Print definition generation device, print definition generation method, and print definition generation program

Abstract

To make it easy to generate print definition information. [Solution] The input unit 21 inputs the image information of the target form to a trained model 40 that has been given sample information, which is a pair of image information 31 of a sample form and print definition information 32 that is setting information for printing the sample form. The output unit 22 acquires and outputs the print definition information, which is setting information for printing the target form and is generated by the trained model 40 in accordance with the image information of the target form.

Description

Technical Field

[0001] The present disclosure relates to a technique for generating a print definition for printing a form from image information of the form.

Background Art

[0002] Recently, the paperless trend of digitizing paper forms has been promoted. However, in, for example, a salary system, due to the need to prevent information leakage, the paperless trend has not progressed, and many paper forms are still used at present. The paper forms in the salary system include management forms used in operations such as salary slips, and forms of a predetermined format designated by the relevant government agency, such as a tax withholding declaration form.

[0003] For salary slips and the like, the format changes every time the number of details increases, and for forms designated by the government agency, the format changes every time there is a law amendment. Every time the format changes, it is necessary to modify the print definition information of the form. The print definition information is setting information for printing a form, such as the printing position of each item, the font, and the ruled line. Forms designated by the government agency often have image information presented with ruled lines and items intricately intertwined. The items are display items and input items. There are also cases where only image information can be obtained for other forms. In order to create print definition information from the image information, it is necessary to visually check the positions of the ruled lines and items and generate the definition information one by one. For example, Patent Document 1 describes generating form data of a form using form editor software.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Every time the form format changes, it's necessary to go through the time-consuming process of creating or modifying print definition information. This disclosure aims to facilitate the generation of print definition information. [Means for solving the problem]

[0006] The print definition generation device relating to this disclosure is A trained model is given sample information, which is a pair of image information of a sample form and print definition information, which is setting information for printing the sample form. The input unit is for inputting the image information of the target form. An output unit outputs print definition information which is setting information for printing the target form, generated by the trained model in response to the image information of the target form input by the input unit. It is equipped with. [Effects of the Invention]

[0007] In this disclosure, image information of the target form is input into a trained model to obtain print definition information, which is the setting information for printing the target form. This makes it possible to easily generate print definition information without having to visually check the position of lines and items and generate definition information one by one. [Brief explanation of the drawing]

[0008] [Figure 1] Configuration diagram of the print definition generation device 10 according to Embodiment 1. [Figure 2] A flowchart showing the processing flow of the print definition generation device 10 according to Embodiment 1. [Figure 3] A diagram showing an example of print definition information according to Embodiment 1. [Figure 4] A diagram showing an example configuration of the trained model 40 according to Embodiment 1. [Figure 5] An explanatory diagram of the prompts generated by the first model 41 according to Embodiment 1. [Figure 6] Configuration diagram of the print definition generation device 10 according to Embodiment 2. [Figure 7] A flowchart showing the processing flow of the print definition generation device 10 according to Embodiment 2. [Figure 8] A flowchart showing the processing flow of the print definition generation device 10 according to Embodiment 3. [Figure 9] Configuration diagram of the print definition generation device 10 according to Embodiment 4. [Figure 10] A flowchart showing the processing flow of the print definition generation device 10 according to Embodiment 4. [Modes for carrying out the invention]

[0009] Embodiment 1. ***Explanation of the structure*** Referring to Figure 1, the configuration of the print definition generation device 10 according to Embodiment 1 will be described. The print definition generation device 10 is a computer. The print definition generation device 10 comprises hardware including a processor 11, memory 12, storage 13, and a communication interface 14. The processor 11 is connected to the other hardware via signal lines and controls this other hardware.

[0010] Processor 11 is an IC that performs processing. IC stands for Integrated Circuit. Specific examples of processor 11 include CPU, DSP, and GPU. CPU stands for Central Processing Unit. DSP stands for Digital Signal Processor. GPU stands for Graphics Processing Unit.

[0011] Memory 12 is a storage device that temporarily stores data. Specific examples of memory 12 include SRAM and DRAM. SRAM stands for Static Random Access Memory. DRAM stands for Dynamic Random Access Memory.

