Information processing system, information processing method and information processing program

Abstract

To provide an information processing system further accurately generating form information including detailed amounts corresponding to forms, even when the forms are issued in various formats.SOLUTION: An information processing system 1 includes a column classification unit P2, a row acquisition unit P3 and a form information generation unit P5. The column classifying unit P2 classifies, into amount columns Cm, columns that satisfy column classification conditions for classifying the amount columns Cm in which amounts are stored, from table data T that corresponds to a table area A of a form image I and in which elements E having a text format are arranged in rows and columns. The row acquisition unit P3 acquires an acquisition object row Tu, the object to be acquired, from a body part B in the table data T on the basis of row acquisition conditions. The form information generation unit P5 generates form information having details D including the amount columns Cm on the basis of the elements E included in the acquisition object row Tu.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an information processing system, an information processing method, and an information processing program for generating form information. [Background technology]

[0002] 2. Description of the Related Art Conventionally, there is an information processing system that reads a form image, which is an image of a form such as an invoice, and generates form information based on the read form image.

[0003] For example, Japanese Patent Publication No. 2022-013524 discloses a document information generation device that uses document information associated with similar past features for the features of a target document image, and character information read from the target document image, and generates document information by reading the "billing date" and "payment deadline" from a new document image using OCR. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2022-013524 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in actual transactions, although various formats of reports are used by different issuers, conventional information processing systems can sometimes have difficulty generating report information that includes accurate amount details, especially for new formats of reports for which no template report information is prepared in advance.

[0006] Therefore, an embodiment of the present invention aims to provide an information processing system, an information processing method, and an information processing program that more accurately generate report information including details of amounts corresponding to the report, even for reports issued in various formats. [Means for solving the problem]

[0007] An information processing system according to an embodiment includes a column classification unit, a row acquisition unit, and a form information generation unit. The column classification unit classifies, from table data corresponding to a table region of a form image and in which elements having text format are arranged in rows and columns, columns that satisfy column classification conditions for classifying an amount column containing amounts into the amount column. The row acquisition unit acquires target rows to be acquired from a body portion of the table data based on the row acquisition conditions. The form information generation unit generates form information having details including the amount column based on the elements included in the target rows.

[0008] In the information processing method according to the embodiment, a column classification unit classifies columns that satisfy column classification conditions for classifying amount columns into amount columns from table data in which elements having text format are arranged in rows and columns, corresponding to the table area of ​​a report image; a row acquisition unit acquires target rows to be acquired from the body portion of the table data based on the row acquisition conditions; and a report information generation unit generates report information having details including the amount columns based on the elements contained in the target rows.

[0009] An information processing program according to an embodiment causes a computer to execute code for a column classification unit, code for a row acquisition unit, and code for a form information generation unit. The code for the column classification unit corresponds to a table region of a form image, and classifies, from table data in which elements having text format are arranged in rows and columns, columns that satisfy column classification conditions for classifying an amount column storing amounts into the amount column. The code for the row acquisition unit acquires target rows to be acquired from a body portion of the table data based on the row acquisition conditions. The code for the form information generation unit generates form information having details including the amount column based on the elements included in the target rows. [Effects of the Invention]

[0010] Embodiments of the present invention can provide an information processing system, an information processing method, and an information processing program that more accurately generate report information including details of amounts corresponding to the report, even for reports issued in various formats. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a block diagram illustrating an example of the overall configuration of an information processing system according to an embodiment. [Figure 2] 1 is a block diagram illustrating an example of an internal configuration of an information processing system according to an embodiment. [Figure 3] 10 is a diagram showing an example of a form image for acquiring table data by a table data acquiring unit in the information processing system according to the embodiment; FIG. [Figure 4] 10A and 10B are examples of table data acquired by a table data acquisition unit in the information processing system according to the embodiment. [Figure 5] 10 is a table illustrating an example of column sorting conditions in the information processing system according to the embodiment. [Figure 6] FIG. 10 is a diagram illustrating an example of a form processing screen displayed by the information processing system according to the embodiment. [Figure 7] FIG. 4 is a partial enlarged view of a form processing screen displayed by the information processing system according to the embodiment. [Figure 8] 10 is a diagram showing an example of a total amount display on a form processing screen displayed by the information processing system according to the embodiment. FIG. [Figure 9] 10A and 10B are diagrams illustrating examples of input of account items and tax classifications on a form processing screen displayed by the information processing system according to the embodiment. [Figure 10] FIG. 10 is a diagram illustrating an example of a form information display screen displayed by the information processing system according to the embodiment. [Figure 11] 10 is a flowchart illustrating an example of a form reading process in the information processing system according to the embodiment. [Figure 12] 10 is a flowchart illustrating an example of a form information generation process in the information processing system according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0012] (Configuration of the embodiment) Hereinafter, an embodiment will be described with reference to the drawings. FIG. 1 is a block diagram showing an example of the overall configuration of an information processing system 1. As shown in FIG.

