WO2021047182A1 - 基于ocr的图片数据识别方法、装置、及计算机设备 - Google Patents

基于ocr的图片数据识别方法、装置、及计算机设备 Download PDF

Info

Publication number
WO2021047182A1
WO2021047182A1 PCT/CN2020/087132 CN2020087132W WO2021047182A1 WO 2021047182 A1 WO2021047182 A1 WO 2021047182A1 CN 2020087132 W CN2020087132 W CN 2020087132W WO 2021047182 A1 WO2021047182 A1 WO 2021047182A1
Authority
WO
WIPO (PCT)
Prior art keywords
picture
recognized
standardized
value
pictures
Prior art date
Application number
PCT/CN2020/087132
Other languages
English (en)
French (fr)
Inventor
张�杰
Original Assignee
深圳壹账通智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳壹账通智能科技有限公司 filed Critical 深圳壹账通智能科技有限公司
Publication of WO2021047182A1 publication Critical patent/WO2021047182A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40—Document-oriented image-based pattern recognition
    • G06V30/41—Analysis of document content
    • G06V30/413—Classification of content, e.g. text, photographs or tables
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/04—Billing or invoicing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition

Definitions

  • This application relates to the field of image recognition technology, and in particular to an OCR-based image data recognition method, device, computer equipment and storage medium.
  • the reimbursement staff needs to fill in the reimbursement form and paste the invoice, and then the financial staff will review and calculate whether the invoice amount is consistent with the amount in the form.
  • the embodiments of the application provide an OCR-based image data recognition method, device, computer equipment, and storage medium, which are intended to solve the reimbursement filled in by the reimbursement personnel in the online reimbursement function module of the online office collaboration system in the prior art
  • Information and scanned files are only saved for users to query historical data.
  • Manual calculations are still required based on the reimbursement form and pasted invoices. The manual calculation process is cumbersome, resulting in low calculation efficiency and error-prone problems.
  • an embodiment of the present application provides an OCR-based image data recognition method, which includes: receiving a set of pictures to be recognized uploaded by an uploader; and rotating all non-forward pictures in the set of pictures to be recognized to obtain a standard Forward pictures to update the set of pictures to be recognized to obtain a set of standardized pictures to be recognized; obtain the picture types corresponding to each standardized picture to be recognized in the set of standardized pictures to be recognized; wherein, the picture types include special values corresponding to value-added tax
  • the first picture type of an invoice or ordinary value-added tax invoice corresponds to the second picture type of machine-printed invoices and the third picture type of fixed-value invoices; the standardized pictures to be recognized in the set of standardized pictures to be recognized are acquired through image recognition Recognition values corresponding to preset designated areas; obtain the number of pictures of each picture type in the standardized picture set to be recognized to obtain the total number of pictures, and create a sub-data table corresponding to the number of rows according to the number of pictures of each picture type.
  • Form a total data table fill in the identification values corresponding to each standardized picture to be identified into the corresponding sub-data tables for storage, respectively sum the identification values of each sub-data table and then accumulate the sum to obtain the corresponding data table The actual total value; and sending the actual total value to the uploader.
  • an OCR-based image data recognition device which includes:
  • the picture collection receiving unit is used to receive the to-be-identified picture collection uploaded by the uploader;
  • a picture standardization unit configured to rotate all non-forward pictures in the picture set to be recognized to obtain a standard forward picture, so as to update the picture set to be recognized to obtain a standardized picture set to be recognized;
  • the picture type obtaining unit is configured to obtain the picture types corresponding to each standardized picture to be recognized in the standardized picture to be recognized; wherein, the picture type includes the first picture type corresponding to the special value-added tax invoice or the ordinary value-added tax invoice, Corresponding to the second picture type of machine-printed invoices, and corresponding to the third picture type of fixed-amount invoices;
  • An identification value acquisition unit configured to acquire, through image recognition, the identification values corresponding to the preset designated areas in each standardized to-be-recognized picture set in the standardized to-be-recognized pictures;
  • the total data table obtaining unit is used to obtain the number of pictures of each picture type in the standardized picture to be identified to obtain the total number of pictures, and create a sub-data table corresponding to the number of rows according to the number of pictures of each picture type to form the total data table;
  • the summation unit is used to fill the identification value corresponding to each standardized picture to be identified into the corresponding sub-data table for storage, respectively sum the identification values of each sub-data table and then accumulate the sum to obtain the corresponding total data table The actual total value of;
  • the sum value sending unit is used to send the actual sum value to the uploader.
  • an embodiment of the present application provides a computer device, which includes a memory, a processor, and a computer program stored on the memory and running on the processor, and the processor executes the computer
  • the program implements an OCR-based image data recognition method, which includes: receiving a set of pictures to be recognized uploaded by the uploader; rotating all non-forward pictures in the set of pictures to be recognized to obtain a standard forward picture to update all pictures.
  • the picture set to be recognized obtains a standardized picture set to be recognized; the picture type corresponding to each standardized picture to be recognized in the standardized picture set to be recognized is obtained; wherein, the picture type includes a special value-added tax invoice or an ordinary value-added tax invoice
  • the first picture type corresponds to the second picture type of the machine-printed invoice and the third picture type of the fixed invoice; the designated areas in the standardized pictures to be recognized in the standardized pictures to be recognized in the standardized pictures to be recognized are obtained through image recognition, respectively Obtain the number of pictures of each picture type in the standardized to-be-recognized picture set to obtain the total number of pictures, and create a sub-data table corresponding to the number of rows according to the number of pictures of each picture type to form a total data table; Standardize the identification value corresponding to the picture to be identified and fill it into the corresponding sub-data table for storage, respectively sum the identification values of each sub-data table and then accumulate the sum to obtain the actual sum value corresponding to the total data table;
  • the embodiments of the present application also provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes a An OCR-based image data recognition method, which includes: receiving a set of pictures to be recognized uploaded by an uploader; rotating all non-forward pictures in the picture set to be recognized to obtain a standard forward picture, so as to update the picture set to be recognized Obtain a set of standardized to-be-recognized pictures; obtain the picture type corresponding to each standardized picture to be recognized in the standardized-to-be-recognized picture set; wherein, the picture type includes the first picture type corresponding to a special value-added tax invoice or a general value-added tax invoice, The second picture type corresponding to the machine-printed invoice corresponds to the third picture type of the fixed invoice; the recognition value corresponding to the preset designated area in each standardized to-be-recognized picture in the standardized to-be-
  • the embodiments of the present application provide an OCR-based image data recognition method, device, computer equipment, and storage medium. This method realizes that after the non-forward pictures are rotated to obtain the standard forward picture, the invoice amount is recognized and the calculation is performed through the image recognition technology, and the calculation efficiency is improved, and the calculation accuracy rate is high.
  • FIG. 1 is a schematic diagram of an application scenario of an OCR-based image data recognition method provided by an embodiment of the application
  • FIG. 2 is a schematic flowchart of an OCR-based image data recognition method provided by an embodiment of the application
  • FIG. 3 is a schematic diagram of another process of the OCR-based image data recognition method provided by an embodiment of the application.
  • Fig. 4a is a schematic diagram of a non-forward picture in an OCR-based picture data recognition method provided by an embodiment of the application;
  • 4b is a schematic diagram of a standard forward picture in an OCR-based picture data recognition method provided by an embodiment of the application;
  • FIG. 5 is a schematic diagram of a sub-flow of an OCR-based image data recognition method provided by an embodiment of this application;
  • FIG. 6 is a schematic block diagram of an OCR-based image data recognition apparatus provided by an embodiment of the application.
  • FIG. 7 is another schematic block diagram of an OCR-based image data recognition apparatus provided by an embodiment of the application.
  • FIG. 8 is a schematic block diagram of subunits of an OCR-based image data recognition apparatus provided by an embodiment of the application.
  • FIG. 9 is a schematic block diagram of a computer device provided by an embodiment of the application.
  • FIG. 1 is a schematic diagram of an application scenario of an OCR-based image data recognition method provided by an embodiment of the application
  • FIG. 2 is a schematic flowchart of an OCR-based image data recognition method provided by an embodiment of the application.
  • the OCR-based image data recognition method is applied to the server, and the method is executed by application software installed in the server.
  • the method includes steps S110 to S170.
  • the specific application scenario is financial reimbursement
  • you need to operate the uploader such as smart phone, tablet computer, etc.
  • the server calculates the reimbursement amount based on the uploaded scanned documents or photos of the invoice, without the need for manual calculation by the user.
  • S120 Rotate all non-forward pictures in the picture set to be recognized to obtain a standard forward picture, so as to update the picture set to be recognized to obtain a standardized picture set to be recognized.
  • the server since there may be pictures whose scanning direction is not the positive direction in the picture set to be recognized, the server needs to rotate all the non-forward pictures in the picture set to be recognized to obtain the standard forward picture.
  • the standard of all pictures to be recognized is normalized.
  • the method before step S120, the method further includes:
  • the scanning direction of the scanned invoice file may be included in it is not the positive direction (the positive direction of the scanned invoice file refers to the connection between the center points of each text on the invoice header).