[0012] Storage 13 is a storage device for storing data. As a specific example, Storage 13 is an SSD. SSD is an abbreviation for Solid State Drive. Also, Storage 13 may be a portable recording medium such as an SD (registered trademark) memory card, CompactFlash (registered trademark), NAND flash, flexible disk, optical disk, compact disk, Blu-ray (registered trademark) disk, or DVD. SD is an abbreviation for Secure Digital. DVD is an abbreviation for Digital Versatile Disk.

[0013] Communication interface 14 is an interface for communicating with external devices. As a specific example, Communication interface 14 is a port for Ethernet (registered trademark), USB, or HDMI (registered trademark). USB is an abbreviation for Universal Serial Bus. HDMI is an abbreviation for High-Definition Multimedia Interface.

[0014] The print definition generation device 10 includes an input unit 21 and an output unit 22 as functional components. The functions of each functional component of the print definition generation device 10 are realized by software. The storage 13 stores a program for realizing the functions of each functional component of the print definition generation device 10. This program is read into the memory 12 by the processor 11 and executed by the processor 11. Thereby, the functions of each functional component of the print definition generation device 10 are realized.

[0015] The storage 13 stores image information 31 and print definition information 32. The image information 31 is an image of a form and is data in a format such as PDF, for example. PDF is an abbreviation for Portable Document Format. The print definition information 32 is setting information for printing the form indicated by the image information 31. The print definition information 32 indicates the printing positions, fonts, ruled lines, etc. of each item constituting the form.

[0016] The print definition generation device 10 is connected to the trained model 40 via a communication interface 14. The trained model 40 is a so-called generative AI. AI stands for Artificial Intelligence. The trained model 112 may be constructed using algorithms such as BERT and GPT, as specific examples. BERT stands for Bidirectional Encoder Representations from Transformers. GPT stands for Generative Pretrained Transformer. The trained model 112 may be constructed by combining multiple algorithms, including these algorithms.

[0017] In Figure 1, the trained model 40 was located outside the print definition generation device 10. However, the trained model 40 may also be included as a functional component of the print definition generation device 10.

[0018] In Figure 1, only one processor 11 was shown. However, there may be multiple processors 11, and multiple processors 11 may work together to execute programs that implement each function.

[0019] ***Explanation of operation*** Referring to Figures 2 to 5, the operation of the print definition generation device 10 according to Embodiment 1 will be explained. The operation procedure of the print definition generation device 10 according to Embodiment 1 corresponds to the print definition generation method according to Embodiment 1. Furthermore, the program that implements the operation of the print definition generation device 10 according to Embodiment 1 corresponds to the print definition generation program according to Embodiment 1.

[0020] Referring to Figure 2, the processing flow of the print definition generation device 10 according to Embodiment 1 will be explained. Here, it is assumed that, as a prerequisite, the image information 31 and print definition information 32 for existing forms are already stored in the storage 13.

[0021] (Step S11: Sample information acquisition process) The input unit 21 acquires sample information, which is a pair of image information 31 of a sample form and print definition information 32. The image information 31 of the sample form is the image information 31 of any existing form stored in the storage 13. The print definition information 32 is the setting information for printing the sample form. The input unit 21 may use multiple existing forms as sample forms. In other words, the input unit 21 may acquire a set of image information 31 and print definition information 32 for multiple existing forms as sample information.

[0022] (Step S12: Image acquisition process) The input unit 21 acquires image information of the target document. For example, the input unit 21 acquires image information of a withholding tax statement, using the withholding tax statement as the target document. For example, if the target form is one whose format is specified by a government agency, the input unit 21 will acquire image information of the form that has been made public by the government agency. Also, if the target form is printed on paper, the input unit 21 will acquire image information generated by scanning it.

[0023] (Step S13: Input Processing) The input unit 21 inputs the sample information obtained in step S11 and the image information of the target form obtained in step S12 to the trained model 40. In other words, the input unit 21 provides the sample information and the image information of the target form as input to the trained model 40. In this case, the input unit 21 inputs sample information to the trained model 40 when it inputs the image information of the target form. However, the input unit 21 may also provide the trained model 40 with sample information in advance.