[0013] 1, the information processing system 1 includes a terminal device 10 and a server 20. The terminal device 10 is connected to the server 20 via a network such as a LAN, the Internet, or mobile communications. The terminal device 10 is used by a user such as an accountant in an organization such as a company.

[0014] FIG. 2 is a block diagram showing an example of the internal configuration of the information processing system 1. As shown in FIG.

[0015] 2, the terminal device 10 is an information terminal such as a personal computer or a smartphone. The terminal device 10 transmits a form image I to the server 20 based on an instruction input by a user. The terminal device 10 has a control unit 11, a storage unit 12, a display unit 13, and an instruction input unit 14.

[0016] The control unit 11 has a processing unit such as a CPU, and executes a program read from the storage unit 12 to realize its functions.

[0017] The storage unit 12 has storage elements such as a ROM, a RAM, an HDD, and an SSD, and stores various information for controlling the terminal device 10 and various programs that constitute the processing unit.

[0018] The display unit 13 is configured with an LCD, an LED, an OLED, or the like, and is capable of displaying various types of information input from the control unit 11.

[0019] The instruction input unit 14 has input devices such as a keyboard and a mouse, and outputs to the control unit 11 various instruction inputs input by the user.

[0020] The server 20 includes a control unit 21 and a storage unit 22 .

[0021] The control unit 21 has a processing unit such as a CPU, and executes a program read from the storage unit 22 to realize its functions.

[0022] The storage unit 22 has storage elements such as a ROM, RAM, HDD, and SSD, and stores various programs that constitute a processing unit for controlling the server 20, as well as various information used by the programs. For example, the storage unit 22 stores programs for a table data acquisition unit P1, a column classification unit P2, a row acquisition unit P3, a table data selection unit P4, a form information generation unit P5, and an accumulation unit P6.

[0023] FIG. 3 is a diagram showing an example of the form image I.

[0024] As shown in Figure 3, the document image I from which table data T is acquired by the table data acquisition unit P1 is, for example, an invoice, and includes one or more table areas A as well as the invoice date, invoice number, registration number, issuer, subject, and invoice amount, etc.

[0025] 4 shows an example of table data T acquired by table data acquisition unit P1. Element E "6" in row 4 of column 1 of table data T1 shown in FIG. 4(A) shows an example in which "c" in form image I has been erroneously converted to text.

[0026] 4, the table data acquisition unit P1 acquires table data T in which elements E in text format are arranged in rows and columns based on a table region A of a form image I. For example, the table data acquisition unit P1 extracts text based on the form image I using a text extraction process, and acquires the table data T by detecting multiple pieces of text arranged in a grid pattern from the position of the extracted text. Alternatively, the table data acquisition unit P1 may acquire the table region A and elements E surrounded by ruled lines extracted from the form image I using a ruled line extraction technique such as a Hough transform, and acquire table data T converted into text for each element E.

[0027] When the form image I has multiple table areas A, the table data acquisition unit P1 also acquires multiple pieces of table data T. For example, in the form image I illustrated in FIG. 3, the table data acquisition unit P1 acquires, as table data T1, from the table area A1: elements E "Summary," "Quantity," "Unit Price," and "Detailed Amount" in the header H of the first row; elements E "a," "1," "100," and "100" in the body B of the second row; and elements E (element values ​​omitted) in the body B of each of the third to fifth rows. Furthermore, the table data acquisition unit P1 also acquires, as table data T2, from the table area A2: "Subtotal" "1,000 yen" in the first row; "Consumption Tax" "100 yen" in the second row; and "Total" "1,100 yen" in the third row. Hereinafter, when referring to both or either of the table data T1 and the table data T2, the term "table data T" will be used.