  • the angle between the obtained direction line and the bottom edge of the scanned document page is 0, that is, the two are parallel, and the ticket header is located at the top of the scanned document).
  • the non-forward picture shown in Figure 4a the non-forward picture shown in Figure 4a.
  • the rotation angle can be obtained based on the position of the first line of text corresponding to the non-forward image and the corresponding position of the same text in the corresponding standard forward image.
  • the first line of text recognized in the non-forward picture as shown in FIG. 4a is "the Xth link: XX link”; the above-mentioned characters are in the middle of the upper side of the non-forward picture.
  • the corresponding position of the same text "Xth link: XX link” is in the middle right part of the standard forward picture.
  • step S1201 includes:
  • the OCR image recognition model is used to first recognize the first line of characters of each to-be-recognized picture in the to-be-recognized picture set, which uses the principle of scanning from left to right line by line using OCR technology.
  • OCR technology is the abbreviation of Optical Character Recognition (Optical Character Recognition), which converts the text of various bills, newspapers, books, manuscripts and other printed materials into image information through optical input methods such as scanning, and then uses text recognition technology to convert image information It is a computer input technology that can be used. It can be applied to the input and processing fields of bank bills, large amounts of text data, file files, and copywriting. It is suitable for automatic scanning identification and long-term storage of a large number of bill forms in banking, taxation and other industries.
  • Optical Character Recognition Optical Character Recognition
  • the first line of text does not include the keywords in the preset first keyword list (for example, the first keyword list that is set first includes special invoices, ordinary invoices, fixed invoices and other keywords), it means that the picture to be recognized is Non-positive image.
  • the invoice issued by the on-board terminal of a taxi is a machine-printed invoice
  • the invoice issued by a general taxpayer to an individual or other general taxpayer is a special value-added tax invoice or a general value-added tax invoice.
  • the parking ticket is a fixed invoice.
  • Invoice content generally includes: ticket header, character track number, number and purpose, customer name, bank account number, business (product) product name or business item, measurement unit, quantity, unit price, amount, as well as upper and lower case amount, and person who handles it , Unit seal, invoice date, etc.
  • the special value-added tax invoices used by units that implement value-added tax should also include tax types, tax rates, and tax amounts.
  • step S130 includes:
  • the OCR image recognition model is used to identify the header of each standardized picture to be recognized, so as to obtain the picture type corresponding to each standardized picture to be recognized.
  • the ticket header of each standardized picture to be recognized can be recognized through the OCR image recognition model, and then each picture in the picture set to be recognized can be obtained.
  • the header of a standardized picture to be recognized is a special XXX value-added tax invoice, which indicates that the picture type of the standardized picture to be recognized is the first picture type.
  • S140 Obtain, through image recognition, the recognition values corresponding to the designated areas preset in the standardized pictures to be recognized in the standardized pictures to be recognized respectively.
  • the keyword total or the total price and tax is included.
  • the recognition value after the total or total price and tax keyword can be obtained (for example, the total price and tax in Figure 4b The value shown after the column).
  • step S140 includes:
  • S142 Locate and obtain the text content in the text of the picture content of each standardized picture to be recognized that is the same as the keyword in the preset second keyword list, and use the corresponding value after the text content as the recognition value corresponding to each standardized picture to be recognized.
  • the keyword "price and tax total” set in the second keyword list is located in each image content text.
  • the values after the keyword (such as 300, 14) are respectively obtained, and the corresponding value after the text content is used as the recognition value corresponding to each standardized image to be recognized.
  • the sub-data table corresponding to each picture type is created to correspondingly store the recognition value of the standardized picture to be recognized of that type, so as to facilitate subsequent summation and use. For example, there are 10 standardized pictures to be recognized for the first picture type, and 10 recognition values are obtained after recognition. Then the above 10 recognition values are stored in the first sub-data table corresponding to the first picture type; the same way is obtained
  • the second sub-data table corresponding to the second picture type and the third sub-data table corresponding to the third picture type are composed of the first sub-data table, the second sub-data table, and the third sub-data table to form a total data table.
  • the identification values of each sub-data table are respectively summed and then accumulated and summed to obtain the sum of the identification values corresponding to each standardized image to be identified in the standardized image to be identified, that is, to obtain all uploaded
  • the total invoice value of the scanned invoice file is recorded as the actual total value.
  • the actual total value can be sent to the uploader to notify the server that the automatic verification of the invoice amount has been completed, and the user can Proceed to the next step.
  • step S170 the method further includes:
  • the second notification information for notifying that the review has not passed is sent to the uploader.
  • the uploader after the uploader receives the actual total value, it can also choose to set the expected amount of expected reimbursement (understood as a target value).
  • This target value is directly uploaded to the server and calculated before.
  • the actual total value is compared. If the actual total value is greater than or equal to the target value, it means that the expected amount of expected reimbursement is less than or equal to the actual total value, and the reimbursement process can pass the review and continue. If the actual total value is less than the target value, it means that the expected amount of reimbursement is greater than the actual total value, and it cannot be reviewed and the user is prompted to continue uploading another set of pictures to be identified or reduce the target value until it is less than or equal to the total value. After the actual total value is stated, the reimbursement process can be continued.
  • This method realizes that after the non-forward pictures are rotated to obtain the standard forward picture, the invoice amount is recognized and the calculation is performed through the image recognition technology, and the calculation efficiency is improved, and the calculation accuracy rate is high.
  • An embodiment of the present application also provides an OCR-based picture data recognition device, which is used to execute any embodiment of the aforementioned OCR-based picture data recognition method.
  • FIG. 6, is a schematic block diagram of an OCR-based image data recognition apparatus provided by an embodiment of the present application.
  • the OCR-based image data recognition device 100 can be configured in a server.
  • the OCR-based picture data recognition device 100 includes a picture set receiving unit 110, a picture standardization unit 120, a picture type acquisition unit 130, an identification value acquisition unit 140, a total data table acquisition unit 150, a summation unit 160, Sum value sending unit 170.
  • the picture collection receiving unit 110 is configured to receive the to-be-identified picture collection uploaded by the uploader.
  • the specific application scenario is financial reimbursement
  • you need to operate the uploader such as smart phone, tablet computer, etc.
  • the server calculates the reimbursement amount based on the uploaded scanned documents or photos of the invoice, without the need for manual calculation by the user.
  • the picture standardization unit 120 is configured to rotate all non-forward pictures in the picture set to be recognized to obtain a standard forward picture, so as to update the picture set to be recognized to obtain a standardized picture set to be recognized.
  • the server since there may be pictures whose scanning direction is not the positive direction in the picture set to be recognized, the server needs to rotate all the non-forward pictures in the picture set to be recognized to obtain the standard forward picture.
  • the standard of all pictures to be recognized is normalized.
  • the OCR-based image data recognition apparatus 100 further includes:
  • the non-forward picture judging unit 1201 is used to determine whether there is a non-forward picture in the set of pictures to be recognized; if there is a non-forward picture in the set of pictures to be recognized, execute the position according to the first line of text corresponding to the non-forward picture The step of obtaining the rotation angle at the position corresponding to the same text in the corresponding standard forward picture; if there is no non-forward picture in the set of pictures to be recognized, perform the step of obtaining the pictures corresponding to each standard picture to be recognized in the set of standardized pictures to be recognized Type of steps;
  • the rotation angle obtaining unit 1202 is configured to obtain the rotation angle according to the position of the first line of text corresponding to the non-forward image and the corresponding position of the same text in the corresponding standard forward image.
  • the scanning direction of the scanned invoice file may be included in it is not the positive direction (the positive direction of the scanned invoice file refers to the connection between the center points of each text on the invoice header).
  • the angle between the obtained direction line and the bottom edge of the scanned document page is 0, that is, the two are parallel, and the ticket header is located at the top of the scanned document).
  • the non-forward picture shown in Figure 4a the non-forward picture shown in Figure 4a.
  • the rotation angle can be obtained based on the position of the first line of text corresponding to the non-forward image and the corresponding position of the same text in the corresponding standard forward image.
  • the first line of text recognized in the non-forward picture as shown in FIG. 4a is "the Xth link: XX link”; the above-mentioned characters are in the middle of the upper side of the non-forward picture.
  • the corresponding position of the same text "Xth link: XX link” is in the middle right part of the standard forward picture.
  • the non-forward picture determining unit 1201 is further configured to:
  • the OCR image recognition model is used to first recognize the first line of characters of each to-be-recognized picture in the to-be-recognized picture set, which uses the principle of scanning from left to right line by line using OCR technology.
