JP2002133367A - Character recognition device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an OCR that can permit instantaneous identification of an input image corresponding to an illegible character and a reliable and easy check and correction of a recognition result. SOLUTION: A field detection part 13 extracts a character input image from a picture image read by a scanner 11 and stored in a picture memory 12, and stores the input image in a field memory 15. The input image is then subjected to character image extraction for every single character in a character extraction part 16 and to sequential recognition in a character recognition part 17, whose recognition results are stored in a recognition result memory 18. The input image and the recognition results are read out by a screen display generation part 19A, in which a character image corresponding to an illegible character is provided with a display color and a character frame different from those for a correctly read character. The input image and the recognition results are displayed comparatively on a display part 21 as screen display for checking and correcting.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition device (hereinafter referred to as "OCR"), and more particularly to a display format of a recognition result.

[0002]

2. Description of the Related Art FIG. 2 is a block diagram showing an example of a conventional OCR. The OCR has a scanner 11 that optically reads a form to be recognized. An image memory 12 is connected to the scanner 11 so that an image of a read form is temporarily stored.

The image memory 12 has a field detector 1
3 are connected. The field detection unit 13 detects a position of a field to be read in accordance with the format information of the form registered in the form format registration unit 14 and outputs the field image. On the output side of the field detecting unit 13, a field memory 15 and a character extracting unit 16 are connected. The field memory 15
The field image is stored to correct the recognition result.

The character extracting section 16 extracts a character image from a field image in units of one character, and a character recognizing section 17 is connected to the output side. The character recognizing unit 17 extracts the features of the cut-out character image, recognizes the character, and outputs the character code of the corresponding character. A recognition result memory 18 for storing a recognition result is connected to an output side of the character recognition unit 17.

[0005] Field memory 15 and recognition result memory 1
A display screen generation unit 19 that generates a screen for checking and correcting the recognition result is connected to the output side of the display unit 8. A display memory 20 for storing screen information is connected to an output side of the display screen generation unit 19, and the screen information stored in the display memory 20 is displayed on the display unit 21.

The OCR has an input unit 22 for correcting the recognition result displayed on the display unit 21 as necessary. The input unit 22 is connected to the recognition result correction unit 23, and can correct the recognition result stored in the recognition result memory 18 by the correction data input from the input unit 22. The recognition result for which the correction processing has been completed is output from the recognition result memory 18 as an output signal OUT.

FIGS. 3 (a) and 3 (b) are diagrams showing display examples of recognition results in the OCR of FIG. Hereinafter, the operation of FIG. 2 will be described.

An image of a form to be read is read by a scanner 11 and temporarily stored in an image memory 12. A field image of a character portion is cut out from the image image in the image memory 12 by the field detection unit 13, stored in the field memory 15, and given to the character cutout unit 16. The character image in the field image is cut out character by character by the character cutout unit 16 and sequentially recognized by the character recognition unit 17 and converted into a character code. The converted character code is stored in the recognition result memory 18.

When the recognition of all the characters in one field is completed, the field image stored in the field memory 15 and the character code of the recognition result stored in the recognition result memory 18 are read out by the display screen generation unit 19. Screen information for confirmation / correction of the recognition result is generated,
It is stored in the display memory 20. The contents of the display memory 20 are displayed on the screen by the display unit 21 as shown in FIG.

As shown in FIG. 3A, on the display screen before correction, the field image as input is displayed as an input image, and the character code of the recognition result is displayed in black with a predetermined font. . For unrecognized characters that are rejected without being recognized, the recognition result is displayed as red "?", And a cursor for correction input is set.

The operator confirms the input image corresponding to the unread character, selects an appropriate character from a plurality of presented candidate characters with a mouse, or inputs a correct character from a keyboard. As a result, the screen of the display unit 21 is changed to a corrected display screen as shown in FIG. 3B, and the unread character “?” Is replaced with the corrected and input blue character.

As described above, since the input image and the recognition result are displayed in comparison with each other, and the unread characters are displayed as red "?", The unread portion can be corrected without overlooking the unread portion. .

[0013]

However, the conventional OCR has the following problems. That is, since there is no correspondence between the unread characters displayed as the recognition result and the characters in the input image, when the number of characters in the field increases, the image corresponding to the unread characters cannot be instantaneously specified. .