[0028] FIG. 5 is a table illustrating an example of column classification conditions.

[0029] As shown in Figure 5, the column classification unit P2 classifies columns that satisfy the column classification conditions for classifying the amount column Cm, which stores amounts, from table data T that corresponds to a table area A included in the form image I and in which elements E having text format are arranged in rows and columns, into amount column Cm.

[0030] More specifically, the column classification unit P2 sets the evaluation value of the amount of a column in the header section H of the column included in the table data T, where an element E contains a predetermined amount character indicating that the element is an amount column Cm, to a first predetermined amount evaluation value. For example, if the predetermined amount character is "amount" or "subtotal," the first predetermined amount evaluation value is 1.0. Furthermore, the column classification unit P2 sets the evaluation value of the amount of a column in the body section B of the column, where the ratio of the number of elements E containing numbers to the total number of elements E is equal to or greater than a predetermined amount threshold, to a second predetermined amount evaluation value. For example, the predetermined amount threshold is 0.7, and the second predetermined amount evaluation value is 0.6. The second predetermined amount evaluation value is set to a value lower than the first predetermined amount evaluation value.

[0031] In addition, the column classification unit P2 classifies columns that satisfy the column classification conditions for classifying the unit price column Cu in which unit prices are stored as unit price columns Cu. In the header section H, the column classification unit P2 sets the evaluation value of the unit price of a column in which element E contains a predetermined unit price character indicating that it is a unit price column Cu as the predetermined unit price evaluation value. For example, the predetermined unit price character is "unit price." The predetermined unit price evaluation value is, for example, 1.0.

[0032] Furthermore, the column classification unit P2 classifies columns that satisfy the column classification conditions for classifying the quantity columns Cn in which quantities are stored into the quantity columns Cn. In the header section H, the column classification unit P2 sets the evaluation value of the quantity of a column in which the element E contains a predetermined quantity character indicating that it is a quantity column Cn to a predetermined quantity evaluation value. For example, the predetermined quantity character is "quantity." The predetermined quantity evaluation value is, for example, 1.0.

[0033] The column classification unit P2 also classifies columns that satisfy the column classification conditions for classifying an item column Ci containing items into an item column Ci. The column classification unit P2 sets the evaluation value of an item in a column in the header section H, where an element E contains a first predetermined item character that indicates that the item is an item column Ci, to a first predetermined item evaluation value. For example, the first predetermined item character is "item" and the first predetermined item evaluation value is 1.0. The column classification unit P2 also sets the evaluation value of an item in a column in the header section H, where an element E contains a second predetermined item character, to a second predetermined item evaluation value. For example, the second predetermined item character is "summary" and the second predetermined item evaluation value is 0.8. The column classification unit P2 also sets the evaluation value of an item in a column in the body section B, where the ratio of the number of elements E that do not contain numbers to the total number of elements E is equal to or greater than a predetermined item threshold, to a third predetermined item evaluation value. For example, the predetermined item threshold is 0.7 and the third predetermined item evaluation value is 0.6. The first predetermined item evaluation value is the lowest, followed by the second predetermined item evaluation value and the third predetermined item evaluation value.

[0034] The column classification unit P2 also classifies columns that satisfy the column classification conditions for classifying date columns Cd, which contain dates, into date columns Cd. The column classification unit P2 sets the date evaluation value of a column in the header section H, where an element E contains a predetermined date character indicating that the column is a date column Cd, to a predetermined date evaluation value. The predetermined date character is, for example, "date." The predetermined date value is, for example, 1.0.

[0035] The column classification unit P2 can use regular expressions that specify a forward match or a backward match, etc., to detect whether a column header H contains a predetermined amount character, a predetermined unit price character, a predetermined quantity character, a first predetermined item character, a second predetermined item character, or a predetermined date character. The predetermined amount character, the predetermined unit price character, the predetermined quantity character, the first predetermined item character, the second predetermined item character, and the predetermined date character can be adjusted empirically or experimentally so that the column classification unit P2 can correctly classify columns, in addition to the examples given here.