[0024] Then, the trained model 40 generates print definition information corresponding to the image information input in step S13. In other words, the trained model 40 generates print definition information, which is the setting information for printing the target form. At this time, the trained model 40 generates print definition information that follows the format of the print definition information 32 included in the sample information, following the sample information input in step S13. For example, as shown in Figure 3, print definition information defines the printing position, font, and borders of each item that makes up the form.

[0025] (Step S14: Print definition acquisition process) The output unit 22 acquires print definition information generated by the trained model 40 in response to the image information input in step S13.

[0026] (Step S15: Output processing) The output unit 22 outputs the print definition information acquired in step S14. For example, the output unit 22 outputs print definition information to a user terminal or the like connected via the communication interface 14.

[0027] (Step S16: Recording process) The output unit 22 writes the image information of the target form acquired in step S12 as image information 31, and the print definition information acquired in step S14 as print definition information 32 for the target form to the storage 13.

[0028] As shown in Figure 4, the trained model 40 may include two models: the first model 41 and the second model 42. The first model 41 is a model that takes sample information and image information of the target form as input and generates a prompt that instructs the generation of print definition information, which is setting information for printing the target form. Here, the first model 41 generates a prompt that instructs the generation of print definition information according to the description rules for print definition information included in the sample information. The first model 41 also generates a prompt in text format. For example, it generates a prompt like the one shown in Figure 5. The second model 42 is a model that takes the prompt generated by the first model 41 as input and generates print definition information, which is configuration information for printing the target form.

[0029] ***Effects of Embodiment 1*** As described above, the print definition generation device 10 according to Embodiment 1 inputs image information of the target form into the trained model 40 and obtains print definition information, which is setting information for printing the target form. This makes it possible to easily generate print definition information without having to visually check the position of lines and items and generate definition information one by one. The print definition generation device 10 allows the generated print definition information to be transferred to other systems, enabling each system to easily obtain the necessary print definition information. For example, by transferring print definition information related to withholding tax slips to the payroll management system, the previous method of system administrators visually checking the position of lines and items and generating definition information one by one becomes unnecessary.

[0030] The trained model 40 takes image information of the target form as input. Therefore, the trained model 40 needs to be able to handle image format input. However, some generation AI models only handle text format input. In such cases, as explained with reference to Figure 4, it is possible to configure the trained model 40 to include a first model 41, which is configured using a model that can handle image format input, and a second model 42, which does not need to handle image format input. This makes it possible to configure the main processing part, which generates print definition information, using a model that does not need to handle image format input.

[0031] Embodiment 2. Embodiment 2 differs from Embodiment 1 in that it extracts the differences between the print image obtained from the print definition information and the image information of the target form, and feeds this back to the trained model 40. Embodiment 2 explains these differences, while omitting explanations of the similarities.

[0032] ***Explanation of the structure*** Referring to Figure 6, the configuration of the print definition generation device 10 according to Embodiment 2 will be described. The print definition generation device 10 differs from the print definition generation device 10 shown in Figure 1 in that it includes a difference extraction unit 23 and a repetition determination unit 24 as functional components. The difference extraction unit 23 and the repetition determination unit 24 are implemented by software, just like the other functional components.

[0033] ***Explanation of operation*** Referring to Figure 7, the processing flow of the print definition generation device 10 according to Embodiment 2 will be explained. The processes from step S21 to step S24 are the same as the processes from step S11 to step S14 in Figure 2.

[0034] (Step S25: Difference extraction process) The difference extraction unit 23 extracts the differences between the print image obtained from the print definition information acquired in step S24 and the image information of the target form. The print image obtained from the print definition information is generated by the printing system after providing the print definition information to the form's printing system. The print image represents the state that will be printed when the printing system uses the print definition information to print. Therefore, by extracting the differences between the print image and the image information, it is possible to extract the differences between what will actually be printed and the given image information. The differences here include, for example, the degree of misalignment in the printing position of items or the position of ruled lines. It also includes differences such as incorrect displayed characters, different fonts, or different font sizes.

[0035] (Step S26: Iterative judgment process) The repeat determination unit 24 determines whether the difference extracted in step S25 is smaller than the standard. The repetition determination unit 24 proceeds to step S27 if the difference is greater than or equal to the standard. On the other hand, if the difference is less than the standard, the repetition determination unit 24 proceeds to step S28. The repetition determination unit 24 determines whether the difference is less than the standard by comparing the printed image with the image information.