An object of the present invention is to provide an OCR capable of solving the problems of the prior art and confirming and correcting the recognition result reliably and easily.

[0015]

In order to solve the above problems, a first aspect of the present invention is a reading means for optically reading a document to be read in an OCR to generate an image image, A registration unit in which format information of the document is registered in advance; a detection unit that detects a character field in which a plurality of characters are described from the image image based on the format information; and a character field detected by the detection unit. Recognizing means for recognizing a plurality of characters, outputting a corresponding character code when each character can be correctly recognized, and outputting a symbol code indicating unread when not correctly recognizing the character field image and the recognition. Display means for displaying the recognition result of the means on the same screen, and an image displayed on the display means based on the correction data provided from the input unit. While changing the display format of di-, and a correction means for correcting the recognition result of said recognition means.

According to the first invention, as described above, the OCR
, The following operation is performed.

The document to be read is optically read by the reading means to generate an image image. The image is provided to the detection means, and a character field in which a plurality of characters are described is detected based on the format information of the document registered in the registration means in advance. A plurality of characters included in the character field are recognized by the recognizing means, and the characters that have been correctly recognized have a corresponding character code,
For characters that cannot be recognized correctly, a symbol code indicating unreadness is output.

The image of the character field and the recognition result of the recognition means are displayed on the same screen by the display means. Further, when correction data is given from the input unit to the correction means, the correction means changes the display format of the image based on the correction data and corrects the recognition result.

According to a second aspect of the present invention, the display means in the first aspect displays a character image in a character field corresponding to a character that has not been correctly recognized by the recognition means in a display color different from that of the correctly recognized character. It is configured to be.

In a third aspect of the present invention, the display means according to the first or second aspect displays a character image in a character field corresponding to a character that has not been correctly recognized by the recognition means with a character frame added thereto. It is composed.

According to a fourth aspect of the present invention, the display means according to the first or the second aspect of the present invention comprises the steps of adding a character frame to all the character images in the character field and adding a character frame to each character and symbol of the recognition result. It is configured to be displayed.

According to a fifth aspect, in the fourth aspect, the display means is provided with a character frame for adding a character in a character field corresponding to a character which has not been correctly recognized by the recognition means.
It is configured to display in a display color different from a character frame to be added to a character that has been correctly recognized.

In a sixth aspect, the correction means in the first to fifth aspects is configured to change a display color of a character image corrected by the correction data and a display color of a character as a recognition result. .

According to a seventh aspect of the present invention, in addition to the character field image and the recognition result of the recognizing means, the display means of the first to sixth aspects enlarges a character image which cannot be correctly recognized by the recognizing means. It is configured to be displayed on the same screen.

According to an eighth aspect of the present invention, in addition to the character field image and the recognition result of the recognizing means, the display means of the first to sixth aspects further comprises a character image which cannot be correctly recognized by the recognizing means and a peripheral portion thereof. Is enlarged and displayed on the same screen.

According to a ninth aspect of the present invention, in addition to the character field image and the recognition result of the recognizing means, the display means in the first to sixth aspects is further provided with a character image and a character frame which cannot be correctly recognized by the recognizing means. The image of the periphery is enlarged and displayed on the same screen.

According to a tenth aspect, in the first to ninth aspects, there is provided changing means for changing a display range of an image displayed on the display means within a certain area based on an instruction given from the input unit. ing.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) FIG. 1 is a block diagram of an OCR showing a first embodiment of the present invention.
Elements common to those in the middle are denoted by common reference numerals.

This OCR has reading means (for example, a scanner) 11 for optically reading a form on which characters to be recognized are described, similarly to the OCR of FIG. An image memory 12 is connected to the scanner 11 so that an image of a read form is temporarily stored.

The image memory 12 has detection means (for example,
Field detector 13 is connected. The field detection unit 13 includes a registration unit (for example, a form format registration unit) 1
According to the format information of the form registered in No. 4, the position of the field where the character to be read exists is detected, and the field image is output. On the output side of the field detecting unit 13, a field memory 15 and a character extracting unit 16 are connected.

The field memory 15 stores a field image for correction processing of the recognition result. The character cutout unit 16 cuts out a character image in a field image in units of one character for character recognition. A recognition unit (for example, a character recognition unit) 1 is provided on the output side.
7 is connected.

