JPH09218922A - Character recognition device and method therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate operations such as the connection and separation of character segmentation frames or the like by instructing the correction of the character segmentation frame to a character image which is the object of character recognition. SOLUTION: When a text string to be corrected is specified from the text strings by character recognized results, the specified text string (shown by Figure) is displayed in a text window 3b and partial images corresponding to the text string are displayed in an image window 3a. In the image window 3a, the character segmentation frames 31, 32 and 33 set in the character recognition for the character images in the partial images are displayed together. In the state, when it is instructed to connect the character segmentation frames 31, 32 and 33, the character images inside the character frames 31, 32 and 33 are segmented as one character and recognized as the Chinese character shown by Figure by performing character recognition to it.

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 and method thereof. In particular, the present invention relates to a character recognition device and method capable of performing a correction operation for erroneous recognition due to a character cutout error.

[0002]

2. Description of the Related Art Conventionally, character recognition application software (hereinafter referred to as character recognition software) cuts out characters as preprocessing and performs character recognition on the cut out image. Therefore, if the character is cut out incorrectly, the recognition result is also wrong.

In the above case, general character recognition software corrects the resulting text. Further, in some character recognition software, when one character is separated into two or more by cutting out the character, it is attempted to correct the character cutting frame by an operation of connecting the text sides.

[0004]

As described above, the method of correcting the character cutting frame on the text side is a combination when a character cutting error separates one character into two or more. It is limited to operations. That is, when two or more characters or ruled lines are included in one character, the text cannot be separated. For example, when trying to separate the character "work" into two or more, a combination of "a" and "motion" or "i" and "heavy"
There is a combination of and "force", but if you try to modify the character cutting frame on the text side, you cannot specify how to cut out the character.

The present invention has been made in view of the above problems, and makes it possible to instruct correction of a character cutting frame for a character image which is a target of character recognition, and to combine the character cutting frames, An object of the present invention is to provide a character recognition device and a method thereof that can easily separate the characters.

[0006]

A character recognition apparatus according to the present invention for achieving the above object has the following configuration.
That is, a first storage means for storing the original image data, a recognition means for performing character recognition by cutting out a character from the original image data, and a second storage means for storing the recognition result obtained by the recognition means. , A portion of the original image data corresponding to a text string to be corrected in the recognition result is extracted and displayed as a partial image, and a character cutout is changed for the partial image displayed by the display portion. And a re-recognition unit for recognizing the character of the partial image based on the cut-out of the character changed by the changing unit.

[0007] Preferably, the display means further comprises a third storage means for storing frame information indicating a position and a size of a character cutting frame obtained by cutting out the character by the recognition means. An area including a character cutting frame indicated by frame information corresponding to each character of the text string is extracted from the original image data and displayed as a partial image. The recognition result and the original image data are associated with each other by using the character cutting frame obtained by the character recognition processing by the recognition means. Therefore, the original image data can be extracted at an appropriate position and range for the text string to be corrected, and a partial image suitable for correcting the character cutting frame can be generated.

[0008] Further, preferably, the display means adds and displays a character cutting frame indicated by the frame information when displaying the partial image. This is because the character cutting frame obtained by the recognition processing can be grasped and the operability is improved.

Further, preferably, the changing means includes a specifying means for specifying a plurality of character cutting frames in the partial image,
A combining unit that combines a plurality of designated character cutting frames into one character cutting frame. This is because desired character cutting frames in the partial images can be specified and combined, and operability is improved.

Further, preferably, the changing means is designated by the designating means for designating a dividing position for dividing one character cutting frame in the partial image into a plurality of character cutting frames, and the designating means. Separation means for separating the one character cutting frame into a plurality of character cutting frames at a dividing position. This is because the desired character cutting frame in the partial image can be separated into the desired character cutting frame.

Further, preferably, the display means displays the partial image and the text string to be corrected.

[0012] Preferably, it further comprises display updating means for updating the text string to be corrected displayed by the displaying means with the result recognized by the re-recognizing means.

Further, it is preferable that a designation means for designating the text string to be modified by designating a desired text string from the text string display based on the recognition result stored in the second storage means is further provided. Prepare This is because the text string to be modified can be easily specified.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the arrangement of the character recognition apparatus according to this embodiment. In the figure, 1 is a CPU
And executes various processes in the character recognition device. Reference numeral 2 is an interface for communicating with the scanner 21. Reference numeral 21 is a scanner for optically reading an image written on a document or the like. Original image data obtained by reading an original image with the scanner 21 is supplied to the bus 7 via the interface 21.