[0036] The column classification unit P2 classifies, from among the columns included in the table data T, the column with the highest evaluation value for the amount into an amount column Cm, the column with the highest evaluation value for the unit price into a unit price column Cu, the column with the highest evaluation value for the quantity into a quantity column Cn, the column with the highest evaluation value for the item into an item column Ci, and the column with the highest evaluation value for the date into a date column Cd.

[0037] For example, as shown in table data T1 in FIG. 4A, for the amount column Cm, in the fourth column, the ratio of the number of elements E containing numbers (4) to the total number of elements E (4) in the body part B is calculated as 1.0, which is greater than or equal to the predetermined amount threshold (0.7), so the condition for an amount evaluation value of 0.6 is met. However, because the header H contains the predetermined amount character "amount," the overall amount evaluation value is 1.0. In the second and third columns, all of the elements E in the body part B contain numbers, so the amount evaluation value is 0.6. In the first column, the misrecognition "6" in the fourth row causes the ratio of the number of elements E containing numbers (1) to the total number of elements E (4) to be calculated as 0.25, but because this is less than the predetermined amount threshold (0.7), the amount evaluation value is 0. Therefore, the fourth column, which has the highest amount evaluation value among the first through fourth columns included in table data T, is classified as amount column Cm.

[0038] Furthermore, for the unit price column Cu, the third column contains the specified unit price character "unit price" and has a unit price evaluation value of 1.0. On the other hand, the first, second, and fourth columns have unit price evaluation values ​​of 0. Therefore, the third column, which has the largest unit price evaluation value among the columns included in the table data T, is classified as unit price column Cu.

[0039] Furthermore, for the quantity column Cn, the second column contains the predetermined number of characters "quantity" and has a quantity evaluation value of 1.0. On the other hand, the first, third, and fourth columns have quantity evaluation values ​​of 0. Therefore, the second column, which has the largest quantity evaluation value among the columns included in the table data T, is classified as quantity column Cn.

[0040] Furthermore, for item sequence Ci, the first column in body part B satisfies the condition for an item evaluation value of 0.6 because the ratio (0.75) of the number of elements E that do not contain numbers to the total number of elements E (4) due to the misrecognition of "6" in the fourth row is greater than the predetermined item threshold (0.7). However, because it contains a second predetermined item character, the overall item evaluation value is 0.8. On the other hand, columns 2 to 4 have an item evaluation value of 0. Therefore, column 1, which has the largest item evaluation value among the columns included in table data T, is classified as item sequence Ci.

[0041] Furthermore, with regard to the date column Cd, none of the first to fourth columns of the table data T1 contain the predetermined date characters, and the table data T1 does not have the date column Cd.

[0042] As a result, the column classification unit P2 classifies the first column into an item column Ci, the second column into a unit price column Cu, the third column into a quantity column Cn, and the fourth column into an amount column Cm for the table data T1 obtained from the table area A1.

[0043] That is, the column classification conditions include a condition that the header H of a column included in table data T contains a predetermined amount character indicating that the column is an amount column Cm. The column classification conditions also include a condition that the ratio of the number of elements E containing numbers to the total number of elements E in the body B of a column included in table data T is equal to or greater than a predetermined amount threshold. The column classification conditions set the evaluation value of a column in which the header H of the column included in table data T contains a predetermined amount character indicating that the column is an amount column Cm, set the evaluation value of a column in which the ratio of the number of elements E containing numbers to the total number of elements E in the column body B is equal to or greater than the predetermined amount threshold, set a second predetermined amount evaluation value that is smaller than the first predetermined amount evaluation value, and classify the column with the largest evaluation value among the columns included in table data T as an amount column Cm.