The character recognizing section 17 extracts the features of the extracted character image, recognizes the character, and converts the character into the character code of the corresponding character. The character recognition unit 17
Converts a character image that cannot be correctly recognized into a symbol code of a symbol "?" Indicating unreadness and outputs the code. A recognition result memory 18 for storing a recognition result is connected to an output side of the character recognition unit 17.

Field memory 15 and recognition result memory 1
A display unit (for example, a display screen generation unit) 19A for generating a screen for checking and correcting the recognition result is connected to the output side of the display unit 8. The display screen generation unit 19A generates a display screen that displays a first display color (for example, black) by comparing the field image with the recognition result. Furthermore,
When the recognition result includes an unread character that has been rejected without being recognized, the display screen generation unit 19A displays the recognition result of the unread character in the second display color (for example, red) as “ "?" And a function of changing an unread portion in the field image to a third display color (for example, pink) for display.

A display memory 20 for storing screen information is connected to the output side of the display screen generation unit 19A, and the screen information stored in the display memory 20 is read out and displayed by the display unit 21. ing.

The OCR has an input unit 22 such as a keyboard and a mouse for correcting the recognition result displayed on the display unit 21 as necessary.
It is connected to correction means (for example, a correction processing unit) 30.

The correction processing unit 30 includes a correction control unit 31, a recognition result generation unit 32, and a display screen correction unit 33, and the input unit 22 is connected to the correction control unit 31. The correction control unit 31 controls the recognition result generation unit 32 and the display screen correction unit 33 based on the correction data input from the input unit 22 and the format information of the form registered in the form format registration unit 14. .

The recognition result generator 32 corrects the recognition result stored in the recognition result memory 18 under the control of the correction controller 31. The display screen correction unit 33 similarly corrects the contents of the screen information stored in the display memory 20 under the control of the correction control unit 31. The recognition result corrected according to the correction data input from the input unit 22 is output from the recognition result memory 18 to the output signal O.
The data is output as UT.

FIGS. 4A and 4B are diagrams showing examples of displaying the recognition result in the OCR shown in FIG. Hereinafter, FIG.
The operation of FIG. 1 will be described with reference to (a) and (b).

The image of the form to be read is read by the scanner 11 and is temporarily stored in the image memory 12. From the image image stored in the image memory 12, a field image of a character portion is cut out by the field detection unit 13, stored in the field memory 15, and given to the character cutout unit 16. The character image in the field image is cut out character by character by the character cutout unit 16 and is sequentially recognized by the character recognition unit 17. A correctly recognized character image is converted to a character code. A character image that is not correctly recognized is rejected as an unread character, and a character code of a plurality of candidate characters corresponding to the unread character is output together with a symbol code of “?”. These recognition results are stored in the recognition result memory 18.

When the recognition of all the characters in one field is completed, the field image stored in the field memory 15 and the recognition result stored in the recognition result memory 18 are read out by the display screen generating unit 19A, and are confirmed and corrected. Screen information is generated and stored in the display memory 20. The contents of the display memory 20 are displayed by the display unit 21 as shown in FIG.
The screen is displayed as shown in FIG.

As shown in FIG. 4A, the field image as input is displayed on the display screen before correction as an input image. In the input image, a field image corresponding to a correctly recognized character is displayed in black, and a field image corresponding to an unreadable character is displayed in pink.

On the other hand, in the recognition result, a correctly recognized character code is displayed in black with a predetermined font, unread characters are displayed as red "?", And a correction input is made at the position of the unread character. The cursor for is set.

The operator checks the input image corresponding to the unread character, selects an appropriate character from the plurality of presented candidate characters with a mouse, or inputs a correct character from the keyboard. As a result, the display screen correction unit 33 corrects the "?" Of the recognition result in the display memory 20 to the corrected input character, and changes the display color to the fourth display color (for example, blue). You. As a result, the screen of the display unit 21 is changed to the corrected display screen shown in FIG.

As described above, the OC of the first embodiment
R has a display screen generation unit 19A that generates a display screen that can easily identify correctly recognized characters and unreadable characters in an input image, and also changes the contents of the display memory 20 according to a correction input from the input unit 22. It has a correction processing unit 30 for correcting. This makes it possible to instantaneously specify the image corresponding to the unreadable character, and it is possible to reliably and easily confirm and correct the recognition result.

