Information processing device

Abstract

To solve the problem in which it was difficult for a user to flexibly set a requirement because conventionally it is required to previously define an inhibited character when attribute information such as a file name is given on the basis of a character string obtained by character recognition processing for a scan image.SOLUTION: Correction content can be learned on the basis of a correction operation performed by a user on an attribute setting screen when attribute information such as a file name is given on the basis of a character string obtained by character recognition processing for a scan image.SELECTED DRAWING: Figure 9

Description

[Technical field]

[0001] The present invention relates to an apparatus, a method, and a program for acquiring a desired character string based on a character recognition result of a scanned image. [Background technology]

[0002] Conventionally, there is a system that sets a file name for a form image based on the recognition result obtained by performing character recognition processing on a form image obtained by scanning a paper form. Patent Document 1 describes that a predetermined area in a form image is subjected to character recognition processing and used as the file name of the form image. Patent Document 2 describes that a file naming rule can be imposed when an extracted area selected by a user is OCR-ed and used as the file name. The file naming rule describes imposing requirements on the length of the file name (minimum length and maximum length), erasing prohibited characters, and preventing the same file name from being reused. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 62-051866 [Patent Document 2] Special Publication No. 2007-503032 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the method in Patent Document 2 requires users to set requirements in advance, such as characters that are prohibited from being used in file names, making it difficult for users to flexibly set requirements using the recognition results of form images. [Means for solving the problem]

[0005] In order to solve the above problems, an information processing device of the present invention has: a character recognition means for performing character recognition processing on a character area in a first scanned image to obtain a character recognition result; and a learning means for learning a correction content when a part of a character string in the character recognition result obtained by the character recognition means is corrected when attribute information for the first scanned image is set using the character recognition result obtained by the character recognition means, when the correction made when setting the attribute information for the first scanned image is a deletion of a part of a character string of the character recognition result, the learning means learns a regular expression based on the deleted character string as the correction content; The character recognition means performs character recognition processing on a character area in the second scanned image. and a character string resulting from character recognition in a character region in the second scanned image matches a regular expression learned by the learning means. case 、 Character recognition results for character regions in the second scanned image. String Against Delete a part of the string that matches the regular expression and The method is characterized in that it outputs the character recognition result after the character recognition. Effect of the Invention

[0006] The present invention can automatically perform corrections on similar scanned images by learning the correction content based on the correction operations performed by the user when assigning attribute information such as a file name based on the character string obtained by character recognition processing on the scanned image. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram showing the overall configuration of the present system. [Diagram 2] 2 is an example of the hardware configuration of an MFP. [Diagram 3] 2 is an example of a hardware configuration of a client PC and an MFP linkage service. [Figure 4] 2 is an example of a software configuration of the present system. [Diagram 5] FIG. 2 is a sequence diagram showing a process flow between the devices. [Figure 6] 13A and 13B are diagrams illustrating examples of screens displayed by the MFP or the client PC. [Figure 7] 10 is a flowchart showing details of a learning process for correcting a character recognition result. [Figure 8] FIG. 13 is a diagram showing an example of an attribute setting screen when a character recognition result is corrected. [Figure 9] 10 is a flowchart showing details of a correction process when performing character recognition processing on a new image. [Figure 10] FIG. 13 is a diagram showing an example of an attribute setting screen. [Figure 11] FIG. 11 is a diagram showing details of data output as a character recognition result. [Figure 12] FIG. 13 is a diagram showing an example of an attribute setting screen. [Figure 13] FIG. 11 is a diagram showing details of data output as a character recognition result. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] <System configuration> 1 is a diagram showing the overall configuration of a system for realizing this embodiment. The image processing system includes an MFP (Multifunction Peripheral) 110, a client PC 111, an MFP collaboration service 120, and a cloud storage 130. The MFP 110 and the client PC 111 are communicably connected to servers (such as the MFP collaboration service 120 and the cloud storage 130) that provide various services on the Internet via a LAN (Local Area Network).

[0009] The MFP 110 is a multifunction device having multiple functions such as a scanner and a printer, and is an example of an image processing device. The client PC 111 is a computer that executes an application program that receives a service requested from the MFP linkage service 120. The MFP linkage service 120 is an example of a service that has a function of saving an image file scanned by the MFP 110 on its own server, or transferring it to a service that allows file saving, such as another storage service. The cloud storage 130 is a service that allows files to be saved via the Internet and files to be obtained using a web browser.