[0044] The row acquisition unit P3 acquires the target row Tu to be acquired from the body part B of the table data T based on the row acquisition condition. The row acquisition condition is a condition for excluding rows containing an element E having a predetermined exclusion character from the row to be acquired. The row acquisition unit P3 acquires all elements E of the row to be processed from the body part B of the table data T, and if the acquired element E has a predetermined exclusion character, it excludes that row from the row to be acquired. Examples of predetermined exclusion characters are "subtotal," "total," "consumption tax," "withholding," and "tax amount." In addition to the examples given here, the predetermined exclusion character string can be adjusted empirically or experimentally so that the target row Tu can be correctly acquired.

[0045] Furthermore, the row acquisition unit P3 acquires all elements E of the row to be processed, and when the acquired elements E contain the characters "subtotal," it acquires the number stored in the amount column Cm of the row to be processed as the subtotal and stores it in the memory unit 22. Furthermore, the row acquisition unit P3 acquires all elements E of the row to be processed, and when the acquired elements E contain the characters "tax amount," it acquires the number stored in the amount column Cm of the row to be processed as the tax amount and stores it in the memory unit 22.

[0046] For example, in table data T1 in Fig. 4(A), all of body part B is set as rows Tu to be acquired. In table data T2 in Fig. 4(B), the first row is determined to be a header H and is excluded from the acquisition target, the second row is excluded from the acquisition target because it contains the predetermined exclusion character "consumption tax", and the third row is excluded from the acquisition target because it contains the predetermined exclusion character "total", so that all rows are excluded from the acquisition target.

[0047] Fig. 6 is a diagram showing an example of the form processing screen G1. Fig. 7 is a partial enlarged view of the form processing screen G1. Fig. 8 is a diagram showing an example of a total display on the form processing screen G1. Fig. 9 is a diagram showing an example of input of account items and tax classifications on the form processing screen G1.

[0048] When there are multiple pieces of table data T, the table data selection unit P4 accepts a user's selection input of the table data T to be acquired. More specifically, as shown in FIG. 6, when multiple pieces of table data T are acquired, the table data selection unit P4 displays check boxes on the form processing screen G1, allowing the user to select and input the table data T to be selected. When the user inputs a selection, the table data selection unit P4 includes the selected table data T as the table data to be acquired. For example, in the example of FIG. 6, the table data T1 corresponding to the table area A1 displayed as "Table 1" on the form processing screen G1 is selected by the user and included as the table data to be acquired.

[0049] The form information generation unit P5 generates form information having details D including at least an amount column Cm based on the elements E included in the target row Tu. The form information generation unit P5 displays the target row Tu in the target row display area Gu on the upper side of the form processing screen G1, and displays the detail row Td in the detail row display area Gd on the lower side.

[0050] The user can specify the item in the detail line Td that will set the element E of the classified column. In the example of Figure 6, the upper section of the form processing screen G1 has columns for "item that reflects item," "item that reflects unit price," "item that reflects quantity," and "item that reflects amount," and the "item that reflects item" is specified and input in the "summary" of the detail line Td, and the "item that reflects amount" is specified and input in the "product amount" of the detail line Td.

[0051] 7, the form information generation unit P5 can exclude lines specified by the user from the detail lines Td. When the user checks the check box provided to the left of the target line Tu displayed in the target line display area Gu and clicks the "Do not reflect" button, the form information generation unit P5 excludes the checked line from the detail lines Td displayed in the detail line display area Gd.

[0052] The summing unit P6 sums up the amounts included in the target rows Tu specified and input by the user, and generates a detail row Td for generating form information. In the example of Fig. 7, by checking the checkbox to the left of the table data T displayed on the form processing screen G1 and clicking the "Sum up" button, a detail row Td is generated that sums up the amounts of the target rows Tu that are checked in the form information, as shown in Fig. 8. The generated detail row Td is identified by a summation key.

[0053] The report information generation unit P5 accepts user input of account titles and tax categories for the detail line Td. The account title is information used for journalization, and in the example input of Figure 9, it is "consumable expenses." The tax category is information for classifying transactions related to consumption tax, and in the example input of Figure 9, it is "taxable purchases 10%." When a tax category instruction such as "taxable purchases 10%" is input, the report information generation unit P5 multiplies the amount by the consumption tax rate indicated in the tax category to calculate and display the amount including tax. The report information generation unit P5 also accepts user input for corrections if the amount, etc., has been incorrectly converted to text, for example.