(Second Embodiment) FIGS. 5A and 5B
FIG. 9 is a diagram illustrating a display example according to the second embodiment of the present invention.
The OCR used in the second embodiment is shown in FIG.
It is almost the same as However, in addition to the function of correcting the unreadable character in the recognition result based on the correction data of the input unit 22, the display image correction unit 33 changes the display color of the input image of the corrected character to the fourth color same as the recognition result. It has a function of changing to a color (ie, blue). In this OCR, almost the same operation as in FIG. 1 is performed.

That is, the recognition result of the form to be read is
As shown in FIG. 5A, among the input images, the field image corresponding to the correctly recognized character is displayed in black, and the field image corresponding to the unreadable character is displayed in pink. On the other hand, in the recognition result, a correctly recognized character code is displayed in black with a predetermined font, unread characters are displayed as red “?”, And the position of the unread character for correction input is displayed. The cursor is set.

Next, a correction input corresponding to the unread character is performed by the operator. As a result, the recognition result correcting unit 32 replaces the unread character “?” In the recognition result with the corrected and input blue character, and the display screen correcting unit 33 changes the input image corresponding to the corrected character. Changed to blue. The modified screen is shown in FIG.
It is displayed as shown in (b).

As described above, the OC of the second embodiment
R is a display screen generation unit 19A that generates a display screen that can identify input images of correctly recognized characters and unreadable characters.
And a correction processing unit 30 for correcting the contents of the display memory 20 in accordance with the correction input from the input unit 22. This makes it possible to instantaneously specify the image corresponding to the unreadable character, and it is possible to reliably and easily confirm and correct the recognition result. Further, since the display color of the corrected input image is changed, it is possible to easily distinguish uncorrected and corrected characters, thereby preventing omission of correction.

(Third Embodiment) FIG. 6 shows a third embodiment of the present invention.
FIG. 2 is a configuration diagram of an OCR showing the embodiment of FIG. 1, in which components common to those in FIG. 1 are denoted by common reference numerals.

In this OCR, a character frame generator 41 is provided in the OCR of FIG. 1, and a display screen generator 19B having a different function is provided in place of the display screen generator 19A. The character frame generation unit 41 generates information of a character frame corresponding to a character to be recognized according to the format information of the form registered in the form format registration unit 14 and generates the display screen generation unit 19.
B. The display screen generation unit 19B
Generates a display screen that compares the field image with the recognition result. The display screen generation unit 19B displays the unread character in the recognition result as “?” In the second display color (that is, red), and displays the unread character in the input image in the third display color (that is, red). , Pink). Other configurations are the same as those in FIG.

FIGS. 7A and 7B are diagrams showing display examples of recognition results in the OCR of FIG.

The form format registration unit 14 is referred to by the character frame generation unit 41, and input based on the format information of the number of characters, character width, character height, and character spacing for each field registered in the form format registration unit 14. The position of the character frame added to the image is calculated and provided to the display screen generation unit 19B. Then, in the display screen generation unit 19B,
In the input image of the form to be read, the character image is displayed in the first display color (that is, black), and a pink character frame is added to the unread characters. In the recognition result, a correctly recognized character code is displayed in black with a predetermined font, unread characters are displayed as red "?", And a cursor for correction input is placed at the position of the unread character. The screen information in which is set is generated. The generated screen information is stored in the display memory 20 and displayed on the display unit 21 as shown in FIG.

Next, when a correction input corresponding to the unread character is performed by the operator, the correction processing unit 30
As shown in FIG. 7B, the unread character "?"
Is replaced with the corrected blue text,
The input image corresponding to the corrected character is changed to blue.

As described above, the OC of the third embodiment
R is a character frame generation unit 41 that generates character frame data;
It has a display screen generation unit 19B that generates screen information in which a character frame is added to an input image of unread characters. This makes it possible to instantaneously specify the image corresponding to the unreadable character, and it is possible to reliably and easily confirm and correct the recognition result.

(Fourth Embodiment) FIGS. 8A and 8B
FIG. 14 is a diagram showing a display example according to the fourth embodiment of the present invention.
The OCR used in the fourth embodiment is shown in FIG.
It is almost the same as However, the display screen generating unit 19 </ b> B according to the character frame data given from the character frame generating unit 41
It has a function of generating screen information in which a character frame is added to both the input image and the recognition result. In this OCR,
Operation substantially similar to that of FIG. 6 is performed.