  • OCR technology is the abbreviation of Optical Character Recognition (Optical Character Recognition), which converts the text of various bills, newspapers, books, manuscripts and other printed materials into image information through optical input methods such as scanning, and then uses text recognition technology to convert image information It is a computer input technology that can be used. It can be applied to the input and processing fields of bank bills, large amounts of text data, file files, and copywriting. It is suitable for automatic scanning identification and long-term storage of a large number of bill forms in banking, taxation and other industries.
  • Optical Character Recognition Optical Character Recognition
  • the first line of text does not include the keywords in the preset first keyword list (for example, the first keyword list that is set first includes special invoices, ordinary invoices, fixed invoices and other keywords), it means that the picture to be recognized is Non-positive image.
  • the picture type obtaining unit 130 is configured to obtain the picture type corresponding to each standardized picture to be recognized in the standardized picture to be recognized; wherein, the picture type includes the first picture type corresponding to the special value-added tax invoice or the ordinary value-added tax invoice , Corresponds to the second picture type of machine-printed invoices, and corresponds to the third picture type of fixed-amount invoices.
  • the invoice issued by the on-board terminal of a taxi is a machine-printed invoice
  • the invoice issued by a general taxpayer to an individual or other general taxpayer is a special value-added tax invoice or a general value-added tax invoice.
  • the parking ticket is a fixed invoice.
  • Invoice content generally includes: ticket header, character track number, number and purpose, customer name, bank account number, business (product) product name or business item, measurement unit, quantity, unit price, amount, as well as upper and lower case amount, and person who handles it , Unit seal, invoice date, etc.
  • the special value-added tax invoices used by units that implement value-added tax should also include tax types, tax rates, and tax amounts.
  • the picture type obtaining unit 130 is further configured to:
  • the OCR image recognition model is used to identify the header of each standardized picture to be recognized, so as to obtain the picture type corresponding to each standardized picture to be recognized.
  • the ticket header of each standardized picture to be recognized can be recognized through the OCR image recognition model, and then each picture in the picture set to be recognized can be obtained.
  • the header of a standardized picture to be recognized is a special XXX value-added tax invoice, which indicates that the picture type of the standardized picture to be recognized is the first picture type.
  • the recognition value obtaining unit 140 is configured to obtain, through image recognition, the recognition values corresponding to the preset designated areas in each standardized to-be-recognized picture set in the standardized to-be-recognized pictures.
  • the keyword total or the total price and tax is included.
  • the recognition value after the total or total price and tax keyword can be obtained (for example, the total price and tax in Figure 4b The value shown after the column).
  • the identification value acquisition unit 140 includes:
  • the picture content text obtaining unit 141 is configured to obtain the picture content text corresponding to each standardized picture to be recognized in the standardized picture to be recognized;
  • the keyword locating unit 142 is used to locate and obtain the text content of the image content text of each standardized image to be recognized that is the same as the keyword in the preset second keyword list, and use the corresponding value after the text content as each standardized image to be recognized The corresponding identification value.
  • the keyword "price and tax total” set in the second keyword list is located in each image content text.
  • the values after the keyword (such as 300, 14) are respectively obtained, and the corresponding value after the text content is used as the recognition value corresponding to each standardized image to be recognized.
  • the total data table obtaining unit 150 is configured to obtain the number of pictures of each picture type in the standardized picture set to be recognized to obtain the total number of pictures, and create a sub-data table corresponding to the number of rows according to the number of pictures of each picture type to form a total data sheet.
  • the sub-data table corresponding to each picture type is created to correspondingly store the recognition value of the standardized picture to be recognized of that type, so as to facilitate subsequent summation and use. For example, there are 10 standardized pictures to be recognized for the first picture type, and 10 recognition values are obtained after recognition. Then the above 10 recognition values are stored in the first sub-data table corresponding to the first picture type; the same way is obtained
  • the second sub-data table corresponding to the second picture type and the third sub-data table corresponding to the third picture type are composed of the first sub-data table, the second sub-data table, and the third sub-data table to form a total data table.
  • the summation unit 160 is used to fill the identification values corresponding to each standardized picture to be identified into the corresponding sub-data tables for storage, respectively sum the identification values of the sub-data tables, and then accumulate and sum them to obtain the sum total data table The corresponding actual total value.
  • the identification values of each sub-data table are respectively summed and then accumulated and summed to obtain the sum of the identification values corresponding to each standardized image to be identified in the standardized image to be identified, that is, to obtain all uploaded
  • the total invoice value of the scanned invoice file is recorded as the actual total value.
  • the sum value sending unit 170 is configured to send the actual sum value to the uploader.
  • the actual total value can be sent to the uploader to notify the server that the automatic verification of the invoice amount has been completed, and the user can Proceed to the next step.
  • the image data recognition apparatus 100 based on OCR further includes:
  • the target value acquisition unit uses the target value uploaded by the receiving uploader
  • a numerical value judging unit for judging whether the actual total value is less than the target value
  • the first notification unit is configured to send the first notification information used to notify the approval of the approval to the uploader if the actual total value is greater than or equal to the target value;
  • the second notification unit is configured to send second notification information for notifying that the review has not passed to the uploader if the actual total value is less than the target value.
  • the uploader after the uploader receives the actual total value, it can also choose to set the expected amount of expected reimbursement (understood as a target value).
  • This target value is directly uploaded to the server and calculated before.
  • the actual total value is compared. If the actual total value is greater than or equal to the target value, it means that the expected amount of expected reimbursement is less than or equal to the actual total value, and the reimbursement process can pass the review and continue. If the actual total value is less than the target value, it means that the expected amount of reimbursement is greater than the actual total value, and it cannot be reviewed and the user is prompted to continue uploading another set of pictures to be identified or reduce the target value until it is less than or equal to the total value. After the actual total value is stated, the reimbursement process can be continued.
  • the device realizes that after the non-forward pictures are rotated to obtain the standard forward picture, the invoice amount is recognized and the calculation is performed through the image recognition technology, and the calculation efficiency is improved, and the calculation accuracy rate is high.
  • the above-mentioned OCR-based image data recognition apparatus can be implemented in the form of a computer program, and the computer program can be run on a computer device as shown in FIG. 9.
  • FIG. 9 is a schematic block diagram of a computer device according to an embodiment of the present application.
  • the computer device 500 is a server, and the server may be an independent server or a server cluster composed of multiple servers.
  • the computer device 500 includes a processor 502, a memory, and a network interface 505 connected through a system bus 501, where the memory may include a non-volatile storage medium 503 and an internal memory 504.
  • the non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032.
  • the processor 502 can execute the OCR-based image data recognition method.
  • the processor 502 is used to provide computing and control capabilities, and support the operation of the entire computer device 500.
  • the internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503.
  • the processor 502 can execute the OCR-based image data recognition method.
  • the network interface 505 is used for network communication, such as providing data information transmission.
  • the structure shown in FIG. 9 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied.
  • the specific computer device 500 may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.
  • the processor 502 is configured to run a computer program 5032 stored in a memory to implement the OCR-based image data recognition method disclosed in the embodiment of the present application.
  • the embodiment of the computer device shown in FIG. 9 does not constitute a limitation on the specific configuration of the computer device.
  • the computer device may include more or less components than those shown in the figure. Or some parts are combined, or different parts are arranged.
  • the computer device may only include a memory and a processor. In such embodiments, the structures and functions of the memory and the processor are consistent with the embodiment shown in FIG. 9 and will not be repeated here.
  • the processor 502 may be a central processing unit (Central Processing Unit, CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (Digital Signal Processors, DSPs), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor may be a microprocessor or the processor may also be any conventional processor.
  • a computer-readable storage medium may be non-volatile or may be a volatile computer-readable storage medium.
  • the computer-readable storage medium stores a computer program, where the computer program is executed by a processor to implement the OCR-based image data recognition method disclosed in the embodiments of the present application.
  • the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
  • the above-mentioned integrated unit can be implemented in the form of hardware or software functional unit.
  • the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium.
  • the technical solution of this application is essentially or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium. It includes several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), magnetic disk or optical disk and other media that can store program codes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Development Economics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Finance (AREA)
  • Artificial Intelligence (AREA)
  • Economics (AREA)
  • Accounting & Taxation (AREA)
  • Character Input (AREA)
  • Character Discrimination (AREA)