[0010] Note that the image processing system of this embodiment is configured to include the MFP 110, the client PC 111, the MFP cooperation service 120, and the cloud storage 130, but is not limited to this configuration. For example, the MFP 110 may also serve as the client PC 111 or the MFP cooperation service 120. Further, the MFP cooperation service 120 may be arranged in a connection form on a server on the LAN instead of on the Internet. Further, the cloud storage 130 may be replaced with a mail server or the like, and the scanned image may be attached to and sent by mail.

[0011] <Hardware Configuration of MFP> FIG. 2 is a hardware configuration diagram of the MFP 110. The MFP 110 is composed of a control unit 210, an operation unit 220, a printer unit 221, a scanner unit 222, and a modem 223. The control unit 210 is composed of the following units 211 to 219, and controls the operation of the entire MFP 110. The CPU 211 reads out a control program stored in a ROM 212, and executes and controls various functions of the MFP 110, such as reading, printing, and communication. The RAM 213 is used as a temporary storage area such as a main memory and a work area of ​​the CPU 211. In this embodiment, one CPU 211 executes each process shown in a flowchart described later using one memory (RAM 213 or HDD 214), but is not limited to this. For example, each process may be executed by cooperating multiple CPUs and multiple RAMs or HDDs. The HDD 214 is a large-capacity storage unit that stores image data and various programs. The operation unit I / F 215 is an interface that connects the operation unit 220 and the control unit 210. The operation unit 220 is provided with a touch panel, a keyboard, and the like, and accepts operations / inputs / instructions from a user. The printer I / F 216 is an interface that connects the printer unit 221 and the control unit 210. Image data for printing is transferred from the control unit 210 to the printer unit 221 via the printer I / F 216, and is printed on a recording medium. The scanner I / F 217 is an interface that connects the scanner unit 222 and the control unit 210. The scanner unit 222 reads an original placed on an original table or an ADF (Auto Document Feeder), not shown, to generate image data, and inputs the image data to the control unit 210 via the scanner I / F 217. The MFP 110 has a function of printing (copying) the image data generated by the scanner unit 222 from the printer unit 221, a function of transmitting the image data as a file over a network, a function of attaching the image data to an e-mail and transmitting the same, and the like. The modem I / F 218 is an interface that connects the modem 223 and the control unit 210. A modem 223 communicates image data by facsimile with a facsimile device on the PSTN. A network I / F 219 is an interface that connects the control unit 210 (MFP 110) to a LAN.The MFP 110 uses a network I / F 219 to transmit image data and information to various services on the Internet and to receive various types of information.

[0012] <Hardware configuration of client PC and MFP linkage service> 3 is a hardware configuration example of an information processing device capable of realizing the client PC 111 and the MFP cooperation service 120. The device for realizing the client PC 111 and the MFP cooperation service 120 is composed of a CPU 311, a ROM 312, a RAM 313, a HDD 314, and a network I / F 315. The CPU 311 reads out a control program stored in the ROM 312 and executes various processes to control the overall operation. The control program is installed via various networks such as the Internet and an intranet, or a computer-readable storage medium such as a CD-ROM or a USB memory. The RAM 313 is used as a temporary storage area such as the main memory and work area of ​​the CPU 311. The HDD 314 is a large-capacity storage unit that stores image data and various programs. The network I / F 315 is an interface that connects the MFP cooperation service 120 to the Internet. The MFP cooperation service 120 and the cloud storage 130 receive processing requests from other devices (such as the MFP 110) via the network I / F 315 and transmit and receive various information.

[0013] <Image processing system software configuration> 4 shows an example of a software configuration related to programs executed in the MFP 110 and the MFP cooperation service 120 in the image processing system according to this embodiment. The MFP 110 is roughly divided into two parts: a native function unit 410 and an additional function unit 420. Each processing unit included in the native function unit 410 is provided as standard in the MFP 110, whereas the additional function unit 420 is realized by an application additionally installed in the MFP 110. The additional function unit 420 is, for example, an application based on Java (registered trademark), and can easily realize the addition of functions to the MFP 110. Note that other additional applications not shown may be installed in the MFP 110.

[0014] The native function unit 410 includes a scan execution unit 411 and an image data storage unit 412. The additional function unit 420 includes a display control unit 421, a scan instruction unit 422, and a cooperation service request unit 423.

[0015] The display control unit 421 displays a UI screen for receiving operations by a user on a liquid crystal display unit having a touch panel function of the operation unit 220 of the MFP 110. For example, the display control unit 421 displays UI screens for inputting authentication information for accessing the MFP linkage service 120, scan settings, operations for starting scanning, preview screens, etc. The scan instruction unit 422 requests the scan execution unit 411 to perform scanning processing together with scan settings according to user instructions input via the UI screen.

