area position of the identified data results according to the relationship between labels and values ​​and the specific display position of the template specifications; The label merging module is used to merge labels on the data report with m label columns and n data columns to obtain a data report with merged labels. The intelligent method for recognizing chart data supports manual modification of label relationships and label merging. It generates new label matching rules for missing label rules. Different needs require different data or different data names, so mappings are added to the selected template. After background review, the system will automatically fill in the report template saved by the company according to accounting standards. The system will automatically complete the mapping the next time the same situation is encountered. When the same situation is encountered again, the mapping matching will be performed directly. Among them, label merging means that label A and label B have the same data recognition. The label matching rule means that label A is configured with multiple fuzzy matching schemes while performing exact matching. That is, after labels B, C, D, etc. are identified, they are processed as label A. The tag-subject cross-reference calculation module is used to calculate the cross-reference relationships between tag subjects in the data report after tag merging according to the set cross-reference relationship formula, and to obtain erroneous data. The calculation result of the erroneous data does not conform to the set cross-reference relationship formula. The intelligent chart data recognition method is based on data deep optimization, accurate to each value, and identifies each data as the corresponding tag subject. The recognition result data is used to assist in the cross-reference relationship calculation between tag subjects. If the calculation result does not conform to the set cross-reference relationship formula, it will prompt business personnel to perform manual verification. Clicking on the recognition result row can view the original image of the corresponding area, i.e., the chart data, to facilitate result comparison. In addition, the recognition platform also has an automatic balancing function, and there will be intelligent reminders for uneven areas. The error data correction module is used to correct the error data and obtain a corrected data report; The revised data report display module is used to display the revised data report.

5. The system for intelligently recognizing chart data according to claim 4, characterized in that, The chart data conversion module specifically includes: The chart data conversion unit is used to convert the chart data using Baidu image processing technology to obtain relational data.

6. The system for intelligently recognizing chart data according to claim 4, characterized in that, The system also includes: The data positioning module is used to locate the data in the corrected data report to the chart data based on the row and column number of the subject using graphic positioning technology.

Images