Reference numeral 3 denotes a display unit, which displays various displays under the control of the CPU 1. An external storage device 4 is a recognition processing program 41 for performing character recognition on original image data.
Alternatively, the recognition dictionary 42 referred to in the recognition processing is stored. These recognition processing program 41 and recognition dictionary 42
Are read out to the RAM 5 as necessary and executed by the CPU 1.

Reference numeral 5 denotes a RAM, which provides a main memory of the CPU 1 and a work area. The RAM 5 also stores original image data 51, cutout data 52, and recognition data 53. The original image data 51 is original image data supplied from the scanner 21 via the interface 2. Further, the character is cut out from the original image data by executing the recognition processing program 41, and the information of the cutout frame at this time (hereinafter referred to as frame data) is stored as the cutout data 52. Further, the result of performing the recognition process according to the result of cutting out the character is held as the recognition result data 53.

Reference numeral 6 denotes a ROM, which is a boot program and
Stores various data. Reference numeral 7 is a bus for communicatively connecting the above-mentioned components. An input unit 8 is composed of a keyboard and a pointing device.

FIG. 2 is a diagram for explaining the respective data structures of the original image data 51, the cutout data 52, and the recognition data 53 stored in the RAM 5. In the figure, an example of the original image data 51 obtained as a result of reading the original image with the scanner 21 is shown. By cutting out characters from the original image data 51 and executing recognition processing on each of the cut out characters, cutout data 52 and recognition result data 53 are obtained.

As shown in FIG. 2, the cut-out data 5
Reference numeral 2 is composed of frame data indicating the position and size in the original image data of each character cutting frame obtained as a result of performing the character cutting process on the original image data 51. The recognition result data 53 is composed of a text code obtained by recognizing the character image in each character cutting frame. Each text code of the recognition result data 53 corresponds to each frame data of the cutout data 52.

The operation of correcting the recognition result in the character recognition apparatus of this embodiment having the above-mentioned configuration will be described.
The case where the original image data 51, the cutout data 52, and the recognition result data 53 as shown in FIG. 2 are obtained as a result of the character recognition processing will be described.

FIG. 3 is a flowchart showing a processing procedure at the time of correction processing according to this embodiment. It is assumed that the text display according to the recognition result data 53 is performed on the display unit 3 before this processing is executed. Therefore, in this example, as shown in FIG. 4, a text string “Motion is moving here and gravity ...” is displayed. Here, a character image that should be recognized as "working" is recognized as "a gravity" and a character image that should be recognized as "gravity" is recognized as "moving" due to a cutout error in the character cutting frame.

First, in step S11, the display unit 3
Specify the text string to be modified from the text string displayed in. Here, the text string is composed of one or a plurality of texts. The text string is specified using the input unit 8 such as a pointing device.

Subsequently, in step S12, frame data corresponding to each text of the designated text string is obtained from the cut-out data 52. For example, when "a gravity" in the text display of FIG. 4 is selected, it is detected that the text codes 531, 532, 533 of FIG. 2 are selected, and the corresponding frame data 521, 522, 523 are displayed. Be won.

Next, in step S13, step S1
Image data in a range including the frame data acquired in 2 is extracted from the original image data, and a partial image is generated by the extracted image data. Here, the area including the character cutting frame indicated by the frame data 521 to 523 is extracted as the partial image from the original image data 51.

In step S14, a text window is generated and displayed on the display unit 3, and in this step, step S11.
Display the text string to be modified specified by. Also,
In step S15, an image window is generated and displayed on the display unit 3, and the partial image generated in step S13 is displayed therein. In addition, in the partial image, the frame data 521 to
A character cutting frame is also displayed based on 523.

FIG. 5 is a diagram showing a display state during the combining process according to this embodiment. Here, the text display according to the recognition result data 53 in FIG. 2 (“the movement is gravitational ...”), the text window 3b, and the image window 3 are displayed.
The display state of a is shown. Text window 3b
And each display of the image window 3a is performed in the above step S
14 and S15. Here, the frames 31, 32, and 33 in the image window 3a are the character cutting frames used for recognition. Further, "a gravity" in the text string is underlined to indicate that it is being designated as a correction processing target by the designation operation in step S11.