[0016] The scan execution unit 411 receives a scan request including scan settings from the scan instruction unit 422. In accordance with the scan request, the scan execution unit 411 controls the scanner unit 222 via the scanner I / F 217 to read a document placed on the document glass platen, thereby generating scanned image data. The generated scanned image data is sent to the image data storage unit 412. The scan execution unit 411 sends a scanned image identifier that uniquely identifies the stored scanned image data to the scan instruction unit 422. The scanned image identifier is a number, symbol, alphabet, or the like (not shown) for uniquely identifying an image scanned in the MFP 110. The image data storage unit 412 stores the scanned image data received from the scan execution unit 411 in the HDD 214.

[0017] The scan instruction unit 422 further acquires, from the image data storage unit 412, the scan image data corresponding to the scan image identifier received from the scan execution unit 411. Then, the scan instruction unit 422 requests the cooperation service request unit 423 to issue an instruction to cause the MFP cooperation service 120 to perform desired processing on the acquired scan image data.

[0018] The collaboration service request unit 423 requests various processes to the MFP collaboration service 120. For example, the collaboration service request unit 423 requests login, analysis of a scanned image, transmission of a scanned image, etc. Protocols such as REST and SOAP are used for communication with the MFP collaboration service 120, but other communication means may also be used.

[0019] The MFP linkage service 120 includes a request control unit 431 , an image processing unit 432 , a cloud storage access unit 433 , a data management unit 434 , and a display control unit 435 .

[0020] A request control unit 431 waits in a state in which it can receive a request from an external device. When a processing request is received, the request control unit 431 instructs an image processing unit 432, a cloud storage access unit 433, and a data management unit 434 to perform the processing as appropriate in response to the request.

[0021] The image processing unit 432 performs character area analysis processing on the image, character recognition processing on the analyzed character areas (hereinafter referred to as OCR processing), similar document determination processing (described in the processing of S507, S508, and S509 in FIG. 5 below), and image recognition processing and image processing such as image rotation and tilt correction.

[0022] The cloud storage access unit 433 issues processing requests to the cloud storage. Cloud services generally use protocols such as REST and SOAP to publish various interfaces for saving files in the cloud storage and retrieving the saved files. The cloud storage access unit 433 operates the cloud storage using the published cloud storage interface.

[0023] A data management unit 434 holds user information and various setting data managed by the MFP cooperation service 120 .

[0024] The display control unit 435 receives a request from a web browser running on another terminal (not shown), such as a PC or mobile device connected via the Internet, and returns screen configuration information (HTML, CSS, etc.) required for screen display. The user can check user information registered in the MFP linkage service 120 or change settings for scanning via the screen displayed by the web browser.

[0025] 4, an example of a configuration in which the additional function unit 420 is installed in the MFP 110 has been described, but the present invention is not limited to this configuration. For example, a program for realizing the function of the additional function unit 420 may be included in the client PC 111.

[0026] <Overall process flow> Fig. 5 is a sequence diagram showing the flow of processing between each device when an image scanned by the MFP 110 is converted into a file and transmitted to cloud storage. Here, the explanation will be centered on the exchange between each device. Note that, although Fig. 5 shows that the MFP 110 exchanges with the MFP linkage service 120, the acquisition of analysis results (S513), the display of a screen (S514), instruction of learning (S515), and the like, which will be described later, may be performed by the client PC 111 instead of the MFP 110.

[0027] In normal operation, the MFP 110 displays a main screen on the touch panel with buttons for performing various functions. By installing an additional application (hereinafter referred to as a scan application) for sending scanned forms to cloud storage on the MFP 110, buttons for using the application functions are displayed on the main screen of the MFP 110. When this button is pressed by a user, the CPU of the MFP 110 executes the scan application to display a screen for sending scanned forms to cloud storage, and the process shown in the sequence of FIG. 5 is started.

[0028] In S501, the scan application displays a login screen for prompting the user to input authentication information (user name and password) for accessing the MFP collaboration service 120. When the user inputs the authentication information and presses the login button, in S502, a login request is sent to the MFP collaboration service. In S503, the MFP collaboration service 120 verifies whether the authentication information (user name and password) included in the received login request is correct, and if correct, returns an access token to the MFP 110. Subsequent requests made from the MFP 110 to the MFP collaboration service 120 are sent together with this access token, and the user who instructed the request can be identified based on the information in this access token. User authentication is performed using a generally known method (Basic authentication, Digest authentication, authorization using OAuth, etc.).