That is, the input image of the form to be read is, as shown in FIG. 8A, a pink character frame for dividing the character image in accordance with the character frame position information calculated by the character frame generating unit 41. Displayed in black.

The character of the recognition result is also displayed in the character frame. Correctly recognized characters are displayed in black with a predetermined font, unread characters are displayed as red "?", And a cursor for correction input is set at the position of the unread character.

Next, when a correction input corresponding to the unread character is performed by the operator, as shown in FIG.
The unread character “?” In the recognition result is replaced with the corrected blue character.

As described above, the OC of the fourth embodiment is
R is a character frame generation unit 41 that generates character frame data;
It has a display screen generation unit 19B that generates a display screen in which a character frame is added to the entire input image and the recognition result.
As a result, the position of “(blank)” can be easily specified, and confirmation and correction of the recognition result can be performed reliably and easily.

(Fifth Embodiment) FIGS. 9A and 9B
FIG. 14 is a diagram showing a display example according to the fifth embodiment of the present invention.
The OCR used in the fifth embodiment is shown in FIG.
It is almost the same as However, the display screen generating unit 19 </ b> B according to the character frame data given from the character frame generating unit 41
It has a function of generating display screen data in which a character frame is added to both the input image and the recognition result. In particular, in the input image, the character frame corresponding to the unread character is displayed in the third display color (that is, pink), and the character frame corresponding to the normally read character is displayed in the first display color (that is, pink). (Black). In this OCR, almost the same operation as in FIG. 6 is performed.

That is, as shown in FIG. 9 (a), the input image of the form to be read is such that the correctly recognized character image is displayed in black in a black character frame, and the unread character image is pink. Is displayed in black in the character frame of.

The character of the recognition result is also displayed in a black character frame. Correctly recognized characters are displayed in black with a predetermined font, unread characters are displayed as red "?", And a cursor for correction input is set at the position of the unread character.

Next, a correction input corresponding to the unread character is performed by the operator, and as shown in FIG. 9B, the unread character "?" In the recognition result is replaced with the corrected blue character. Can be

As described above, the OC of the fifth embodiment
R is a character frame generation unit 41 that generates character frame data;
It has a display screen generation unit 19B that generates screen information in which a character frame is added to the entire input image and the recognition result.
As a result, the position of “(blank)” can be easily specified, and confirmation and correction of the recognition result can be performed reliably and easily. Further, in the display screen generation unit 19B,
Since a character frame of a color different from that of a correctly recognized character is added to the image of the unread character, the unread portion can be immediately identified.

(Sixth Embodiment) FIG. 10 is a block diagram of an OCR showing a sixth embodiment of the present invention. Elements common to those in FIG. 1 are denoted by the same reference numerals. .

This OCR is obtained by adding an enlarged image generation unit 42 to the OCR of FIG. The enlarged image generation unit 42
An enlarged image of the input image corresponding to the unreadable character is generated, and additional writing is performed on the display memory 20. Other configurations are the same as those in FIG.

FIG. 11 is a diagram showing a display example of the recognition result in the OCR of FIG. That is, as shown in FIG. 11, the recognition result of the form to be read is such that the input image and the recognition result are displayed in units of fields, and the character image corresponding to the unread character is added as an enlarged image. Is displayed.

As described above, the OC of the sixth embodiment
R has an enlarged image generation unit 42 for enlarging and displaying an input image of unreadable characters. This makes it possible to instantaneously specify the image corresponding to the unreadable character, and it is possible to reliably and easily confirm and correct the recognition result.

(Seventh Embodiment) FIG. 12 is a diagram showing a display example of a seventh embodiment of the present invention. Note that this seventh
The OCR used in the embodiment is almost the same as that in FIG. However, the enlarged image generating unit 42 includes not only the image corresponding to the unreadable character but also the image around the unread character.
It has a function to additionally display as an enlarged image. In this OCR, almost the same operation as in FIG. 10 is performed.

That is, the recognition result of the form to be read is
As shown in FIG. 12, the input image and its recognition result are:
In addition to being displayed on a field-by-field basis, a character image corresponding to the unreadable character and its surrounding image are additionally displayed as an enlarged image.