Abstract

本申请公开了基于OCR的图片数据识别方法、装置、计算机设备及存储介质。该方法包括将待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新得到标准化待识别图片集；获取各标准化待识别图片分别对应的图片类型；通过图像识别获取各标准化待识别图片中指定区域分别对应的识别数值；将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值。该方法实现了将非正向图片均进行旋转得到标准正向图片后，通过图像识别技术识别发票金额和进行核算，提高了核算效率，而且计算准确率高。

Description

基于OCR的图片数据识别方法、装置、及计算机设备
本申请要求于2019年9月11日提交中国专利局、申请号为201910858699.8，发明名称为“基于OCR的图片数据识别方法、装置、及计算机设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及图像识别技术领域，尤其涉及一种基于OCR的图片数据识别方法、装置、计算机设备及存储介质。
背景技术
在财务报销的时候，报销人员需要填写报销表格和粘贴发票，然后由财务人员审核、计算发票的金额和表格中金额是否一致，发票金额必须>＝表格中所报销的金额才能进行后续报销流程。
目前，已出现了在线的办公协同系统，其中也存在在线报销的功能模块。用户在系统上报销时需填写报销信息，并上传报销所提供发票的扫描文件。但是财务人员在使用在线报销的功能模块时，发明人意识到报销人员所填写的报销信息，以及扫描文件均只是保存于在线的办公协同系统的服务器中以供用户查询历史数据，并未利用其中的信息进行金额的自动核算，仍需人工根据报销表格和所粘贴发票核算，而人工核算的过程比较繁琐，这就导致核算效率低下，而且易出错。
发明内容
本申请实施例提供了一种基于OCR的图片数据识别方法、装置、计算机设备及存储介质，旨在解决现有技术中在线的办公协同系统的线报销的功能模块中，报销人员所填写的报销信息，以及扫描文件均只是保存以供用户查询历史数据，仍需人工根据报销表格和所粘贴发票核算，而人工核算的过程比较繁琐，导致核算效率低下，而且易出错的问题。
第一方面，本申请实施例提供了一种基于OCR的图片数据识别方法，其包括：接收上传端所上传的待识别图片集；将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型；通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总 数据表对应的实际总和值；以及将所述实际总和值发送至上传端。
第二方面，本申请实施例提供了一种基于OCR的图片数据识别装置，其包括：
图片集接收单元，用于接收上传端所上传的待识别图片集；
图片标准化单元，用于将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；
图片类型获取单元，用于获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型；
识别数值获取单元，用于通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；
总数据表获取单元，用于获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；
求和单元，用于将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值；以及
和值发送单元，用于将所述实际总和值发送至上传端。
第三方面，本申请实施例又提供了一种计算机设备，其包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现一种基于OCR的图片数据识别方法，其包括：接收上传端所上传的待识别图片集；将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型；通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值；以及将所述实际总和值发送至上传端。
第四方面，本申请实施例还提供了一种计算机可读存储介质，其中所述计算机可读存储介质存储有计算机程序，所述计算机程序当被处理器执行时使所述处理器执行一种基于OCR的图片数据识别方法，其包括：接收上传端所上传的待识别图片集；将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发 票的第二图片类型，对应于定额发票的第三图片类型；通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值；以及将所述实际总和值发送至上传端。
本申请实施例提供了一种基于OCR的图片数据识别方法、装置、计算机设备及存储介质。该方法实现了将非正向图片均进行旋转得到标准正向图片后，通过图像识别技术识别发票金额和进行核算，提高了核算效率，而且计算准确率高。
附图说明
图1为本申请实施例提供的基于OCR的图片数据识别方法的应用场景示意图；
图2为本申请实施例提供的基于OCR的图片数据识别方法的流程示意图；
图3为本申请实施例提供的基于OCR的图片数据识别方法的另一流程示意图；
图4a为本申请实施例提供的基于OCR的图片数据识别方法中非正向图片的示意图；
图4b为本申请实施例提供的基于OCR的图片数据识别方法中标准正向图片的示意图；
图5为本申请实施例提供的基于OCR的图片数据识别方法的子流程示意图；
图6为本申请实施例提供的基于OCR的图片数据识别装置的示意性框图；
图7为本申请实施例提供的基于OCR的图片数据识别装置的另一示意性框图；
图8为本申请实施例提供的基于OCR的图片数据识别装置的子单元示意性框图；
图9为本申请实施例提供的计算机设备的示意性框图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
应当理解，当在本说明书和所附权利要求书中使用时，术语“包括”和“包含”指示所描述特征、整体、步骤、操作、元素和/或组件的存在，但并不排除一个或多个其它特征、整体、步骤、操作、元素、组件和/或其集合的存在或添加。
还应当理解，在此本申请说明书中所使用的术语仅仅是出于描述特定实施例的目的而并不意在限制本申请。如在本申请说明书和所附权利要求书中所使 用的那样，除非上下文清楚地指明其它情况，否则单数形式的“一”、“一个”及“该”意在包括复数形式。
还应当进一步理解，在本申请说明书和所附权利要求书中使用的术语“和/或”是指相关联列出的项中的一个或多个的任何组合以及所有可能组合，并且包括这些组合。
请参阅图1和图2，图1为本申请实施例提供的基于OCR的图片数据识别方法的应用场景示意图；图2为本申请实施例提供的基于OCR的图片数据识别方法的流程示意图，该基于OCR的图片数据识别方法应用于服务器中，该方法通过安装于服务器中的应用软件进行执行。
如图2所示，该方法包括步骤S110～S170。
S110、接收上传端所上传的待识别图片集。
在本实施例中，具体的应用场景为财务报销时，需操作上传端(如智能手机、平板电脑等)在线上直接上传发票扫描文件或照片至服务器，之后还可以选填所需报销的期望金额。由服务器根据所上传的发票扫描文件或照片进行报销金额的计算，无需用户人工核算。
S120、将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集。
在本实施例中，由于所述待识别图片集中可能存在扫描方向不是正方向的图片，此时需服务器统一对所述待识别图片集中非正向图片均进行旋转得到标准正向图片，从而实现所有待识别图片的标准正向化。
在一实施例中，如图3所示，步骤S120之前还包括：
S1201、判断所述待识别图片集中是否存在非正向图片；若所述待识别图片集中存在非正向图片，执行步骤S1202；若所述待识别图片集中不存在非正向图片，执行步骤S130；
S1202、根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度。
在本实施例中，上传端上传了所述待识别图片集时，可能其中包括的发票扫描文件的扫描方向并不是正方向(发票扫描文件的正方向是指票头各文字的中心点连线得到的方向线与扫描文件的页面的底边的夹角为0，也即两者是平行的，而且票头位于扫描文件的最上方)，此时需要对其中非正向图片对应进行旋转，例如如图4a所示的非正向图片。
此时，可根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度。例如如图4a所示的非正向图片中识别得到的首行文字为“第X联：XX联”；上述这些文字是在非正向图片上侧的中部。而参考如图4b中的标准正向图片中“第X联：XX联”这些相同文字对应位置是在标准正向图片中右侧中部。
在一实施例中，步骤S1201中包括：
通过图像识别获取待识别图片集中各待识别图片的首行文字，若有待识别图片的首行文字不包括预先设置的第一关键词列表中的关键词，将对应的待识别图片作为非正向图片。
在本实施例中，通过OCR图像识别模型先识别待识别图片集中各待识别图片的首行文字，是利用OCR技术的逐行从左至右的扫描原理。