[0029] When the login process is completed, the MFP 110 displays a scan setting screen in S504. After configuring various scan reading settings, placing the paper form to be scanned on the platen glass or ADF, and pressing the "Start Scan" button, scanning is performed in S505 to generate image data that digitizes the paper form. Then, in S506, a request for analyzing the scanned image is sent to the MFP collaboration service 120 together with the generated scanned image data. When the MFP collaboration service 120 receives the request for analyzing the scanned image, in S507, the MFP collaboration service 120 controls the image processing unit 432 of the MFP collaboration service 120 to start image analysis of the scanned image. At this time, the MFP collaboration service 120 returns to the MFP 110 a "processId" that is an identifier for uniquely identifying the analysis requested to the MFP collaboration service 120, without waiting for the image analysis process to end.

[0030] In the image analysis process executed by the image processing unit 432, first, in S508, an analysis of a character area existing in a scanned image is performed. Then, using the layout information of the character area in the analyzed scanned image, in S509, the layout information of the character area of ​​an image scanned in the past and the layout information of the character area of ​​the current scanned image are compared to determine whether there is an image (similar form) having a similar layout of character areas. This process is called similar form determination. Note that the layout information of the character area in the past scanned image used in this determination is saved and accumulated by the process of S517 described later. Based on the determination result in S509, in S510, OCR processing is performed on the analyzed character area. Note that the details of the OCR processing in S510 will be described with reference to FIG. 7.

[0031] In S511, the MFP 110 uses the “processId” received as a response to the analysis request in S506 to periodically (for example, every several hundred milliseconds to a few milliseconds) transmit a confirmation request for the processing status of the image analysis corresponding to “processId” to the MFP collaboration service 120. Although not shown in the figure, the processing of S511 is performed periodically and continuously until a response indicating completion of image processing from the MFP collaboration service 120 can be obtained (until the timing of S512). When the MFP collaboration service 120 receives a request for confirmation of the processing status from the MFP 110, it checks the processing status of the requested “processId” and returns a response. In the response, a character string indicating the current processing status is stored in “status”. For example, when “status” is “processing”, it indicates that the processing is being performed by the MFP collaboration service 120, and when “completed”, it indicates that the processing is completed. If the processing fails, a status such as “failed” may be returned. In addition, the response upon completion of the process (when the status is completed) includes the status indicating the completion of the process, as well as the results of analyzing the scanned image and information such as the scan settings.

[0032] In S512, if a response including a status indicating completion of the process is received as a result of checking the processing status, in S513, the MFP accesses a URL related to the analysis result information included in the response to acquire information on the analysis result of the scanned image. In S514, the MFP uses the analysis result information acquired in S513 to display an attribute setting screen for setting a file name and metadata for the scanned image. FIG. 6B shows an example of the attribute setting screen. Details of the attribute setting screen will be described later. Note that, in the case of displaying the attribute setting screen on the client PC 111 to set a file name and metadata, the attribute setting screen can be displayed in the following procedure. That is, when the user logs in to the MFP cooperation service from the web browser of the client PC 111, a list of jobs of scanned forms held in the MFP cooperation service (FIG. 6B) is displayed, and a job related to a desired form is selected from the list based on a user's instruction. Then, the web browser of the client PC 111 accesses a URL related to the analysis result information of the scanned form image corresponding to the selected job, acquires the analysis result information, and displays an attribute setting screen for setting a file name and metadata for the form image.

[0033] After setting a file name and metadata for the scanned form on the attribute setting screen, when the user presses the "OK" button 617, in S515 the MFP sends information about the character area used to set the file name and metadata along with a learning request to the MFP linkage service 120. If the user corrects a character string extracted by OCR processing on the file name setting screen using a keyboard or the like, information about the corrected character string (information about the correction content) is also sent.

[0034] Upon receiving the learning request, the request control unit 431 of the MFP linkage service 120 requests the image processing unit 432 to learn the form in S516. In S517, the image processing unit 432 saves the layout information of all character areas included in the scanned image and the information of the character areas used by the user for the file name and metadata. In addition, if the user has corrected the character string extracted by OCR processing on the file name setting screen, in S518, the request control unit 431 requests the image processing unit 432 to learn about the correction of the OCR processing result. In S519, the image processing unit 432 performs learning processing based on the correction content of the OCR processing result and saves the information of the learning result. Details of S519 will be described later.

[0035] Thereafter, in S520, the MFP 110 transmits the scanned image and information such as the file name set on the attribute setting screen to the MFP collaboration service 120. When the MFP collaboration service 120 receives the request, it starts a file generation process based on the received scanned image and returns to the MFP 110 that the request was received normally. When the MFP 110 receives a response to the transmission, it ends the process and returns to the scan setting screen display of S504. In S521, the MFP collaboration service 120 acquires information on the file format when transmitting the scanned image to the cloud storage 130 based on the scan settings registered in the MFP collaboration service, and generates a file from the scanned image based on the format information. Then, in S522, the MFP collaboration service 120 sets the file name received in S520 to the file generated in S521 and transmits the file to the cloud storage 130.