[0036] (Step S27: Re-entry process) The input unit 21 inputs the differences extracted in step S25 into the trained model 40. At this time, the input unit 21 may also input instructions to the trained model 40 to generate print definition information so that the differences become smaller. As a result, the input unit 21 causes the trained model 40 to regenerate the print definition information, which is the setting information for printing the target form. In addition, the input unit 21 may also input not only the differences, but also the sample information acquired in step S21 and the image information of the target form acquired in step S22 to the trained model 40.

[0037] (Step S28: Output processing) The output unit 22 outputs the print definition information obtained in the last executed step S24.

[0038] (Step S29: Recording process) The output unit 22 writes the image information of the target form acquired in step S22 as image information 31, and the print definition information acquired in the last executed step S24 as print definition information 32 for the target form to the storage 13.

[0039] ***Effects of Embodiment 2*** As described above, the print definition generation device 10 according to Embodiment 2 extracts the difference between the print image obtained from the print definition information and the image information of the target form, and feeds it back to the trained model 40. This makes it possible to generate print definition information that is close to the image information of the target form and can be printed.

[0040] Embodiment 3. Embodiment 3 differs from Embodiments 1 and 2 in that, along with the image information of the target form, changes to the image information of the target form are input to the trained model 40. Embodiment 3 explains this difference, while the same points are omitted from the explanation. Embodiment 3 describes a case in which a modification has been made to Embodiment 1. However, it is also possible to modify Embodiment 2.

[0041] ***Explanation of operation*** Referring to Figure 8, the processing flow of the print definition generation device 10 according to Embodiment 3 will be explained. The processes in steps S31 and S32 are the same as the processes in steps S11 and S12 in Figure 2. The processes in steps S34 to S36 are the same as the processes in steps S14 to S16 in Figure 2.

[0042] (Step S33: Input Processing) The input unit 21 inputs the sample information obtained in step S11, the image information of the target form obtained in step S12, and the changes made to the image information of the target form to the trained model 40. The changes may include adding a grid line to indicate an area for displaying a company seal image or other stamp image at a reference position, deleting information for specified items in the target form, changing the design, or changing the font or font size. For example, when printing a target form with a grid line indicating the area for the company seal added, the input unit 21 inputs the change to indicate the addition of the grid line to the trained model 40.

[0043] The trained model 40 then generates print definition information, which is configuration information for printing a form that has been modified according to the changes indicated in the target form.

[0044] ***Effects of Embodiment 3*** As described above, the print definition generation device 10 according to Embodiment 3 inputs the changes to the image information of the target form into the trained model 40. This makes it possible to easily modify the target form and then print it.

[0045] Embodiment 4. Embodiment 4 differs from Embodiments 1 to 3 in that it removes information from specified items from the image information of the target form before inputting it into the trained model 40. Embodiment 4 explains this difference, while omitting explanations of the same points. Embodiment 4 describes a case in which a modification has been made to Embodiment 1. However, it is also possible to modify Embodiments 2 and 3.

[0046] ***Explanation of the structure*** Referring to Figure 9, the configuration of the print definition generation device 10 according to Embodiment 4 will be described. The print definition generation device 10 differs from the print definition generation device 10 shown in Figure 1 in that it includes a pre-processing unit 25 as a functional component. The pre-processing unit 25, like the other functional components, is implemented by software.

[0047] ***Explanation of operation*** Referring to Figure 10, the processing flow of the print definition generation device 10 according to Embodiment 4 will be explained. The processes in steps S41 and S42 are the same as the processes in steps S11 and S12 in Figure 2. The processes in steps S45 to S47 are the same as the processes in steps S14 to S16 in Figure 2.

[0048] (Step S43: Pre-treatment) The preprocessing unit 25 generates a post-removed image from the image information of the target form by removing information from specified items. For example, the specified items are items where personal information is set. This removes the personal information. Alternatively, specified items may be display characters or other elements that the system administrator has deemed unnecessary.