[0054] FIG. 10 is a diagram showing an example of the form information display screen G2.

[0055] As shown in FIG. 10, after the user inputs account titles and tax classifications into the detail line Td and clicks the "Reflect" button, the form information generation unit P5 generates form information including a detail D in which account titles are associated with amounts, and displays the form information display screen G2. In the example of FIG. 10, the detail D includes the amount including tax, account titles, and tax classification. In addition to the detail D, the form information display screen G2 in FIG. 10 also displays information such as the client, amount, transaction date, payment due date, payment date, payment method, transfer account, team (department), person in charge, tag, and case. This information is displayed if it can be extracted by a predetermined text extraction process based on the form image I; if it cannot be extracted, it is left blank. For example, in the case of a client, if the text extracted from the form image I is included in the client list stored in advance in the storage unit 22, it is displayed in the client column.

[0056] When the user inputs an instruction, the form information generation unit P5 outputs the form information including the details D. A business system such as an accounting system can read the form information including the details D and perform subsequent processing.

[0057] (operation) The report reading process will be explained. FIG. 11 is a flowchart showing an example of a form reading process.

[0058] When the user inputs an instruction via the instruction input unit 14 and transmits the form image I to the server 20, the server 20 starts processing in the table data acquisition unit P1.

[0059] The table data acquisition unit P1 acquires table data T, in which elements E are arranged in rows and columns, based on the form image I (S1). The column classification unit P2 classifies the columns included in the table data T into one of an amount column Cm, a unit price column Cu, a quantity column Cn, an item column Ci, or a date column Cd based on the column classification conditions (S2). The row acquisition unit P3 acquires all elements E of the row to be processed, and if any of the elements E contains a predetermined exclusion character, the row is excluded from the acquisition target, and processing proceeds to S5 (S3: NO). If none of the elements E contains the predetermined exclusion character, the row is designated as the acquisition target row Tu, and processing proceeds to S4 (S4).

[0060] In S4, the target row Tu is acquired and a detail row Td is generated.

[0061] If there are rows remaining to be processed, the process returns to S3 (S5: NO). On the other hand, if all rows to be processed have been processed, the process proceeds to S6 (S5: YES).

[0062] If table data T to be processed remains, the process returns to S2 (S6: NO). On the other hand, when processing of all table data T to be processed has been completed, the process ends.

[0063] The processes of S1 to S6 constitute the form reading process.

[0064] Next, the form information generation process will be described. FIG. 12 is a flowchart showing an example of the form information generation process.

[0065] When multiple pieces of table data T are acquired by the table data acquisition unit P1, the table data selection unit P4 displays check boxes on the form processing screen G1 so that the user can select and input the table data T to be processed (S11). When the user operates the check boxes, the table data selection unit P4 adds or deletes detail lines Td depending on the check status of the check boxes (S12: YES, S13). If the user has not operated the check boxes, S13 is skipped and the process proceeds to S14 (S12: NO).

[0066] When the user inputs an instruction to exclude the target row Tu, the form information generation unit P5 excludes the detail row Td corresponding to the target row Tu for which the instruction was input (S14: YES, S15).If the user does not input an instruction to exclude the target row Tu, S15 is skipped and the process proceeds to S16 (S14: NO).

[0067] When the user inputs an instruction to sum up the target rows Tu, the form information generation unit P5 generates and displays a detail row Td that sums up the amounts of the target rows Tu for which the instruction was input, and displays a sum key in the target rows Tu (S16: YES, S17).If the user does not input an instruction to sum up the target rows Tu, S17 is skipped and the process proceeds to S18.

[0068] If the user has not yet finished inputting the account titles and tax classification instructions, the process returns to S11 (S18: NO). On the other hand, if the user has finished inputting the account titles and tax classification instructions and has entered instructions to create form information, the form information generation unit P5 generates form information based on the detail line Td (S18: YES, S19).

[0069] The processes of S11 to S19 constitute the form information generation process.

[0070] As a result, the column classification unit P2 calculates evaluation values ​​for the heading section H and the body section B based on the column classification conditions and performs classification, so that columns can be classified more accurately even if the form image I is in a new format or if characters are misrecognized when extracting text.