[0036] 6A and 6B are diagrams showing examples of screens displayed by the MFP 110 (or the client PC 111). FIG. 6A is a diagram showing an example of a scanned form list screen 600. The screen 600 displays a list 601 of jobs related to scanned form images that have been temporarily stored in the MFP linkage service 120 after the scanning and image analysis processes (S505 to S510) have been completed and before being sent to the cloud storage 130. In addition to the scanned form list 601, the screen 600 also includes a send button 602, an edit button 603, and a delete button 604.

[0037] The scanned form list 601 is composed of a form name 605, a destination 606, a status 607, and a type 608. The form name 605 is an identifier that uniquely identifies the name of the form. The destination 606 indicates the name of the cloud storage 130 to which the file of the form image is to be sent. The status 607 indicates the result of performing similar form determination on the form image, and either "unlearned" or "learned" is displayed. "Unlearned" means that it has been determined that no similar form exists, and "learned" means that it has been determined that a similar form exists. The type 608 indicates the type of form that has been automatically determined based on the result of OCR of the title character string in the form image, or the type of the similar form when a similar form exists by the similar form determination, and for example, "quote" or "invoice" is displayed. In addition, for forms whose status 607 is learned, a detailed type indicating which invoice format the form corresponds to is displayed, such as "invoice AAA" or "invoice BBB". This detailed type is the type associated with the most similar form determined by the similar form determination process.

[0038] The send button 602 is a button for sending a form to the cloud storage 130. When a job of an arbitrary form image is selected from the scanned form list 601 by a user operation and the send button 602 is pressed, the MFP cooperation service sends the selected form image to the cloud storage 130 set as the destination 606. Note that if the transmission is completed normally, the job related to that form image is deleted from the list.

[0039] The edit button 603 is a button for moving to an attribute setting screen 610, which will be described later. When a user operates to select a job of an arbitrary form image from the scanned form list 601 and presses the edit button 603, the screen moves to the attribute setting screen 610 for setting attributes related to the selected form image.

[0040] The delete button 604 is a button for deleting a form. When a job of an arbitrary form image is selected from the scanned form list 601 by a user operation and the delete button 604 is pressed, the job of the selected form image is deleted.

[0041] FIG. 6B is a diagram showing an example of an attribute setting screen 610 for setting attribute information such as a file name and metadata for a form image. An "attribute area" 611 is an area for displaying attribute information set for a form image. In the attribute area 611, the item names of each attribute, such as "file name" and "company name", are displayed, and a text field and a cutout image display area are provided for each item name. For example, a text field 618 for inputting attribute information for the item name "company name" and a cutout image 619 for the character area used to input the attribute information are displayed, as described later. If it is determined in S509 that there is no similar form (i.e., in the case of an unlearned form), the attribute setting screen 610 is initially displayed in a state in which no character area is selected (the text field is empty and no cutout image is set). On the other hand, if it is determined in S509 that there is a similar form, based on the position information of the character area previously selected in the similar form, the character recognition result of the character area at the corresponding position of the form image to be processed and the image of the character area are cut out, and the text field and the cutout image are automatically displayed.

[0042] In the "preview area" 612, a preview of the scanned form image is displayed.

[0043] When attribute information is set for each item in the attribute area 611 in the client PC, it can be set in the following procedure. For example, after selecting the item "company name", when the mouse is placed over a desired character area in the preview image, the character area over which the mouse is placed is highlighted (for example, the character area is displayed with a colored frame) based on the analysis result information acquired in S513. At this time, a character string of the character recognition result of the character area over which the mouse is placed is displayed in the text field 618, and a cutout image of the character area is displayed in the cutout image display area 619. Every time the character area over which the mouse is placed in the preview image is changed, the text field 618 and the cutout image 619 are switched in conjunction with each other. Then, when the mouse is clicked on the desired character area, the character recognition result and the cutout image of the character area over which the mouse is clicked are confirmed as setting input to the text field of the selected item name and the cutout image display area. In addition, the character string inputted into the text field based on the character recognition result can be corrected by using the keyboard (not shown) of the client PC or the like, such as correcting misrecognized characters or deleting unnecessary characters.

[0044] In the case of setting attribute information in each item of the attribute area 611 by touch operation in the MFP, the user touches the desired character area after selecting the desired attribute item. Then, the character recognition result and the cutout image corresponding to the touched character area are set. In addition, the character area used to set each attribute information is displayed in a highlighted state on the preview image.