OCR技术是光学字符识别的缩写(Optical Character Recognition)，是通过扫描等光学输入方式将各种票据、报刊、书籍、文稿及其它印刷品的文字转化为图像信息，再利用文字识别技术将图像信息转化为可以使用的计算机输入技术。可应用于银行票据、大量文字资料、档案卷宗、文案的录入和处理领域。适合于银行、税务等行业大量票据表格的自动扫描识别及长期存储。
若首行文字中不包括预先设置的第一关键词列表中的关键词(如先设置的第一关键词列表中包括专用发票、普通发票、定额发票等关键词)，表示该待识别图片为非正向图片。
结合图4a和图4b，根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度时，可知该旋转角度为-90度(其中非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取实际角度为逆时针方向90度，若记逆时针方向为正方向，那为了将非正向图片旋转为标准正向图片，需要顺时针旋转90度)，将所述待识别图片集中的非正向图片根据对应的旋转角度进行图片旋转，得到标准化待识别图片集。
S130、获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型。
在本实施例中，请参考图4b，例如出租车的车载端所打出来的发票就是机打发票，一般纳税人给个人或其他一般纳税人开具的发票为增值税专用发票或增值税普通发票，停车票为定额发票。发票内容一般包括：票头、字轨号码、联次及用途、客户名称、银行开户账号、商(产)品名称或经营项目、计量单位、数量、单价、金额，以及大小写金额、经手人、单位印章、开票日期等。实行增值税的单位所使用的增值税专用发票还应有税种、税率、税额等内容。在对所述标准化待识别图片集中各标准化待识别图片进行图片类型的识别时，即可根据票头实现准确识别。
在一实施例中，步骤S130包括：
通过OCR图像识别模型识别各标准化待识别图片的票头，以得到各标准化待识别图片分别对应的图片类型。
在本实施例中，获取所述待识别图片集中各待识别图片分别对应的图片类型时，可以通过OCR图像识别模型识别各标准化待识别图片的票头，即可获取所述待识别图片集中各待识别图片分别对应的图片类型。例如某一标准化待识别图片票头为XXX增值税专用发票，表示该标准化待识别图片的图片类型为第一图片类型。
S140、通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值。
在本实施例中，通过OCR图像识别模型识别各标准化待识别图片的图片内容文本后，其中包括有合计这一关键词，或者价税合计这一关键词。在定位了 各标准化待识别图片分别对应的图片内容文本中合计或价税合计的关键词位置后，即可获知合计或价税合计的关键词之后的识别数值(例如图4b中的价税合计栏之后所示的这一数值)。通过对标准化待识别图片中预设的指定区域的文本识别，即可准确获取对应的识别数值。
在一实施例中，如图5所示，步骤S140包括：
S141、获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片内容文本；
S142、定位获取各标准化待识别图片的图片内容文本中与预设的第二关键词列表中关键词相同的文本内容，以文本内容之后对应的数值作为各标准化待识别图片对应的识别数值。
在本实施例中，先通过OCR图像识别模型识别各标准化待识别图片的图片内容文本后，在各图片内容文本中分别定位“价税合计”这一设置于第二关键词列表中的关键词，在定位到“价税合计”这一关键词之后，分别获取该关键词之后的数值(如300、14)，以文本内容之后对应的数值作为各标准化待识别图片对应的识别数值。通过这一图像识别的方式，能有效且高效识别各标准化待识别图片对应的识别数值。
S150、获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表。
在本实施例中，创建与每一图片类型对应的子数据表是为了对应存储该类型的标准化待识别图片的识别数值，便于后续求和使用。例如第一图片类型的标准化待识别图片有10张，分别识别后得到10个识别数值，则在与第一图片类型对应的第一子数据表中存储上述10个识别数值；同样的方式获取了第二图片类型对应的第二子数据表，及第三图片类型对应的第三子数据表，由第一子数据表、第二子数据表、第三子数据表组成总数据表。
S160、将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值。
在本实施例中，对各子数据表的识别数值分别求和再累加求和，即可获取所述标准化待识别图片集中各标准化待识别图片对应的识别数值的总和，即得到了所有上传的发票扫描文件的发票总额，记为实际总和值。
S170、将所述实际总和值发送至上传端。
在本实施例中，当服务器中根据所上传的待识别图片集完成了发票总金额核算时，可以将所述实际总和值发送至上传端，以通知服务器已完成发票金额的自动核实，用户可进行下一步操作。
在一实施例中，步骤S170之后还包括：
接收上传端所上传的目标数值；
判断所述实际总和值是否小于所述目标数值；
若所述实际总和值大于或等于所述目标数值，将用于通知审核通过的第一通知信息发送至上传端；
若所述实际总和值小于所述目标数值，将用于通知审核未通过的第二通知 信息发送至上传端。
在本实施例中，当上传端接收到了所述实际总和值之后，也可选择设置期望报销的期望金额(理解为目标数值)，这一目标数值是直接上传至服务器后与之前计算得到的所述实际总和值进行比较。若所述实际总和值大于或等于所述目标数值，表示期望报销的期望金额小于或等于实际总和值，是可以通过审核并继续报销流程。若所述实际总和值小于所述目标数值，表示期望报销的期望金额大于实际总和值，是无法审核并提示用户继续上传另一待识别图片集或是减小所述目标数值直至小于或等于所述实际总和值之后，方可继续报销流程。
该方法实现了将非正向图片均进行旋转得到标准正向图片后，通过图像识别技术识别发票金额和进行核算，提高了核算效率，而且计算准确率高。
本申请实施例还提供一种基于OCR的图片数据识别装置，该基于OCR的图片数据识别装置用于执行前述基于OCR的图片数据识别方法的任一实施例。具体地，请参阅图6，图6是本申请实施例提供的基于OCR的图片数据识别装置的示意性框图。该基于OCR的图片数据识别装置100可以配置于服务器中。
如图6所示，基于OCR的图片数据识别装置100包括图片集接收单元110、图片标准化单元120、图片类型获取单元130、识别数值获取单元140、总数据表获取单元150、求和单元160、和值发送单元170。
图片集接收单元110，用于接收上传端所上传的待识别图片集。
在本实施例中，具体的应用场景为财务报销时，需操作上传端(如智能手机、平板电脑等)在线上直接上传发票扫描文件或照片至服务器，之后还可以选填所需报销的期望金额。由服务器根据所上传的发票扫描文件或照片进行报销金额的计算，无需用户人工核算。
图片标准化单元120，用于将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集。
在本实施例中，由于所述待识别图片集中可能存在扫描方向不是正方向的图片，此时需服务器统一对所述待识别图片集中非正向图片均进行旋转得到标准正向图片，从而实现所有待识别图片的标准正向化。
在一实施例中，如图7所示，基于OCR的图片数据识别装置100还包括：
非正向图片判断单元1201，用于判断所述待识别图片集中是否存在非正向图片；若所述待识别图片集中存在非正向图片，执行根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度的步骤；若所述待识别图片集中不存在非正向图片，执行获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型的步骤；
旋转角度获取单元1202，用于根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度。
在本实施例中，上传端上传了所述待识别图片集时，可能其中包括的发票扫描文件的扫描方向并不是正方向(发票扫描文件的正方向是指票头各文字的中心点连线得到的方向线与扫描文件的页面的底边的夹角为0，也即两者是平行的，而且票头位于扫描文件的最上方)，此时需要对其中非正向图片对应进行旋转，例如如图4a所示的非正向图片。
此时，可根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度。例如如图4a所示的非正向图片中识别得到的首行文字为“第X联：XX联”；上述这些文字是在非正向图片上侧的中部。而参考如图4b中的标准正向图片中“第X联：XX联”这些相同文字对应位置是在标准正向图片中右侧中部。
在一实施例中，非正向图片判断单元1201还用于：
通过图像识别获取待识别图片集中各待识别图片的首行文字，若有待识别图片的首行文字不包括预先设置的第一关键词列表中的关键词，将对应的待识别图片作为非正向图片。
在本实施例中，通过OCR图像识别模型先识别待识别图片集中各待识别图片的首行文字，是利用OCR技术的逐行从左至右的扫描原理。