[0071] Even if the table data acquisition unit P1 acquires multiple pieces of table data T from the form image I, the row acquisition unit P3 can exclude rows that are not required for generating the details D. Furthermore, the table data selection unit P4 can exclude table data T that the user has determined to be unnecessary.

[0072] The report information generation unit P5 can exclude rows Tu that the user deems unnecessary, and the summation unit P6 can sum up rows Tu that the user wishes to sum up, thereby generating report information that includes more accurate details D.

[0073] According to the embodiment, the information processing system 1 can generate form information including a detailed amount statement D more accurately based on the form even if the form is issued in various formats.

[0074] In the embodiment, the example in which the form image I is an image of an invoice has been described as an example, but the form image I is not limited to this. The form image I may be an image of a form such as a billing statement, an estimate, a delivery note, or an insurance policy, or may be an image of another form.

[0075] In the embodiment, the column classification unit P2 classifies the columns in the table data T into any of the amount column Cm, unit price column Cu, quantity column Cn, item column Ci, and date column Cd, but is not limited to this. The column classification unit P2 may classify the columns in the table data T into other categories.

[0076] Furthermore, in the embodiment, an example has been described in which there are table data T1 and T2, but the number of table data T may be one, or three or more.

[0077] In addition, in the embodiment, an example has been described in which the table data acquisition unit P1 acquires table data T from the form image I by text extraction processing or the like, but when table data T is embedded in the form image I, the embedded table data T may also be acquired directly from the form image I.

[0078] That is, the information processing system 1 has a column classification unit P2 that classifies columns that satisfy column classification conditions for classifying amount columns Cm, which store amounts, from table data T, which corresponds to the table area A of the document image I and in which elements E having text format are arranged in rows and columns, into amount columns Cm; a row acquisition unit P3 that acquires target rows Tu, which are the targets for acquisition, from the body part B of the table data T based on the row acquisition conditions; and a document information generation unit P5 that generates document information having details D including the amount columns Cm based on the elements E contained in the target rows Tu.

[0079] The column classification condition includes a condition that the heading H of a column included in the table data T contains a predetermined amount character indicating that it is an amount column Cm.

[0080] The column classification condition includes a condition that the ratio of the number of elements E containing numbers to the total number of elements E in the body part B of a column included in the table data T is equal to or greater than a predetermined monetary threshold.

[0081] The column classification condition is a condition that sets the evaluation value of a column in the header part H of the column contained in the table data T, in which an element E contains a predetermined amount character indicating that it is an amount column Cm, to a first predetermined amount evaluation value, and sets the evaluation value of a column in the body part B of the column, in which the ratio of the number of elements E containing numbers to the total number of elements E is equal to or greater than a predetermined amount threshold, to a second predetermined amount evaluation value smaller than the first predetermined amount evaluation value, and classifies the column with the largest evaluation value among the columns contained in the table data T as an amount column Cm.

[0082] The line acquisition condition is a condition for excluding lines that include an element E having a predetermined exclusion character from the acquisition target.

[0083] When there are multiple pieces of table data T, the table data selection unit P4 accepts a selection input from the user of the table data T to be acquired.

[0084] The summing unit P6 sums up the amounts included in the acquisition target lines Tu specified and input by the user, and generates a detail line Td for generating form information.

[0085] The form information generating unit P5 generates form information having detail lines Td in which amounts are associated with account items.

[0086] The information processing method involves a column classification unit P2 classifying columns that satisfy column classification conditions for classifying amount columns Cm, which store amounts, from table data T, which corresponds to a table area A of a document image I and in which elements E having text format are arranged in rows and columns, into amount columns Cm; a row acquisition unit P3 acquires a target row Tu to be acquired from a body part B in the table data T based on the row acquisition conditions; and a document information generation unit P5 generates document information having details D including the amount column Cm based on the elements E contained in the target row Tu.

