JP2006277001A - Input image displaying method, and input image displaying program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce, quicken, and simplify movement of line of sight to an item image during confirmation and correction of an item in a text field by displaying the corresponding item image near the item pointed by a cursor in the text field displaying a character recognition result obtained by carrying out character recognition on a full-page image acquired from a business form, highlighting the item image in the full-page image, repeating reflection in the item when there is correction input. <P>SOLUTION: In the input image displaying method and the input image displaying program, the character recognition result from the full-page image acquired from the business form is confirmed and corrected. They have a step of displaying the character recognition result obtained by carrying out character recognition of the full-page image in the text field, and a step of repeating acquisition and display of the item image in the full-page image in the neighborhood of the selected image in the displayed text field. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an input image display method and an input image display program for confirming and correcting a character recognition result from a full image acquired from a form.

Conventionally, a form that has been handwritten is scanned, the form image is displayed on a screen, characters are recognized from the form image, and a character recognition result is input to a predetermined input field. At this time, when the operator confirms and corrects the characters and numbers for which the character recognition results have been input, the scanned form image is displayed on one screen, and the form in which the character recognition results are input is displayed on the other screen. (Or split one screen horizontally or vertically and display the form image on one side and the character recognition result on the other), check the two while comparing them, or correct the character recognition result when necessary (Patent Document 1).
Japanese Patent Laid-Open No. 05-303661

  In the conventional technology described above, the operator needs to visually confirm the corresponding image on the form image associated with the cursor for the character recognition result of the handwritten input on the form. It is necessary to move from the screen to the position of the cursor on the screen and check it, and it takes time to view, and the troublesome task of searching for and checking the location associated with the cursor is forced. There was a problem.

  In order to solve these problems, the present invention displays an item image corresponding to the vicinity of the item where the cursor is positioned in a text field displaying a character recognition result obtained by character recognition from the entire image acquired from the form, and also displays the entire image. The item image above is highlighted, and when there is a correction input, it is repeatedly reflected in the item.

  The present invention displays an item image in the vicinity of the item where the cursor is positioned in the text field in which character recognition is performed, highlights the item image on the entire image, and reflects the correction item in the item when there is a correction input. By repeating the above, it is possible to perform confirmation and correction work quickly and easily with very little movement of the line of sight to the item image at the time of confirmation and correction of the item on the text field.

  The present invention displays an item image corresponding to the vicinity of the item where the cursor is positioned in a text field displaying a character recognition result obtained by character recognition from the entire image acquired from the form and highlights the item image on the entire image. Then, when correction input is made, it is repeatedly reflected in the item, and the movement of the line of sight to the item image at the time of confirmation and correction of the item on the text field has been realized extremely quickly and easily.

FIG. 1 shows a system configuration diagram of the present invention.
In FIG. 1, a system 1 executes various processes according to a program, and is a computer such as a personal computer, and includes a data acquisition means 2, a screen display means 3, a data correction means 4, a form database 5, a display device 6, And the input device 7 and the like.

  The data acquisition means 2 acquires data from the form database 5 or the like. For example, the entire image read by a scanner from a handwritten form, and further, the character is recognized from the entire image and set by software not shown. A text field or the like is acquired (see FIGS. 2 to 4).

  The screen display means 3 displays the data acquired by the data acquisition means 2 (data such as full image, text field) on the screen of the display device 6 (see FIGS. 2 to 4).

  The data correction means 4 corrects the data. For example, when the item image displayed near the selected item on the text field is viewed by a person and corrected to the correct character or value, the item is corrected. And the like are corrected (see FIGS. 2 to 4).

The form database 5 is registered with information related to forms (see FIG. 3).
The display device 6 displays a screen. Here, the display device 6 displays the screen of FIG. 4 (entire image, text field) (see FIG. 4).

  The input device 7 inputs data and instructions, and is a keyboard or a mouse.

Next, the operation of the configuration of FIG. 1 will be described in detail according to the order of the flowchart of FIG. 2. FIG. 2 shows an operation description flowchart of the present invention.

  In FIG. 2, S1 acquires a full image corresponding to the form to be displayed. In this case, a full image corresponding to a form to be displayed and to be corrected for the automatic character recognition result is acquired from the form database 5 here. The acquisition is performed by designating, for example, “form A” as the form ID in the form database 5 of FIG. 3 and displaying the entire image A (for example, the entire image A of the record of (1) in FIG. ) To get.

  In S2, a cut-out image corresponding to the form to be displayed is acquired (for all items in consideration of performance). This is the performance of the system (computer) 1 in FIG. 1 among the item images 1, 2, 3,... N as the cut-out image corresponding to the record of the form A of the entire image acquired in S1. Acquire as many item images as possible in consideration, for example, acquire as many item images as possible in consideration of memory capacity, and select items (character recognition) in the text field during display / confirmation / correction after S3 The item image is prepared so that it can be displayed at high speed in the vicinity of the item for which the result is set.