As described above, the OC of the seventh embodiment
R has an enlarged image generator 42 for enlarging and displaying an input image of unreadable characters and an image around the input image. As a result, an image corresponding to the unreadable character can be instantaneously specified, and a correct character can be easily determined, and confirmation and correction of the recognition result can be performed reliably and easily.

(Eighth Embodiment) FIG. 13 is a diagram showing a display example of an eighth embodiment of the present invention. Note that this eighth
The OCR used in the embodiment is almost the same as that in FIG. However, the enlarged image generation unit 42 has a function of adding not only an image corresponding to the unreadable character but also a character frame to the image and additionally displaying an image around the image as an enlarged image. . In this OCR, almost the same operation as in FIG. 10 is performed.

That is, the recognition result of the form to be read is
As shown in FIG. 13, the input image and its recognition result are
Displayed in units of fields. Further, an image around the character image corresponding to the unreadable character with the character frame added thereto is additionally displayed as an enlarged image.

As described above, the OC of the eighth embodiment
R has an input image of an unreadable character added with a character frame and an enlarged image generation unit 42 for enlarging and displaying an image around the input image. As a result, an image corresponding to the unreadable character can be instantaneously specified, and a correct character can be easily determined, and confirmation and correction of the recognition result can be performed reliably and easily.

(Ninth Embodiment) FIG. 14 is a diagram showing the configuration of an OCR according to a ninth embodiment of the present invention. Elements common to those in FIG. 1 are denoted by the same reference numerals. .

In this OCR, a change means (for example, a field range changing unit) 43 is provided in the OCR of FIG. 1, and a display screen generating unit 19C having a different function is provided in place of the display screen generating unit 19A. is there.

The field range changing section 43 is provided with the input section 22
Has a function of scrolling the input image for display up, down, left, and right within a certain range based on the designation from. The display screen generation unit 19C generates screen information for displaying the field image and the recognition result in comparison with each other,
It has a function of additionally displaying a scroll bar or the like for controlling scrolling of the input image on the display screen. Other configurations are the same as those in FIG.

FIGS. 15A to 15C show the OCR of FIG.
FIGS. 4A and 4B are diagrams for explaining the operation of FIG. 4A, FIG. 4A is a diagram schematically showing an image stored in the image memory 12, and FIGS. 21 is a display screen before and after correction displayed in 21. Less than,
The operation of FIG. 14 will be described with reference to FIGS. 15 (a) to 15 (c).

The image of the form to be read is read by the scanner 11 and stored in the image memory 12 as shown in FIG. From the image stored in the image memory 12, a field image of the reference display range is cut out by the field detection unit 13, stored in the field memory 15 via the field range change unit 43, and given to the character cutout unit 16. Can be The character image in the field image is cut out character by character by the character cutout unit 16 and is sequentially recognized by the character recognition unit 17. A correctly recognized character image is converted to a character code. In addition, a character image that is not correctly recognized is rejected as an unreadable character, and "?"
And the character codes of a plurality of candidate characters corresponding to the unread character are output. These recognition results are stored in the recognition result memory 18.

When the recognition of all the characters in one field is completed, the field image stored in the field memory 15 and the recognition result stored in the recognition result memory 18 are read out by the display screen generation unit 19C, and are checked and corrected. Screen information is generated. Further, scroll bars for scroll control are added to the right and lower sides of the display area of the input image. The screen information generated by the display screen generation unit 19C is stored in the display memory 20, and is stored in the display memory 20 in FIG.
The image is displayed on the display unit 21 as shown in FIG.

The operator confirms the display range of the input image on the display screen of the display unit 21. FIG. 15 (b)
If the image is shifted to the lower left, as shown in
Drag the right scroll bar with the mouse to move the displayed image upward. Further, the display image is moved to the right by dragging the lower scroll bar with the mouse, and is moved to a position where characters in the input image can be grasped. Further, the user checks the input image corresponding to the unread character, selects an appropriate character from the plurality of presented candidate characters with a mouse, or inputs a correct character from the keyboard. As a result, the screen after the modification as shown in FIG. 15C is displayed on the display unit 21.