[0087] The information processing program causes a computer to execute the following: code for a column classification unit P2 that classifies columns that satisfy column classification conditions for classifying amount columns Cm, which store amounts, from table data T, which corresponds to table area A of a form image I and in which elements E having text format are arranged in rows and columns, into amount columns Cm; code for a row acquisition unit P3 that acquires target rows Tu, which are to be acquired, from body section B of the table data T based on row acquisition conditions; and code for a form information generation unit P5 that generates form information having details D including amount columns Cm based on elements E contained in the target rows Tu.

[0088] As described above, the embodiments of the present invention have been described, but the embodiments are not limited to these, and various changes, modifications, etc. are possible within the scope that does not change the gist of the present invention.

[0089] Furthermore, the order of execution of each step of each procedure in the embodiments may be changed, multiple steps may be executed simultaneously, or the steps may be executed in a different order each time, as long as this does not contradict the nature of the steps. Furthermore, all or part of the steps of each procedure in the present embodiment may be realized by hardware or by a program. [Explanation of symbols]

[0090] 1. Information Processing System 10. Terminal device 11 Control section 12...Storage section 13...Display section 14. Instruction input section 20 servers 21 Control unit 22...Storage section B Body D...Details Cd...Date column Ci···Item sequence Cm...Amount column Cn...Quantity column Cu... Unit Price Column E element G1... Report processing screen G2 Report information display screen H Heading Gd...Detail line display area Gu... Display area for lines to be acquired I...Form image A, A1, A2...Tablespace P1: Table data acquisition section P2... Column classification section P3...Row acquisition part P4: Table data selection section P5: Report information generation section P6...Total section T, T1, T2...table data Td...detail line Tu... Target row

Claims

1. a column classification unit that classifies, from table data in which elements having a text format are arranged in rows and columns, a price column in which a price is stored, the price column if the column satisfies a column classification condition; and a row acquisition unit that acquires rows to be acquired from a body portion of the table data based on row acquisition conditions; a form information generating unit that generates form information having details including the amount column based on the elements included in the target row; An information processing system having the above.

2. The information processing system according to claim 1 , wherein the column classification condition includes a condition that a header of the column included in the table data contains a predetermined amount character indicating that the column is an amount column.

3. The information processing system according to claim 1 , wherein the column classification condition includes a condition that the ratio of the number of elements including numbers to the total number of elements in the body part of the column included in the table data is equal to or greater than a predetermined monetary threshold.

4. The column sorting condition is: In the heading of the column included in the table data, the evaluation value of the column in which the element contains a predetermined amount character indicating that it is the amount column is set to a first predetermined amount evaluation value; In the body of the column, the evaluation value of the column in which the ratio of the number of elements including numbers to the total number of elements is equal to or greater than a predetermined monetary threshold is set to a second predetermined monetary evaluation value that is smaller than the first predetermined monetary evaluation value; a condition for classifying the column having the largest evaluation value among the columns included in the table data into the amount column, The information processing system according to claim 1 .

5. The information processing system according to claim 1 , wherein the line acquisition condition is a condition for excluding lines including the element having a predetermined exclusion character from the acquisition target.

6. A table data selection unit is provided, the table data selection unit accepts a selection input by a user of the table data to be acquired when there are a plurality of table data; The information processing system according to claim 1 .

7. It has a summing section, the summing unit sums the amounts included in the acquisition target lines designated and input by the user, and generates detail lines for generating the form information; The information processing system according to claim 1 .

8. The information processing system according to claim 1 , wherein the form information generating unit generates the form information having detail lines in which account items are associated with the amounts.

9. a column classification unit classifies, from table data in which elements having a text format are arranged in rows and columns, a column that satisfies a column classification condition for classifying an amount column in which amounts are stored, into the amount column; a row acquisition unit acquires rows to be acquired from a body portion of the table data based on row acquisition conditions; a report information generating unit generating report information having details including the amount column based on the element included in the target row; Information processing methods.

10. a column classification unit code for classifying a column that satisfies a column classification condition for classifying an amount column in which amounts are stored, from table data in which elements having a text format are arranged in rows and columns, corresponding to a table region of the form image; and code for a row acquisition unit that acquires target rows from a body portion of the table data based on row acquisition conditions; code for a form information generation unit that generates form information having details including the amount column based on the element included in the target row; An information processing program that causes a computer to execute the above.

Images