OCR技术是光学字符识别的缩写(Optical Character Recognition)，是通过扫描等光学输入方式将各种票据、报刊、书籍、文稿及其它印刷品的文字转化为图像信息，再利用文字识别技术将图像信息转化为可以使用的计算机输入技术。可应用于银行票据、大量文字资料、档案卷宗、文案的录入和处理领域。适合于银行、税务等行业大量票据表格的自动扫描识别及长期存储。
若首行文字中不包括预先设置的第一关键词列表中的关键词(如先设置的第一关键词列表中包括专用发票、普通发票、定额发票等关键词)，表示该待识别图片为非正向图片。
结合图4a和图4b，根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度时，可知该旋转角度为-90度(其中非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取实际角度为逆时针方向90度，若记逆时针方向为正方向，那为了将非正向图片旋转为标准正向图片，需要顺时针旋转90度)，将所述待识别图片集中的非正向图片根据对应的旋转角度进行图片旋转，得到标准化待识别图片集。
图片类型获取单元130，用于获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型。
在本实施例中，请参考图4b，例如出租车的车载端所打出来的发票就是机打发票，一般纳税人给个人或其他一般纳税人开具的发票为增值税专用发票或增值税普通发票，停车票为定额发票。发票内容一般包括：票头、字轨号码、联次及用途、客户名称、银行开户账号、商(产)品名称或经营项目、计量单位、数量、单价、金额，以及大小写金额、经手人、单位印章、开票日期等。实行增值税的单位所使用的增值税专用发票还应有税种、税率、税额等内容。在对所述标准化待识别图片集中各标准化待识别图片进行图片类型的识别时，即可根据票头实现准确识别。
在一实施例中，图片类型获取单元130还用于：
通过OCR图像识别模型识别各标准化待识别图片的票头，以得到各标准化待识别图片分别对应的图片类型。
在本实施例中，获取所述待识别图片集中各待识别图片分别对应的图片类型时，可以通过OCR图像识别模型识别各标准化待识别图片的票头，即可获取所述待识别图片集中各待识别图片分别对应的图片类型。例如某一标准化待识别图片票头为XXX增值税专用发票，表示该标准化待识别图片的图片类型为第一图片类型。
识别数值获取单元140，用于通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值。
在本实施例中，通过OCR图像识别模型识别各标准化待识别图片的图片内容文本后，其中包括有合计这一关键词，或者价税合计这一关键词。在定位了各标准化待识别图片分别对应的图片内容文本中合计或价税合计的关键词位置后，即可获知合计或价税合计的关键词之后的识别数值(例如图4b中的价税合计栏之后所示的这一数值)。通过对标准化待识别图片中预设的指定区域的文本识别，即可准确获取对应的识别数值。
在一实施例中，如图8所示，识别数值获取单元140包括：
图片内容文本获取单元141，用于获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片内容文本；
关键词定位单元142，用于定位获取各标准化待识别图片的图片内容文本中与预设的第二关键词列表中关键词相同的文本内容，以文本内容之后对应的数值作为各标准化待识别图片对应的识别数值。
在本实施例中，先通过OCR图像识别模型识别各标准化待识别图片的图片内容文本后，在各图片内容文本中分别定位“价税合计”这一设置于第二关键词列表中的关键词，在定位到“价税合计”这一关键词之后，分别获取该关键词之后的数值(如300、14)，以文本内容之后对应的数值作为各标准化待识别图片对应的识别数值。通过这一图像识别的方式，能有效且高效识别各标准化待识别图片对应的识别数值。
总数据表获取单元150，用于获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表。
在本实施例中，创建与每一图片类型对应的子数据表是为了对应存储该类型的标准化待识别图片的识别数值，便于后续求和使用。例如第一图片类型的标准化待识别图片有10张，分别识别后得到10个识别数值，则在与第一图片类型对应的第一子数据表中存储上述10个识别数值；同样的方式获取了第二图片类型对应的第二子数据表，及第三图片类型对应的第三子数据表，由第一子数据表、第二子数据表、第三子数据表组成总数据表。
求和单元160，用于将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值。
在本实施例中，对各子数据表的识别数值分别求和再累加求和，即可获取所述标准化待识别图片集中各标准化待识别图片对应的识别数值的总和，即得到了所有上传的发票扫描文件的发票总额，记为实际总和值。
和值发送单元170，用于将所述实际总和值发送至上传端。
在本实施例中，当服务器中根据所上传的待识别图片集完成了发票总金额核算时，可以将所述实际总和值发送至上传端，以通知服务器已完成发票金额的自动核实，用户可进行下一步操作。
在一实施例中，基于OCR的图片数据识别装置100还包括：
目标数值获取单元，用接收上传端所上传的目标数值；
数值判断单元，用于判断所述实际总和值是否小于所述目标数值；
第一通知单元，用于若所述实际总和值大于或等于所述目标数值，将用于通知审核通过的第一通知信息发送至上传端；
第二通知单元，用于若所述实际总和值小于所述目标数值，将用于通知审核未通过的第二通知信息发送至上传端。
在本实施例中，当上传端接收到了所述实际总和值之后，也可选择设置期望报销的期望金额(理解为目标数值)，这一目标数值是直接上传至服务器后与之前计算得到的所述实际总和值进行比较。若所述实际总和值大于或等于所述目标数值，表示期望报销的期望金额小于或等于实际总和值，是可以通过审核并继续报销流程。若所述实际总和值小于所述目标数值，表示期望报销的期望金额大于实际总和值，是无法审核并提示用户继续上传另一待识别图片集或是减小所述目标数值直至小于或等于所述实际总和值之后，方可继续报销流程。
该装置实现了将非正向图片均进行旋转得到标准正向图片后，通过图像识别技术识别发票金额和进行核算，提高了核算效率，而且计算准确率高。
上述基于OCR的图片数据识别装置可以实现为计算机程序的形式，该计算机程序可以在如图9所示的计算机设备上运行。
请参阅图9，图9是本申请实施例提供的计算机设备的示意性框图。该计算机设备500是服务器，服务器可以是独立的服务器，也可以是多个服务器组成的服务器集群。
参阅图9，该计算机设备500包括通过系统总线501连接的处理器502、存储器和网络接口505，其中，存储器可以包括非易失性存储介质503和内存储器504。
该非易失性存储介质503可存储操作系统5031和计算机程序5032。该计算机程序5032被执行时，可使得处理器502执行基于OCR的图片数据识别方法。
该处理器502用于提供计算和控制能力，支撑整个计算机设备500的运行。
该内存储器504为非易失性存储介质503中的计算机程序5032的运行提供环境，该计算机程序5032被处理器502执行时，可使得处理器502执行基于OCR的图片数据识别方法。
该网络接口505用于进行网络通信，如提供数据信息的传输等。本领域技术人员可以理解，图9中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备500的限定，具体的计算机设备500可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
其中，所述处理器502用于运行存储在存储器中的计算机程序5032，以实 现本申请实施例公开的基于OCR的图片数据识别方法。
本领域技术人员可以理解，图9中示出的计算机设备的实施例并不构成对计算机设备具体构成的限定，在其他实施例中，计算机设备可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。例如，在一些实施例中，计算机设备可以仅包括存储器及处理器，在这样的实施例中，存储器及处理器的结构及功能与图9所示实施例一致，在此不再赘述。
应当理解，在本申请实施例中，处理器502可以是中央处理单元(Central Processing Unit，CPU)，该处理器502还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。其中，通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
在本申请的另一实施例中提供计算机可读存储介质。该计算机可读存储介质可以为非易失性或者可以为易失性的计算机可读存储介质。该计算机可读存储介质存储有计算机程序，其中计算机程序被处理器执行时实现本申请实施例公开的基于OCR的图片数据识别方法。
在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以是两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分，或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，因此，本申请的保护范围应以权利要求的保护范围为准。