In the above display state, "combine"
Alternatively, the combining process or the separating process is executed by instructing the correction process content of "separating" (steps S16 to S19).

FIG. 6 is a flow chart for explaining the combining process according to this embodiment. In step S31, the character cutting frame to be combined is selected in the image window 3a. Here, the character cutting frames 31 to 31 of the image window 3a
It is assumed that 33 is selected. Next, in step S32, when the combining is instructed, the combining process of the selected character cutting frames is performed, and a new character cutting frame is set.
The instruction of “coupling” may be performed using a menu, a tool button, or the like. Then, in step S33, one character is recognized for the character cutting frame generated in step S32, and the obtained result is displayed in the text window 3b.

FIG. 7 is a diagram showing a state in which the character cutting frames of FIG. 5 are combined to generate a new character cutting frame 20.
Character cutting frame 2 obtained by combining the character cutting frames 31 to 33
By performing the character recognition on the character image within 0, the three-character text string "a gravity" is recognized as one character "work", and the result is displayed in the text window 3b.

The correction result is the cut-out data 5
2. Needless to say, it is reflected in the recognition result data 53. And as a result, the displayed text string is
"The movement works here ..."

Next, a separation process for correcting a portion cut out by one character cutting frame from a plurality of character cutting frames will be described. FIG. 8 is a diagram showing a display state during the separation processing according to this embodiment.

An example will be described in which the text string "movement" in the text string "movement works here ..." that is modified by the above-described combining processing is designated. When a text string to be modified (here, "movement") is designated, the processing in steps S11 to S15 described above causes the text shown in FIG.
Is obtained. The text "moving" is displayed in the text window 3b, and the image window 3a is displayed.
Shows a partial image corresponding to this.

When the separation process is instructed in such a state, the separation process shown in FIG. 9 is executed (step S1).
7, S19). FIG. 9 is a flowchart showing the procedure of the separation process in this embodiment.

First, in step S71, a character cutting frame to be separated is designated by a pointing device or the like. In the example of FIG. 8, the character cutting frame 40 is designated. Next, in step S71, when "separation" is instructed, the display is as shown in FIG. FIG. 10 is a diagram showing an image window and a text window in the middle of a separating operation. Four
Reference numeral 1 is a separation line for determining the separation position in the character cutting frame 40, which is moved by the mouse cursor. The “separation” instruction may be given by operating a button on the menu or toolbar.

Next, in step S73, the separation line 4
The separation position is determined by moving 0 with a mouse cursor or the like. In this example, a separation line is determined between “heavy” and “mosquito” of “movement”. In step S74, when the confirmation of the separation position is instructed, the character cutting frame is separated. Character cutting frame 40
When the position of the separation line 41 with respect to is determined as shown in FIG. 10 and the confirmation of the separation position is instructed, the character cutting frame 40 is separated,
It is displayed as shown in FIG. Note that FIG. 11 is a diagram showing a display state of the image window and the text window after the character cutting frame is separated and recognized. Where 61 and 62
Is a new character cutting frame obtained by separating the character cutting frame 40 by a separating line 41. As shown in FIG. 11, with respect to the character cutting frame 40, a new character cutting frame 61 after separation,
62 is generated and displayed. The separation position may be confirmed by clicking a mouse button or the like.

Then, in step S75, as shown in FIG.
One character is recognized for the character image in each of the character cutting frames 61 and 62, correct text is obtained, and the result is displayed in the text window 3b. That is, as shown in FIG. 11, the character misidentified as "movement" is "gravity".
Is corrected and displayed in the text window 3b.

As described above, according to the present embodiment, by correcting the character cutting frame in the image before recognition, it is possible not only to combine the separated characters but also to separate the contact characters. The efficiency of the character recognition operation can be improved.

Although the embodiment on Windows has been described, OS / 2 (trademark), Mac (trademark), UNIX
Any OS can be used for displaying windows such as (trademark).

Further, in the present embodiment, only the text string to be corrected is designated and displayed in the image window 3a and the text window 3b, but these may include text that is not the correction target. For example, in the display of FIG. 4, when "a gravity" is the correction target, if "a gravity" is specified when the text string is specified, a partial image including these is displayed in the image window 3a. .
However, when the character cutting frames to be combined are specified in step S32, if only the character cutting frames of "a", "heavy", and "force" are designated, the character "ki" has no effect. Obviously, the cutting frames can be joined.