[0049] (Step S44: Input Processing) The input unit 21 inputs the sample information acquired in step S41 to the trained model 40, and also inputs the deleted image generated in step S13 to the trained model 40 as image information for the target form.

[0050] ***Effects of Embodiment 4*** As described above, the print definition generation device 10 according to Embodiment 4 inputs the information of the specified items from the image information of the target form into the trained model 40. This makes it possible to generate print definition information in which the information of the specified items has been removed. In cases where personal information or other sensitive data is displayed on the target form, it may be desirable to remove this information before generating print definition information. Since the system administrator can specify which items to delete, it is possible to reuse existing image information to generate print definition information with only minor differences. The print definition generation device 10 according to Embodiment 4 can handle such cases.

[0051] ***Other configurations*** <Example 1> In Embodiments 1 to 4, each functional component was implemented in software. However, in Modification 1, each functional component may be implemented in hardware. The differences between this Modification 1 and Embodiment 1 will be explained below.

[0052] When each functional component is implemented in hardware, the print definition generation device 10 includes electronic circuits instead of the processor 11, memory 12, and storage 13. The electronic circuits are dedicated circuits that implement the functions of each functional component, as well as the functions of the memory 12 and storage 13.

[0053] Electronic circuits can include single circuits, complex circuits, programmed processors, parallel programmed processors, logic ICs, GAs, ASICs, and FPGAs. GA stands for Gate Array. ASIC stands for Application Specific Integrated Circuit. FPGA stands for Field-Programmable Gate Array. Each functional component may be implemented in a single electronic circuit, or it may be implemented by distributing each functional component across multiple electronic circuits.

[0054] <Modification 2> As a second variation, some of the functional components may be implemented in hardware, while others may be implemented in software.

[0055] The processor 11, memory 12, storage 13, and electronic circuitry are collectively referred to as the processing circuit. In other words, the function of each functional component is realized by the processing circuit.

[0056] Furthermore, the term "part" in the above explanation may be replaced with "circuit," "process," "procedure," "processing," or "processing circuit."

[0057] The various aspects of this disclosure are summarized below as an appendix. (Note 1) A trained model is given sample information, which is a pair of image information of a sample form and print definition information, which is setting information for printing the sample form. The input unit is for inputting the image information of the target form. An output unit that acquires and outputs print definition information, which is setting information for printing the target form, generated by the trained model in response to the image information of the target form input by the input unit, and which is print definition information. A print definition generation device equipped with the following features. (Note 2) The aforementioned trained model, A first model generates a prompt that instructs, in text format, to generate print definition information, which is setting information for printing the target form, from the sample information and the image information of the target form, in accordance with the description rules for print definition information included in the sample information. A second model takes the prompt generated by the first model as input and generates print definition information, which is setting information for printing the target form. A print definition generation device as described in Appendix 1, including the following. (Note 3) The print definition generation device further, Difference extraction unit extracts the difference between the print image obtained from the print definition information acquired by the output unit and the image information of the target form. Equipped with, The input unit inputs the differences extracted by the difference extraction unit into the trained model to regenerate the print definition information, which is the setting information for printing the target form. A print definition generation device as described in Appendix 1 or 2. (Note 4) The print definition generation device further, A repeating determination unit determines whether the difference is smaller than the standard, and if the difference is smaller than the standard, it stops the input unit from inputting the difference to the trained model. A print definition generation device as described in Appendix 3, comprising the following features. (Note 5) The input unit inputs the image information of the target document, along with the changes made to the image information of the target document, into the trained model. The output unit outputs print definition information, which is setting information for printing a form that has been modified to reflect the changes indicated by the modification details, generated by the trained model. A print definition generation device as described in any one of the items 1 to 4 in the appendix. (Note 6) The aforementioned changes involve adding grid lines to indicate an area for displaying the seal image at a reference position. The print definition generation device described in Appendix 5. (Note 7) The aforementioned changes indicate the deletion of information for the specified item in the target form. A print definition generation device as described in Appendix 5 or 6. (Note 8) The print definition generation device further, A preprocessing unit generates a post-deletion image from the image information of the target document, by removing the information of the specified items. Equipped with, The input unit inputs the deleted image generated by the preprocessing unit into the trained model as image information of the target form. A print definition generation device as described in any one of the items 1 through 7 of the appendix. (Note 9) The computer inputs the image information of the target form into a trained model that is given sample information, which is a pair of image information of a sample form and print definition information, which is setting information for printing the sample form. A method for generating a print definition, wherein a computer obtains and outputs print definition information, which is setting information for printing the target form, generated by the trained model in correspondence with the image information of the target form. (Note 10) Given a trained model that is provided with sample information, which is a pair of image information of a sample form and print definition information, which is setting information for printing the sample form, the input process involves inputting the image information of the target form. An output process that obtains and outputs print definition information, which is setting information for printing the target form, generated by the trained model in response to the image information of the target form input by the input process, and which is print definition information. A print definition generation program that makes a computer function as a print definition generation device.