Claims (18)

  1. 一种基于OCR的图片数据识别方法，其中，包括：
    接收上传端所上传的待识别图片集；
    将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；
    获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型；
    通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；
    获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；
    将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值；以及
    将所述实际总和值发送至上传端。
  2. 根据权利要求1所述的基于OCR的图片数据识别方法，其中，所述将所述实际总和值发送至上传端之后，还包括：
    接收上传端所上传的目标数值；
    判断所述实际总和值是否小于所述目标数值；
    若所述实际总和值大于或等于所述目标数值，将用于通知审核通过的第一通知信息发送至上传端；
    若所述实际总和值小于所述目标数值，将用于通知审核未通过的第二通知信息发送至上传端。
  3. 根据权利要求1所述的基于OCR的图片数据识别方法，其中，所述将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集之前，还包括：
    判断所述待识别图片集中是否存在非正向图片；若所述待识别图片集中存在非正向图片，执行根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度的步骤；若所述待识别图片集中不存在非正向图片，执行获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型的步骤；
    根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度。
  4. 根据权利要求3所述的基于OCR的图片数据识别方法，其中，所述判断所述待识别图片集中是否存在非正向图片，包括：
    通过图像识别获取待识别图片集中各待识别图片的首行文字，若有待识别图片的首行文字不包括预先设置的第一关键词列表中的关键词，将对应的待识 别图片作为非正向图片。
  5. 根据权利要求1-4任一项所述的基于OCR的图片数据识别方法，其中，所述通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值，包括：
    获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片内容文本；
    定位获取各标准化待识别图片的图片内容文本中与预设的第二关键词列表中关键词相同的文本内容，以文本内容之后对应的数值作为各标准化待识别图片对应的识别数值。
  6. 根据权利要求1-4任一项所述的基于OCR的图片数据识别方法，其中，所述获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型，包括：
    通过OCR图像识别模型识别各标准化待识别图片的票头，以得到各标准化待识别图片分别对应的图片类型。
  7. 一种基于OCR的图片数据识别装置，其中，包括：
    图片集接收单元，用于接收上传端所上传的待识别图片集；
    图片标准化单元，用于将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；
    图片类型获取单元，用于获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型；
    识别数值获取单元，用于通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；
    总数据表获取单元，用于获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；
    求和单元，用于将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值；以及
    和值发送单元，用于将所述实际总和值发送至上传端。
  8. 根据权利要求7所述的基于OCR的图片数据识别装置，其中，还包括：
    目标数值获取单元，用接收上传端所上传的目标数值；
    数值判断单元，用于判断所述实际总和值是否小于所述目标数值；
    第一通知单元，用于若所述实际总和值大于或等于所述目标数值，将用于通知审核通过的第一通知信息发送至上传端；
    第二通知单元，用于若所述实际总和值小于所述目标数值，将用于通知审核未通过的第二通知信息发送至上传端。
  9. 一种计算机设备，包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，其中，所述处理器执行所述计算机程序时实现 一种基于OCR的图片数据识别方法，其中，包括：
    接收上传端所上传的待识别图片集；
    将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；
    获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型；
    通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；
    获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；
    将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值；以及
    将所述实际总和值发送至上传端。
  10. 根据权利要求9所述的一种计算机设备，其中，所述将所述实际总和值发送至上传端之后，还包括：
    接收上传端所上传的目标数值；
    判断所述实际总和值是否小于所述目标数值；
    若所述实际总和值大于或等于所述目标数值，将用于通知审核通过的第一通知信息发送至上传端；
    若所述实际总和值小于所述目标数值，将用于通知审核未通过的第二通知信息发送至上传端。
  11. 根据权利要求9所述的计算机设备，其中，所述将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集之前，还包括：
    判断所述待识别图片集中是否存在非正向图片；若所述待识别图片集中存在非正向图片，执行根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度的步骤；若所述待识别图片集中不存在非正向图片，执行获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型的步骤；
    根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度。
  12. 根据权利要求11所述的计算机设备，其中，所述判断所述待识别图片集中是否存在非正向图片，包括：
    通过图像识别获取待识别图片集中各待识别图片的首行文字，若有待识别图片的首行文字不包括预先设置的第一关键词列表中的关键词，将对应的待识别图片作为非正向图片。
  13. 根据权利要求9-12任一项所述的计算机设备，其中，所述通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别 对应的识别数值，包括：
    获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片内容文本；
    定位获取各标准化待识别图片的图片内容文本中与预设的第二关键词列表中关键词相同的文本内容，以文本内容之后对应的数值作为各标准化待识别图片对应的识别数值。
  14. 根据权利要求9-12任一项所述的计算机设备，其中，所述获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型，包括：
    通过OCR图像识别模型识别各标准化待识别图片的票头，以得到各标准化待识别图片分别对应的图片类型。15、一种计算机可读存储介质，其中，所述计算机可读存储介质存储有计算机程序，所述计算机程序当被处理器执行时使所述处理器执行一种基于OCR的图片数据识别方法，其中，包括：
    接收上传端所上传的待识别图片集；
    将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集；
    获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型；其中，所述图片类型包括对应于增值税专用发票或增值税普通发票的第一图片类型，对应于机打发票的第二图片类型，对应于定额发票的第三图片类型；
    通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值；
    获取所述标准化待识别图片集中各图片类型的图片张数以得到图片总张数，根据各图片类型的图片张数创建对应行数的子数据表以组成总数据表；
    将各标准化待识别图片对应的识别数值分别填充至对应的子数据表中进行存储，对各子数据表的识别数值分别求和再累加求和，得到与总数据表对应的实际总和值；以及
    将所述实际总和值发送至上传端。16、根据权利要求15所述的存储介质，其中，所述将所述实际总和值发送至上传端之后，还包括：
    接收上传端所上传的目标数值；
    判断所述实际总和值是否小于所述目标数值；
    若所述实际总和值大于或等于所述目标数值，将用于通知审核通过的第一通知信息发送至上传端；
    若所述实际总和值小于所述目标数值，将用于通知审核未通过的第二通知信息发送至上传端。
  15. 根据权利要求15所述的存储介质，其中，所述将所述待识别图片集中非正向图片均进行旋转得到标准正向图片，以更新所述待识别图片集得到标准化待识别图片集之前，还包括：
    判断所述待识别图片集中是否存在非正向图片；若所述待识别图片集中存在非正向图片，执行根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度的步骤；若所述待识别图片集中不存在非正向图片，执行获取所述标准化待识别图片集中各标准化待识别图片分别对 应的图片类型的步骤；
    根据非正向图片对应的首行文字的位置与对应的标准正向图片中相同文字对应位置获取旋转角度。
  16. 根据权利要求17所述的存储介质，其中，所述判断所述待识别图片集中是否存在非正向图片，包括：
    通过图像识别获取待识别图片集中各待识别图片的首行文字，若有待识别图片的首行文字不包括预先设置的第一关键词列表中的关键词，将对应的待识别图片作为非正向图片。
  17. 根据权利要求15-18任一项所述的存储介质，其中，所述通过图像识别获取所述标准化待识别图片集中各标准化待识别图片中预设的指定区域分别对应的识别数值，包括：
    获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片内容文本；
    定位获取各标准化待识别图片的图片内容文本中与预设的第二关键词列表中关键词相同的文本内容，以文本内容之后对应的数值作为各标准化待识别图片对应的识别数值。
  18. 根据权利要求15-18任一项存储介质，其中，所述获取所述标准化待识别图片集中各标准化待识别图片分别对应的图片类型，包括：
    通过OCR图像识别模型识别各标准化待识别图片的票头，以得到各标准化待识别图片分别对应的图片类型。
PCT/CN2020/087132 2019-09-11 2020-04-27 基于ocr的图片数据识别方法、装置、及计算机设备 WO2021047182A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910858699.8A CN110781726A (zh) 2019-09-11 2019-09-11 基于ocr的图片数据识别方法、装置、及计算机设备
CN201910858699.8 2019-09-11