As described above, the OC of the ninth embodiment is
R has a field range changing unit 43 that changes the display range of the input image according to an operation from the input unit 22. Thus, for example, when a character entry position is shifted or read at an angle, it is possible to check the input character from the read image without having to check the original form. , And the recognition result can be quickly confirmed and corrected.

Note that the present invention is not limited to the above embodiment, and various modifications are possible. For example, there are the following modifications (a) to (d).

(A) Although reading of a form has been described, the present invention can be similarly applied to character recognition of general documents and the like.

(B) The field detection unit 13, character extraction unit 16, correction processing unit 30, etc. have been described as individual processing units for convenience of explanation, but actually operate according to a program having these functions. Computer.

(C) The display colors of the input image and the recognition result exemplified in the description are merely examples, and any color classification may be used as long as the image of the unread character and the recognition result can be easily identified.

(D) Color-coded display, additional display of character frame,
Although the display of the enlarged image, the scroll display of the input image, and the like have been described individually, a plurality of display methods can be arbitrarily combined.

[0088]

As described in detail above, according to the first aspect, the display means for displaying the image of the character field and the recognition result of the recognition means on the same screen and the input means are provided. Modifying means for changing the display format of the image displayed on the display means based on the modified data and modifying the recognition result of the recognizing means. As a result, the character image to be corrected can be easily specified, and confirmation and correction of the recognition result can be performed reliably and easily.

According to the second aspect, the display means in the first aspect displays a character image in a character field corresponding to a character that could not be correctly recognized by the recognition means in a different display from the character that could be correctly recognized. It is configured to display in color. Thus, the character image to be corrected can be immediately specified.

According to the third invention, the display means in the first or second invention is provided by adding a character frame to a character image in a character field corresponding to a character which cannot be correctly recognized by the recognition means. It is configured to display. Thus, the character image to be corrected can be immediately specified.

According to the fourth aspect of the present invention, the display means in the first or second aspect of the present invention is used to add a character frame to all the character images in the character field, and to add a character frame to each character and symbol of the recognition result. Is displayed. As a result, the position of the “blank” can be reliably specified.

According to the fifth aspect, the display means in the fourth aspect is characterized in that the character frame to be added to the character in the character field corresponding to the character that has not been correctly recognized by the recognition means can be correctly recognized. Is configured to be displayed in a display color different from the character frame to be added to. As a result, the character image to be corrected including the position of the “blank” can be immediately specified.

According to the sixth aspect, the correction means in the first to fifth aspects is configured to change the display color of the character image corrected by the correction data and the display color of the character as the recognition result. are doing. This makes it possible to distinguish between a corrected character and an uncorrected character, thereby preventing omission of correction and the like.

According to the seventh aspect, in addition to the image of the character field and the recognition result of the recognizing means, the display means in the first to sixth aspects can display a character image which cannot be correctly recognized by the recognizing means. It is configured to be enlarged and displayed on the same screen. Thus, the unread character can be accurately guessed and corrected.

According to the eighth invention, in addition to the image of the character field and the recognition result of the recognition means, the display means in the first to sixth inventions can display the character image which cannot be correctly recognized by the recognition means. The image of the periphery is enlarged and displayed on the same screen. This allows
The image around the unread character becomes clearer and can be corrected more accurately.

According to the ninth invention, in addition to the character field image and the recognition result of the recognition means, the display means in the first to sixth inventions can display character images which cannot be correctly recognized by the recognition means. The image of the character frame and its surroundings is enlarged and displayed on the same screen. As a result, the relationship between the unreadable character and the image around the character becomes clear, and the character can be corrected more accurately.

According to the tenth aspect, in the OCR according to the first to ninth aspects, the display range of the image displayed on the display means is changed within a certain area based on an instruction given from the input unit. A changing means is provided. Thus, when the image of the character field is shifted or inclined, the display range of the image can be corrected, and the recognition result can be confirmed and corrected based on the accurate image.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an OCR according to a first embodiment of the present invention.

FIG. 2 is a configuration diagram illustrating an example of a conventional OCR.

FIG. 3 is a diagram showing a display example of a recognition result in the OCR of FIG. 2;

FIG. 4 is a diagram showing a display example of a recognition result in the OCR of FIG. 1;

FIG. 5 is a diagram showing a display example according to the second embodiment of the present invention.