[0045] The "Delete Character Area" button 613 is a button for clearing the character area used to set the attribute information, and is displayed in the upper right of the highlighted character area. When the "Delete Character Area" button 613 is pressed, the attribute information set in the attribute area 611 based on the character area is cleared, and the character area is no longer highlighted in the preview image.

[0046] The "Enlarge preview" button 614 is an operation button for increasing the magnification of the preview image displayed in the preview area 612. The "Reduce preview" button 615 is an operation button for decreasing the magnification of the preview image displayed in the preview area 612. The "Initial preview display" button 616 is an operation button for returning to the initial magnification and display position when the display magnification has been changed by pressing the preview enlargement button or the preview reduction button, or when the display position of the preview image has been moved by an operation such as swiping.

[0047] The "OK" button 620 is a button for issuing an instruction to send the attribute information (metadata such as file name and company name) set on the attribute setting screen 610 and the position information of the character area used to set the attribute information to the MFP collaboration service 120. If correction processing has been performed on the character string resulting from the character recognition in the text field, information regarding the correction content is also sent in response to pressing the OK button. When the "OK" button 620 is pressed and the information transmission is completed, the screen returns to the scanned form list screen 600. Upon receiving this information, the MFP collaboration service 120 executes learning processing (S516 to S519) based on the correction content of the area analysis result and the OCR result.

[0048] Fig. 7 is a flowchart showing details of the learning process (S519) based on the correction contents of the OCR result when the OCR result is corrected when setting attribute information on the attribute setting screen. The process when processing a new form image using the results learned by the learning process in Fig. 7 will be described in detail in Fig. 9 described later.

[0049] When the character string of the character recognition result displayed in the text field on the attribute setting screen 610 is corrected by the user, information on the correction content is transmitted from the client PC (or MFP) in S515. In such a case, in S701, the image processing unit 432 of the MFP cooperation service 120 acquires information on the correction content. For example, as shown in FIG. 6, when "Shimomaruko Co., Ltd." of the OCR processing result of the character area 617 selected on the preview image is displayed in the text field 618, the user determines that the "Co., Ltd." portion is unnecessary and deletes the character string as shown in the text field 813 of FIG. 8, leaving the character string "Shimomaruko". When such a character string deletion operation is performed, the attribute setting screen 810 may be controlled to highlight the character area 811 corresponding to the OCR result character string remaining in the text field 813, and to display the clipped image 812 corresponding to the character area. When such a deletion operation is performed, it is determined that the character string of the OCR processing result has been corrected, and information on the deleted character string is transmitted as information on the correction content. Note that, although an example has been described here in which MFP cooperation service 120 acquires information on the correction contents transmitted from a client PC (or MFP), the present invention is not limited to this. For example, a configuration may be adopted in which the client PC (or MFP) transmits a character string as an OCR result before correction and a character string after correction, and MFP cooperation service 120 acquires the correction contents by extracting the difference between the character string before and after correction to identify the deleted character string.

[0050] In S702, it is determined whether the correction is a deletion of a part of the OCR result character string and whether the deleted character string is a character string registered in the general term dictionary. The general term dictionary refers to a dictionary in which character strings that are not proper nouns, such as "Kaisho" (company name) and "quotation", are registered. For example, as shown in FIG. 8, when "Kaisho" is deleted from "Shimomaruko Kabushiki Gaisha", it is determined that the character string "Kaisho" is deleted and that "Kaisho" is a character string registered in the general term dictionary. In this case, the process proceeds to S703, where a regular expression is generated based on the original character string and the deleted character string, and learning is performed regarding the correction content of the OCR processing result. For example, when "Kaisho" is deleted from "Shimomaruko Kabushiki Gaisha" as shown in FIG. 8, the learning of the correction content of the OCR processing result is performed by registering the regular expression ".*Kaisho" together with the target form number (the number assigned to each form of the same format) as shown in Table 1 below. For the target form number, the form number determined at the time of similar form determination in S509 is used. The same form number will be assigned to forms with the same arrangement of character areas (forms determined to have the same format), and as will be described later, when OCR processing is performed on a new form image, a determination is made based on this form number as to whether or not to reflect the learning results.

[0051] [Table 1]

[0052] In the example of Figure 8, an example of creating a regular expression like that in Table 1 when deleting the rear part of the character string of the OCR result has been described, but this is not limited to this. For example, if characters in the front part or near the center of the OCR result character string are deleted, a regular expression indicating the deletion content will be generated in the same way if the deleted character string is registered in the general term dictionary.