  In S3, the entire image is displayed on the screen. This is because the entire image of the form to be displayed acquired in S1 (the entire image of the form obtained by reading a handwritten form with a scanner) is displayed on the screen, for example, the entire image of the upper stage of FIG. To display.

  S4 displays a text field on the screen. This is because the text field of the form to be displayed acquired in S2 (the text field in which the character or numerical value of the entire image is recognized) is displayed on the screen, for example, the text field in the lower part of FIG. To display.

  In step S5, it is determined whether the item is escaped. This is, for example, from the entire image above the item in the text field at the bottom of the screen of FIG. 4A, here the item “?? Fushi?…” On the right input field of the name. While displaying the item image corresponding to the cut out item, the operator selects another item (for example, the lower item (telephone number)) and determines whether the item is escaped. If YES, the process proceeds to S6. If NO, skip S6 and S7 and proceed to S8.

  S6 is YES when it is S5, and it has been found out from one item in the text field to another item, so if the cut-out image (item image) is being displayed, it is deleted, for example, the lower text in FIG. In the currently selected item (name) in the field, the image (item image handwritten input of the item (name)) cut out from the corresponding full image is displayed on the item (name). Therefore, the item image is deleted.

  S7 deletes the image cursor when it is being displayed. This erases the image cursor highlighted in the illustrated frame (actually a red frame) that recognizes the character corresponding to the item (name) in the lower text field on the entire upper image in FIG. To do.

  In step S8, it is determined whether the item is an input target. This determines whether the item selected in the text field is an input target. If YES, the process proceeds to S9. In the case of NO, S9 and S10 are skipped and the process proceeds to S11.

  In S9, it is determined that the selected item in the text field is an input target in YES of S8, and therefore a cut-out image corresponding to the item is displayed on the upper part of the item. For example, since an item (telephone number) in the lower text field of FIG. 4B to be described later is an input target, an item image (full image) cut out from the entire image above the item (telephone number). The item image input by handwriting in the upper item (telephone number) field is displayed as shown (the image of the handwritten character “1234” is displayed).

  In S10, an image cursor corresponding to the item is displayed on the entire image. This is because the image cursor (actually a red rectangular frame) is displayed on the item image portion of the item (telephone number) on the entire image at the top of the screen in FIG. The correspondence with the value set in the item (telephone number) of the text field in which the result is set is displayed in an easily understandable manner.

Through the above S9 and S10, the character recognition result is set in the input field of the selected item (telephone number) in the text field at the bottom of the screen of FIG. 4B, and the item in the full image is displayed above it. An item image obtained by reading an image input by handwriting in the (telephone number) column with a scanner is cut out and displayed as shown, and the original full-screen image item (telephone number) in the upper part of the screen in FIG. The image cursor (red rectangular frame) can be displayed in the handwritten input image (item image) portion. As a result, the operator compares the character recognition result (1234) set for the item (phone number) selected in the text field with the item image (image 1234) displayed immediately above, and obtains the character recognition result. Here, it is possible to immediately determine the correctness without moving the line of sight, and the position of the item image on the original whole image is displayed in association with the image cursor (red rectangular frame). The location on the original form and the handwritten image can be easily confirmed together. In step S11, it is determined whether there is a correction input. This is because the character recognition result characters and values set in the selected item in the text field at the bottom of the screen of FIG. 4B and the item image displayed immediately above (the item input by handwriting) The operator who sees the above image) determines whether correction input is necessary for the character recognition result. In the case of YES, the input data is reflected in the text field in S12, for example, the characters and numerical values of the character recognition result set in the item are corrected and input by operating the keyboard to make it correct. Then, the process proceeds to S13. On the other hand, if NO in S11, the process proceeds to S13.

  In step S13, it is determined whether the input process is finished. In the case of YES, the input data is stored in the form database 5 of FIG. In the case of NO, S5 and subsequent steps are repeated.

  As described above, when the operator designates a form, the system 1 obtains a full image of the corresponding form from the form database 5, obtains a cut-out image (item image), and obtains a text field as a result of character recognition. Since it is displayed on the screen of the display device 6 as shown in a), when the operator selects an arbitrary item in the text field at the bottom of the screen of FIG. The item image corresponding to the item cut out from the image is displayed, and the part of the item image corresponding to the item in the original entire image is highlighted with the image cursor. As a result, the operator moves the line of sight of the character recognition result character and value set in the item selected in the lower text field of the screen of FIG. 4A and the item image displayed immediately above. It is possible to compare and check immediately without any problems. Further, as to which location or item in the original form the system 1 is, the portion of the entire image in the upper part of the screen in FIG. The operator highlights the image cursor on the form and the image (item image) of the image cursor in the entire image, and easily confirms the image input by handwriting and the location more accurately. It becomes possible.