Publications (1)

Publication Number Publication Date
WO2021047182A1 true WO2021047182A1 (zh) 2021-03-18

Family

ID=69383492

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/087132 WO2021047182A1 (zh) 2019-09-11 2020-04-27 基于ocr的图片数据识别方法、装置、及计算机设备

Country Status (2)

Country Link
CN (1) CN110781726A (zh)
WO (1) WO2021047182A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113936285A (zh) * 2021-11-03 2022-01-14 重庆海创云链数字科技有限公司 一种ocr自动识别方法
CN114140092A (zh) * 2021-12-06 2022-03-04 烟台文凯园林工程有限公司 一种多功能企业财务核算系统
CN114495085A (zh) * 2022-01-27 2022-05-13 北京京能信息技术有限公司 一种多平台发票在线识别管理的报销方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110781726A (zh) * 2019-09-11 2020-02-11 深圳壹账通智能科技有限公司 基于ocr的图片数据识别方法、装置、及计算机设备
CN112257613B (zh) * 2020-10-23 2023-08-08 中国平安人寿保险股份有限公司 体检报告信息结构化提取方法、装置及计算机设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030016870A1 (en) * 2001-07-19 2003-01-23 Fujitsu Limited Pattern identification apparatus, pattern identification method, and pattern identification program
CN108133212A (zh) * 2018-01-05 2018-06-08 东华大学 一种基于深度学习的定额发票金额识别系统
CN108549843A (zh) * 2018-03-22 2018-09-18 南京邮电大学 一种基于图像处理的增值税发票识别方法
CN109446995A (zh) * 2018-10-30 2019-03-08 广西科技大学 票据信息的处理方法和装置
CN110781726A (zh) * 2019-09-11 2020-02-11 深圳壹账通智能科技有限公司 基于ocr的图片数据识别方法、装置、及计算机设备

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107256379B (zh) * 2017-05-05 2021-06-22 深圳市科迈爱康科技有限公司 基于图像识别的信息采集方法、移动终端及存储介质
CN107633239B (zh) * 2017-10-18 2020-11-03 中电鸿信信息科技有限公司 基于深度学习和ocr的票据分类及票据字段提取方法
CN108596179A (zh) * 2018-03-28 2018-09-28 南京邮电大学 一种增值税发票金额检验方法
CN108961021B (zh) * 2018-05-18 2022-03-15 大账房网络科技股份有限公司 利用智能移动设备拍照生成凭证的方法及系统
CN108717545B (zh) * 2018-05-18 2020-12-18 北京大账房网络科技股份有限公司 一种基于手机拍照的票据识别方法及系统
CN109472918A (zh) * 2018-10-12 2019-03-15 深圳壹账通智能科技有限公司 发票验证方法、融资审核方法、装置、设备及介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030016870A1 (en) * 2001-07-19 2003-01-23 Fujitsu Limited Pattern identification apparatus, pattern identification method, and pattern identification program
CN108133212A (zh) * 2018-01-05 2018-06-08 东华大学 一种基于深度学习的定额发票金额识别系统
CN108549843A (zh) * 2018-03-22 2018-09-18 南京邮电大学 一种基于图像处理的增值税发票识别方法
CN109446995A (zh) * 2018-10-30 2019-03-08 广西科技大学 票据信息的处理方法和装置
CN110781726A (zh) * 2019-09-11 2020-02-11 深圳壹账通智能科技有限公司 基于ocr的图片数据识别方法、装置、及计算机设备

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113936285A (zh) * 2021-11-03 2022-01-14 重庆海创云链数字科技有限公司 一种ocr自动识别方法
CN114140092A (zh) * 2021-12-06 2022-03-04 烟台文凯园林工程有限公司 一种多功能企业财务核算系统
CN114495085A (zh) * 2022-01-27 2022-05-13 北京京能信息技术有限公司 一种多平台发票在线识别管理的报销方法
CN114495085B (zh) * 2022-01-27 2024-05-28 北京京能信息技术有限公司 一种多平台发票在线识别管理的报销方法

Also Published As

Publication number Publication date
CN110781726A (zh) 2020-02-11

Similar Documents

Publication Publication Date Title
WO2021047182A1 (zh) 基于ocr的图片数据识别方法、装置、及计算机设备
US11354495B2 (en) System and method for reliable extraction and mapping of data to and from customer forms
US12014350B2 (en) Systems and methods for mobile image capture and processing of documents
US10565568B2 (en) System and method for processing transaction records for users
WO2019237565A1 (zh) 贷款业务处理方法、装置、计算机设备及可读存储介质
US8798354B1 (en) Method and system for automatic correlation of check-based payments to customer accounts and/or invoices
US9824270B1 (en) Self-learning receipt optical character recognition engine
US9037962B1 (en) Method and system for tracking and reporting data importation accuracy
WO2019052226A1 (zh) 自助理赔申请处理方法、装置、计算机设备和存储介质
US12026772B2 (en) Systems and methods for managing a loan application
US20150356545A1 (en) Machine Implemented Method of Processing a Transaction Document
US20160042469A1 (en) System and method for financial transaction management
US20180032804A1 (en) Optical character recognition utilizing hashed templates
US9208551B2 (en) Method and system for providing efficient feedback regarding captured optical image quality
US20220292607A1 (en) Distributed ledger system for asset management and corresponding insurance applications
CN117807967A (zh) 一种基于ocr智能填单的财务报账方法、装置及电子设备
CN112116484A (zh) 一种在线理赔方法、装置及设备
TWI460674B (zh) 透過行動裝置進行訂定保險契約、變更保戶保險資料、進行保險理賠及處理保險業務之方法
US20220335554A1 (en) Distributed ledger system for asset management and corresponding financial instrument applications
KR20230172753A (ko) 근로소득원천징수영수증 분석 데이터를 이용한 개인 재무분석정보 서비스 제공방법
WO2022069930A1 (en) Method and system for facilitating generation and submission of a digital document

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20863145

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 25.07.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20863145

Country of ref document: EP

Kind code of ref document: A1