[0053] In S704, the pre - correction OCR character string and the post - correction OCR character string are analyzed, and learning for correcting the OCR processing result is performed based on the analyzed result. For example, if the character string of the OCR processing result in the selected character area is the character string "artificial intelligence" (where the Chinese character "工" is misrecognized as the katakana "エ"), and the user corrects the misrecognition to the character string "artificial intelligence", as shown in Table 2 below, the target character string (string before replacement) is registered as "artificial intelligence" and the replacement string is registered as "artificial intelligence". The target form number uses the form number determined during similar form determination in S509 as in Table 1.

[0054]

Table 2

[0055] As described above, through the learning processes of S703 and S704, the regular expression information in Table 1 and the replacement character string information in Table 2 will be saved. By using the information (learning data) in Table 1 and Table 2, when a new form image of the same format is scanned later, when performing the character recognition process of S510 on the new form image, the OCR result can be automatically corrected.

[0056] Figure 9 is a detailed flow of the process of correcting the OCR result using the above - mentioned learning data in Table 1 and Table 2 when performing the character recognition process of S510 on a new form image.

[0057] In S901, the image processing unit 432 of the MFP cooperation service 120 performs character recognition (OCR) processing based on the partial image corresponding to each character area analyzed in S508 to obtain the character string of the character recognition result.

[0058] In S902, the image processing unit 432 determines whether the character string resulting from the OCR processing obtained in S901 is a character string that can be corrected by comparing it with the regular expression in Table 1 and the target character string in Table 2. That is, if the character string resulting from the OCR processing matches the regular expression in Table 1 or matches the target character string in Table 2, it is determined that the character string is a character string that can be corrected, and the process proceeds to S903.

[0059] In S903, the image processing unit 432 judges whether the new form image to be processed is judged to be a similar form in S509, and whether the form number of the judged similar form is included in at least one of the form numbers in Table 1 and Table 2. That is, it judges whether the form image newly generated by scanning this time is a form that has been previously corrected by OCR or has the same format as the form. If it is judged to be a form in a format that has been previously corrected, the same correction as the previous correction is automatically applied to the current OCR result (S904). On the other hand, if it is judged to be a form in a different format, the corrected character string is included in the OCR result so that it can be presented to the user (S905), so that it can be determined whether to apply the same correction as the previous correction to the current OCR result based on the user's instruction.

[0060] That is, in S904, the image processing unit 432 executes replacement of the character string of the character recognition result acquired in S901. Specifically, if the character string of the character recognition result matches the regular expression in Table 1, the character string portion of the regular expression is deleted. Also, if the character string of the character recognition result matches the target character string in Table 2, the character string of the character recognition result is replaced with the replaced character string in Table 2. For example, when a form image as shown in FIG. 10 similar to the form image in FIG. 8 is scanned, it is assumed that "Kosugi Corporation" is initially obtained as the character recognition result of the character region corresponding to the character region 811 in FIG. 8. In this case, since the character string of the character recognition result matches the regular expression ".* Corporation" registered in Table 1, the character string from which "Corporation Corporation" has been deleted is output as the OCR processing result. Specifically, the character string "Kosugi" and its corresponding coordinates are output as the OCR processing result data in FIG. 11. Based on this output result, when the attribute setting screen is displayed in S514, the character string "KOSUGI" is displayed in the text field 1013 as shown in FIG. 10, a partial image with "KOSUGI" cut out is displayed in the cut-out image display area 1012, and the corresponding character area 1011 is highlighted on the preview image.

[0061] In addition, in S905, the image processing unit 432 outputs both the character string resulting from the character recognition obtained in S901 and the character string after replacement based on the information in Table 1 or Table 2 as the OCR processing result. For example, it is assumed that the form image to be processed this time is not determined to have a similar format to the form image to be processed this time, but the character string obtained by character recognition of the selected OJI region matches the regular expression ".* Corporation" in Table 1. In this case, two character strings, namely, the character string "Kawasaki Corporation" resulting from the character recognition of the selected character region 1211, and the character string "Kawasaki" obtained by deleting the "Corporation" portion based on the regular expression in Table 1, are output as data of the OCR processing result. Specifically, as shown in FIG. 13, the character recognition result character string "Kawasaki Corporation" of the target character region and its replacement candidate character string "Kawasaki" are output as the OCR processing result data. Then, based on this OCR processing result data, each character string is displayed in a drop-down list in the text field 1212 as shown in FIG. 12. In addition, if the user selects "Kawasaki" from this drop-down list (i.e., selects a candidate string), the report number for the new report image can be added to and registered as the target report number in Table 1 or Table 2, and from the next time onwards, it can be treated as the report to be replaced rather than a candidate.