In the above embodiment, the case where one character separation position is designated and one character cutting frame is separated into two has been described, but the number of separations is not limited to this. For example, if the character "work" is to be separated into three, i, heavy, and force, two separation lines will be used. Furthermore, for example, when the character "country" is separated into a "ball" and a square ruled line "□" that surrounds the character, a rectangle is formed by a separation line to surround the character "ball". The configuration for realizing the formation of such a plurality of separation lines or a rectangle by the separation lines is
It will be apparent to those skilled in the art from the above description of the embodiments.

Even if the present invention is applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), an apparatus (for example, a copying machine, a facsimile, etc.) comprising one device Device).

Another object of the present invention is to provide a storage medium storing a program code of software for realizing the functions of the above-described embodiments to a system or apparatus, and to provide a computer (or CPU) of the system or apparatus.
And MPU) read and execute the program code stored in the storage medium.

In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the program code constitutes the present invention.

As a storage medium for supplying the program code, for example, a floppy disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD
-R, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

When the computer executes the readout program code, not only the functions of the above-described embodiment are realized, but also the OS (Operating System) running on the computer based on the instruction of the program code. ) May perform some or all of the actual processing, and the processing may realize the functions of the above-described embodiments.

Further, after the program code read from the storage medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, based on the instructions of the program code, It goes without saying that the CPU included in the function expansion board or the function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

When the present invention is applied to the above-mentioned storage medium, the storage medium stores the program code corresponding to the above-mentioned flow chart. Briefly, in the memory map example of FIG. Each module shown will be stored in the storage medium.

That is, at least "first storage processing module""recognition processing module""second storage processing module""display processing module""change processing module".
The program code of each module of the “re-recognition processing module” may be stored in the storage medium.

Here, the first storage processing module realizes a first storage processing for storing the original image data in the first memory. Further, the recognition processing module realizes a recognition processing of cutting out characters from the original image data and performing character recognition. Further, the second storage processing module realizes a second storage processing of storing the recognition result obtained by the recognition processing in the second memory. In addition, the display processing module
A display process of extracting a portion of the original image data corresponding to a text string to be corrected in the recognition result and displaying it as a partial image is realized. In addition, the change processing module
A change process for changing the cutout of characters is realized for the partial image displayed in the display process. Further, the re-recognition process realizes a re-recognition process for recognizing the character of the partial image based on the cut-out of the character changed in the changing process.

[0051]

As described above, according to the present invention,
It is possible to instruct the correction of the character cutting frame for the character image that is the object of character recognition, and it is possible to easily combine and separate the character cutting frames.

[0052]

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a character recognition device according to the present embodiment.

FIG. 2 is a diagram illustrating a data configuration of original image data 51, cutout data 52, and recognition data 53 stored in a RAM 5.

FIG. 3 is a flowchart showing a processing procedure at the time of correction processing according to the present embodiment.

FIG. 4 is a diagram showing a state in which a text string is displayed according to the recognition data of FIG.

FIG. 5 is a diagram showing a display state during a combining process according to the present embodiment.

FIG. 6 is a flowchart illustrating a combining process according to this embodiment.

FIG. 7 is a diagram showing a state where a new character cutting frame 20 is generated by combining the character cutting frames of FIG.

FIG. 8 is a diagram showing a display state during separation processing according to the present embodiment.

FIG. 9 is a flowchart showing a procedure of separation processing in the present embodiment.

FIG. 10 is a diagram showing an image window and a text window in the middle of a separating operation.

FIG. 11 is a diagram showing a display state of an image window and a text window after separating and recognizing a character cutting frame.

FIG. 12 is a diagram showing an example of a memory map of a storage medium storing a control program according to the present invention.