  FIG. 3 shows an example of a form database of the present invention. The illustrated form database 5 registers and manages the entire image for each form, the item images 1, 2,... N extracted from the entire image for each item, and the character recognition results (recognition data) in association with each other. Here, the following information shown in the figure is registered and managed in association with each other.

・ Form ID:
-Image data ID:
・ Image data:
・ Recognition data:
·Input data:
・ Other:
Here, the form ID is an ID (identifier) that uniquely identifies the form input by handwriting. For example, the form A is “Fujitsu Department Store Card Application Form” in FIGS. 4A and 4B. . The image data ID is an ID (identifier) that uniquely identifies the entire image of the form, a cut-out image (item image) cut out for each item, and the like. The entire image A is, for example, an image (entire image) displayed on the upper part of the screen shown in FIGS. The item image 1 is, for example, an image (image shown in the figure) “1234” of the item (telephone number) portion cut out from the entire image displayed above the item (telephone number) in the lower text field of FIG. )). The recognition data is a character recognition result obtained by recognizing each item image 1, 2, 3,... N. For example, the character recognition result “1234” set in the item (telephone number) in the lower text field of the screen of FIG. The input data is data input by the operator for each item image 1, 2, 3,... N.

FIG. 4 shows an example screen of the present invention.
FIG. 4A shows a screen example of a full image and a text field when a certain item (here, application date) is selected. When the item (name) in the illustrated text field is selected, the image (item image) of the handwritten item part of the item (name) cut out from the entire image is displayed at the top. Is displayed. In addition, the location of the item (name) in the entire image on the upper part of the screen in FIG. 4A is highlighted with an image cursor (red rectangular frame), and the item image in the original entire image of the cut item image is displayed. The location is clearly displayed.

  When the input target item is switched by selecting another item (phone number) from the selected item (name) in the text field in the state displayed as described above, as shown in FIG. become.

  FIG. 4B shows a screen example when the item (application date) is switched to the item (telephone number). Here, from the state of FIG. 4A, when the operator selects an item (telephone number) and instructs switching, the item image after switching is displayed above the item (telephone number) after switching by the system 1. (The item image of the item (telephone number) after switching in the entire image) is displayed, and the portion of the item (telephone number) in the entire image is highlighted with an image cursor (red rectangular frame). As a result, the item image of the item (phone number) handwritten at the top of the character recognition result (here, 1234) of the item (phone number) in the newly switched text field is displayed as shown in the figure. The operator can immediately confirm the character recognition result and the original item image without moving the line of sight. Then, since the location of the item (phone number) in the original full image is displayed with an image cursor (red rectangular frame), the operator visually observes the location of the image cursor and the handwritten image in it. The character recognition result can be determined more accurately with respect to the character recognition result based on the location of the image cursor in the entire form.

  The present invention displays an item image in the vicinity of the item where the cursor is positioned in the text field in which character recognition is performed, highlights the item image on the entire image, and reflects the correction item in the item when there is a correction input. By repeating the above, it is possible to perform confirmation and correction work quickly and easily with very little movement of the line of sight to the item image at the time of confirmation and correction of the item on the text field.

It is a system configuration diagram of the present invention. It is an operation | movement explanatory flowchart of this invention. It is an example of the form database of this invention. It is an example of a screen of the present invention.

Explanation of symbols

1: System (PC etc.)
2: Data acquisition means 3: Screen display means 4: Data correction means 5: Form database 6: Display device 7: Input device

Claims (5)

In the input image display method that confirms and corrects the character recognition result from the entire image acquired from the form,
Displaying a character recognition result of character recognition from the entire image in a text field;
Repeatedly acquiring and displaying the corresponding item image on the entire image in the vicinity of the selected item on the displayed text field;
An input image display method comprising:   The input image display method according to claim 1, further comprising a step of updating when the selected item is corrected and input on the text field.   When the corresponding item image on the entire image is acquired and displayed near the selected item on the displayed text field, the item image on the entire image is highlighted. The input image display method according to claim 1 or 2.   4. The item image on the entire image is acquired and displayed on the upper side, the lower side, the left side, or the right side of the selected item on the displayed text field. The input image display method according to any one of the above. In the input image display program that confirms and corrects the character recognition result from the entire image acquired from the form,
On the computer,
Displaying a character recognition result of character recognition from the entire image in a text field;
An input image display program for functioning as a step of repeatedly acquiring and displaying a corresponding item image on the entire image in the vicinity of a selected item on the displayed text field.

Images