[0062] In this embodiment, whether to automatically replace the similar form using a replacement character string or to display candidates is determined based on the form number of the similar form in S903 of Fig. 9, but other conditions may be used. For example, it may be determined whether to apply the corrections in Tables 1 and 2 for each user or for each group to which the user belongs.

[0063] As described above, when attribute information such as a file name or metadata is set using a character string resulting from character recognition in a character region in a scanned image, if the character string resulting from character recognition is modified, the modification content is registered. Then, when attribute information is set for a new scanned image, the character string resulting from character recognition is displayed with the registered modification content reflected, thereby reducing the effort required for the user to modify the character string resulting from character recognition.

Claims

1. a character recognition unit for performing character recognition processing on a character area in the first scanned image to obtain a character recognition result; a learning means for learning a content of a correction made when a part of a character string of the character recognition result obtained by the character recognition means is corrected when attribute information for the first scanned image is set using the character recognition result obtained by the character recognition means, when the correction made when setting the attribute information for the first scanned image is a deletion of a part of a character string of the character recognition result, the learning means learns a regular expression based on the deleted character string as the correction content; the character recognition means performs character recognition processing on a character area in the second scanned image, and when a character string resulting from the character recognition of the character area in the second scanned image matches a regular expression learned by the learning means, deletes a part of the character string that matches the regular expression from the character string resulting from the character recognition of the character area in the second scanned image, and outputs the character recognition result after the deletion.

23. An information processing apparatus comprising:

2. a character recognition unit for performing character recognition processing on a character area in the first scanned image to obtain a character recognition result; a learning means for learning a content of a correction made when a part of a character string of the character recognition result obtained by the character recognition means is corrected when attribute information for the first scanned image is set using the character recognition result obtained by the character recognition means, when the correction made when setting the attribute information for the first scanned image is a deletion of a part of a character string of the character recognition result, and the deleted character string is a character string registered in a predetermined term dictionary, the learning means learns a regular expression based on the deleted character string as the correction content; the character recognition means performs character recognition processing on a character area in the second scanned image, and when a character string resulting from the character recognition of the character area in the second scanned image matches the learned regular expression, deletes a part of the character string that matches the regular expression from the character string resulting from the character recognition of the character area in the second scanned image, and outputs the character recognition result after the deletion.

23. An information processing apparatus comprising:

3. when the correction made when setting the attribute information for the first scanned image is a replacement of a part of a character string of the character recognition result, the learning means learns the character string before the replacement and the character string after the replacement as the correction content; the character recognition means performs character recognition processing on a character area in the second scanned image, and when a character string resulting from the character recognition of the character area in the second scanned image matches the character string before replacement learned by the learning means, replaces the character string resulting from the character recognition of the character area in the second scanned image with the learned character string after replacement, and outputs the character recognition result after replacement.

3. The information processing apparatus according to claim 1, wherein the information processing apparatus is a computer.

4. A program for causing a computer to function as each of the means of the information processing apparatus according to any one of claims 1 to 3.

5. a character recognition step of performing a character recognition process on a character region in the first scanned image to obtain a character recognition result; a learning step of learning the correction content when a part of a character string of the character recognition result obtained in the character recognition step is corrected when attribute information for the first scanned image is set using the character recognition result obtained in the character recognition step, When the correction made when setting the attribute information for the first scanned image is a deletion of a part of a character string of the character recognition result, the learning step learns a regular expression based on the deleted character string as the correction content; In the character recognition step, a character recognition process is performed on a character area in the second scanned image, and when a character string resulting from the character recognition of the character area in the second scanned image matches the learned regular expression, a part of the character string that matches the regular expression is deleted from the character string resulting from the character recognition of the character area in the second scanned image, and the character recognition result after the deletion is output.

23. An information processing method comprising:

6. a character recognition step of performing a character recognition process on a character region in the first scanned image to obtain a character recognition result; a learning step of learning the correction content when a part of a character string of the character recognition result obtained in the character recognition step is corrected when attribute information for the first scanned image is set using the character recognition result obtained in the character recognition step, when the correction made when setting the attribute information for the first scanned image is a deletion of a part of a character string of the character recognition result, and the deleted character string is a character string registered in a predetermined term dictionary, the learning step learns a regular expression based on the deleted character string as the correction content; In the character recognition step, a character recognition process is performed on a character area in the second scanned image, and when a character string resulting from the character recognition of the character area in the second scanned image matches the learned regular expression, a part of the character string that matches the regular expression is deleted from the character string resulting from the character recognition of the character area in the second scanned image, and the character recognition result after the deletion is output.

23. An information processing method comprising:

Images