[Explanation of symbols]

 1 CPU 2 Interface 3 Display Unit 4 External Storage Device 5 RAM 6 ROM 7 Bus 8 Input Unit 21 Scanner 41 Recognition Processing Program 42 Recognition Dictionary 51 Original Image Data 52 Cutout Data 53 Recognition Result Data

Claims (17)

[Claims] 1. A first storage means for storing original image data, a recognition means for performing character recognition by cutting out characters from the original image data, and a second storage means for storing a recognition result obtained by the recognition means.
A storage unit, a display unit for extracting a portion of the original image data corresponding to a text string to be corrected in the recognition result and displaying it as a partial image, and a character string for the partial image displayed by the display unit. Change means for changing the cutout, based on the cutout of the character changed by the changing means,
A character recognition device, comprising: a re-recognition unit that performs character recognition of the partial image. 2. A third storage means for storing frame information indicating a position and a size of a character cutting frame obtained by cutting out a character by the recognition means, wherein the display means stores the text string to be corrected. The character recognition device according to claim 1, wherein an area including a character cutting frame indicated by frame information corresponding to each character is extracted from the original image data and is displayed as a partial image. 3. The character recognition device according to claim 2, wherein the display unit adds and displays a character cutting frame indicated by the frame information when displaying the partial image. 4. The changing unit includes a specifying unit that specifies a plurality of character cutting frames in the partial image, and a combining unit that combines the specified plurality of character cutting frames into one character cutting frame. The character recognition device according to claim 2. 5. The changing unit selects a dividing position for dividing one character cutting frame in the partial image into a plurality of character cutting frames, and a dividing position specified by the specifying unit. 3. A separating unit that separates the one character cutting frame into a plurality of character cutting frames.
The character recognition device according to 1. 6. The character recognition device according to claim 1, wherein the display unit displays the partial image and the text string to be corrected. 7. The character recognition device according to claim 6, further comprising display updating means for updating the text string to be corrected displayed by the display means with a result recognized by the re-recognition means. . 8. The display device further comprises designation means for designating the text string to be corrected by designating a desired text string from among the text strings displayed based on the recognition result stored in the second storage means. The character recognition device according to claim 1, wherein the character recognition device is a character recognition device. 9. A first storage step of storing original image data in a first memory; a recognition step of slicing characters from the original image data to perform character recognition; and a recognition result obtained by the recognition step. A second storage step of storing in a second memory; a display step of extracting a portion of the original image data corresponding to a text string to be corrected in the recognition result and displaying it as a partial image; A changing step of changing the cutout of the character for the partial image that has been made, and based on the cutout of the character changed in the changing step,
And a re-recognition step of recognizing the character of the partial image. 10. A third storage step of storing, in a third memory, frame information indicating a position and a size of a character cutting frame obtained by cutting out a character in the recognition step, the display step including the correction 10. The character recognition method according to claim 9, wherein a region including a character cutting frame indicated by frame information corresponding to each character of the power sequence is extracted from the original image data and displayed as a partial image. . 11. The character recognition method according to claim 10, wherein, in the displaying step, a character cutting frame indicated by the frame information is added and displayed when the partial image is displayed. 12. The changing step includes a specifying step of specifying a plurality of character cutting frames in the partial image, and a combining step of combining the specified plurality of character cutting frames into one character cutting frame. The character recognition method according to claim 10, wherein: 13. The changing step is performed in 1 of the partial images.
A step of designating a dividing position for dividing one character cutting frame into a plurality of character cutting frames, and a separation for separating the one character cutting frame into a plurality of character cutting frames at the dividing position specified in the designating step. The character recognition method according to claim 10, further comprising a step. 14. The character recognition method according to claim 9, wherein the displaying step displays the partial image and the text string to be corrected. 15. The character recognition method according to claim 14, further comprising a display update step of updating the text string to be corrected displayed in the display step with the result recognized in the re-recognition step. . 16. The method further comprises a designating step of designating the text string to be corrected by designating a desired text string from the text string display based on the recognition result stored in the second memory. The character recognition method according to claim 9. 17. A computer-readable memory that stores a program code for performing character recognition processing, the code of a first storing step of storing original image data in a first memory, and cutting out characters from the original image data. A code of a recognition step for performing character recognition, a code of a second storage step of storing the recognition result obtained by the recognition step in a second memory, and a correction of the original image data in the recognition result. The code of the display step of extracting the part corresponding to the text string to be displayed and displaying it as a partial image, the code of the changing step of changing the cutout of the characters for the partial image displayed in the display step, and the change of the changing step Based on the cut out characters
And a code for a re-recognition process for recognizing characters of the partial image, the computer-readable memory.