FIG. 6 is a configuration diagram of an OCR according to a third embodiment of the present invention.

FIG. 7 is a diagram showing a display example of a recognition result in the OCR of FIG. 6;

FIG. 8 is a diagram illustrating a display example according to a fourth embodiment of the present invention.

FIG. 9 is a diagram illustrating a display example according to a fifth embodiment of the present invention.

FIG. 10 is a configuration diagram of an OCR according to a sixth embodiment of the present invention.

11 is a diagram showing a display example of a recognition result in the OCR of FIG. 10;

FIG. 12 is a diagram illustrating a display example according to a seventh embodiment of the present invention.

FIG. 13 is a diagram showing a display example according to the eighth embodiment of the present invention.

FIG. 14 is a configuration diagram of an OCR according to a ninth embodiment of the present invention.

FIG. 15 is a diagram for explaining the operation of the OCR of FIG. 14;

[Explanation of symbols]

 REFERENCE SIGNS LIST 11 scanner 12 image memory 13 field detection unit 14 form format registration unit 15 field memory 16 character extraction unit 17 character recognition unit 18 recognition result memory 19A, 19B, 19C display screen generation unit 20 display memory 21 display unit 22 input unit 30 correction Processing unit 31 Correction control unit 32 Recognition result correction unit 33 Display screen correction unit 41 Character frame generation unit 42 Enlarged image generation unit 43 Field range change unit

Continued on the front page F term (reference) 5B064 AA01 AB09 BA01 EA05 EA10 FA05 FA10 FA11 5E501 AA30 BA07 CA03 CB02 CB09 CB13 EB19 FA13 FA42 FB44

Claims (10)

[Claims] A reading unit that optically reads a document to be read to generate an image image; a registration unit in which format information of the document is registered in advance; and a plurality of format information from the image image based on the format information. Detecting means for detecting a character field in which a character is described; recognizing a plurality of characters in the character field detected by the detecting means; and correctly recognizing a corresponding character code when each character can be correctly recognized. Recognizing means for outputting a symbol code indicating unread when not possible; display means for displaying the image of the character field and the recognition result of the recognizing means on the same screen; and correction data provided from the input unit While changing the display format of the image displayed on the display means based on the
Correction means for correcting the recognition result of the recognition means. 2. The display device according to claim 1, wherein the display unit displays a character image in the character field corresponding to a character that has not been correctly recognized by the recognition unit in a display color different from that of the correctly recognized character. The character recognition device according to claim 1, wherein: 3. The apparatus according to claim 2, wherein said display means displays a character image in the character field corresponding to a character which has not been correctly recognized by said recognition means, with a character frame added thereto. Item 3. The character recognition device according to item 1 or 2. 4. The apparatus according to claim 1, wherein said display means is configured to add a character frame to all the character images in said character field and to add a character frame to each character and symbol of said recognition result. The character recognition device according to claim 1 or 2, wherein 5. The display means for displaying a character frame to be added to a character in the character field corresponding to a character that has not been correctly recognized by the recognition means differently from a character frame to be added to a character that has been correctly recognized. 5. The character recognition device according to claim 4, wherein the character is displayed in a color. 6. The apparatus according to claim 1, wherein the correction unit changes a display color of a character image corrected by the correction data and a display color of a character as a recognition result. The character recognition device described in any one of the above. 7. The display means is configured to enlarge, in addition to the image of the character field and the recognition result of the recognizing means, a character image that cannot be correctly recognized by the recognizing means and display the enlarged character image on the same screen. The character recognition device according to any one of claims 1 to 6, wherein: 8. The display means enlarges, on the same screen, a character image that cannot be correctly recognized by the recognition means and an image of a peripheral part thereof, in addition to the image of the character field and the recognition result of the recognition means. The character recognition device according to any one of claims 1 to 6, wherein the character recognition device is configured to display the information. 9. The display device according to claim 1, wherein the display unit enlarges, in addition to the image of the character field and the result of the recognition by the recognition unit, a character image, a character frame, and an image around the character frame that cannot be correctly recognized by the recognition unit. 7. The display according to claim 1, wherein the display is performed on the same screen.
The character recognition device described in any one of the above. 10. A changing means for changing a display range of an image displayed on the display means within a predetermined area based on an instruction given from the input unit. The character recognition device described in any one of the above.