[0058] The embodiments and variations of this disclosure have been described above. Some of these embodiments and variations may be implemented in combination. Alternatively, some or all of them may be implemented in part. However, this disclosure is not limited to the embodiments and variations described above, and various modifications are possible as needed. [Explanation of Symbols]

[0059] 10 Print definition generation device, 11 Processor, 12 Memory, 13 Storage, 14 Communication interface, 21 Input unit, 22 Output unit, 23 Difference extraction unit, 24 Repeat determination unit, 25 Preprocessing unit, 31 Image information, 32 Print definition information, 40 Trained model, 41 First model, 42 Second model.

Claims

1. A trained model is given sample information, which is a pair of image information of a sample form and print definition information, which is setting information for printing the sample form. The input unit is for inputting the image information of the target form. An output unit that acquires and outputs print definition information, which is setting information for printing the target form, generated by the trained model in response to the image information of the target form input by the input unit, and which is print definition information. A print definition generation device equipped with the following features.

2. The aforementioned trained model, A first model generates a prompt that instructs, in text format, to generate print definition information, which is setting information for printing the target form, in accordance with the description rules for print definition information included in the sample information, based on the sample information and the image information of the target form. A second model takes the prompt generated by the first model as input and generates print definition information, which is setting information for printing the target form. A print definition generation apparatus according to claim 1, including the following:

3. The print definition generation device further, Difference extraction unit extracts the difference between the print image obtained from the print definition information acquired by the output unit and the image information of the target form. Equipped with, The input unit inputs the differences extracted by the difference extraction unit into the trained model to regenerate the print definition information, which is the setting information for printing the target form. The print definition generation apparatus according to claim 1.

4. The print definition generation device further, A repeating determination unit determines whether the difference is smaller than the standard, and if the difference is smaller than the standard, it stops the input unit from inputting the difference to the trained model. The print definition generation apparatus according to claim 3, comprising:

5. The input unit inputs the image information of the target document, along with the changes made to the image information of the target document, into the trained model. The output unit outputs print definition information, which is setting information for printing a form that has been modified to reflect the changes indicated by the modification details, generated by the trained model. The print definition generation apparatus according to claim 1.

6. The aforementioned changes involve adding grid lines to indicate an area for displaying the seal image at a reference position. The print definition generation apparatus according to claim 5.

7. The aforementioned changes indicate the deletion of information for the specified item in the target form. The print definition generation apparatus according to claim 5.

8. The print definition generation device further, A preprocessing unit generates a post-deletion image from the image information of the target document, by removing the information of the specified items. Equipped with, The input unit inputs the deleted image generated by the preprocessing unit into the trained model as image information of the target form. The print definition generation apparatus according to claim 1.

9. The computer inputs the image information of the target form into a trained model that is given sample information, which is a pair of image information of a sample form and print definition information, which is setting information for printing the sample form. A method for generating a print definition, wherein a computer obtains and outputs print definition information, which is setting information for printing the target form, generated by the trained model in correspondence with the image information of the target form.

10. Given a trained model that is provided with sample information, which is a pair of image information of a sample form and print definition information, which is setting information for printing the sample form, the input process involves inputting the image information of the target form. An output process that obtains and outputs print definition information, which is setting information for printing the target form, generated by the trained model in response to the image information of the target form input by the input process, and which is print definition information. A print definition generation program that makes a computer function as a print definition generation device.

Images