JPH11250177A - Document reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a document reader capable of easily executing correction processing for a recognition result. SOLUTION: The input image data S1 of a document P are sent from a scanner part 1 to an image input part 22. The data S1 are entered into the input part 22 and stored in a memory 40. An image data recognition part 23 executes character recognition processing for the data S1 stored in the memory 40 and outputs a recognition result. The recognition result is sent to a monitor 51 through a picture display part 26 and displayed on the monitor 51. An operator displays the data S1 on the monitor 51 as a guide for correcting the recognition result. In this case, a character frame extraction part 25 extracts a character frame from the data S1 and a character frame conversion part 25 deletes the character frame. Then an image obtained by removing the character frame from the data S1 is displayed on the monitor 51. The operator executes correction processing by collating each character displayed on the monitor 51 with the recognition result.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention recognizes a document image of a form in which characters are described in a character frame such as a slip of a predetermined format, and makes it possible to easily correct the recognition result. The present invention relates to a document reading apparatus described above.

[0002]

2. Description of the Related Art FIG. 2 is a diagram showing an example of a form document in which characters are described in a character frame. In this figure, characters (for example, “0, 1, 2,...”) Are described in a character frame F, such as a slip in a predetermined format. FIG. 3 is a diagram showing an example of a display screen of the display unit when the document of FIG. 2 is read by a conventional document reading apparatus. In this figure, an input image V1 when the document shown in FIG. 2 is read by a conventional document reading apparatus and a recognition result R1 of characters in the input image V1 are displayed.

In a conventional document reading apparatus, input image data of the document shown in FIG. 2 is taken into an image input section, and a character pattern in the input image data is recognized by an image data recognition section. The character frame F in the input image data is extracted by the character frame extracting unit. The recognition result of the character, the character frame F, and the character in the input image data is displayed on the display unit. In this display section, as shown in FIG. 3, a character frame F is displayed together with each character in the input image V1. In the recognition result R1, a cursor C is displayed at a position of a character to be corrected. The operator compares the character in the input image V1 with the character in the recognition result R1, and in the case of a mismatch, operates the input unit to move the cursor C to the position of the mismatched character and inputs a correction signal. . When a correction signal is input, a correction process is performed so that a mismatched character matches a character in the input image V1.

[0004]

However, in the conventional document reading apparatus, when correcting the recognition result, as shown in FIG. 3, a character frame F is put together with each character in the input image V1 on the display screen. Will be displayed. Therefore, when the operator checks the characters in the input image V1 and the characters in the recognition result R1, the input image V1 is very difficult to see, and there is a problem that the efficiency of the correction work is reduced.

[0005]

According to a first aspect of the present invention, there is provided a document reading apparatus for converting input image data of a document in which characters are described in a character frame. An image input unit to be captured, and an image to cut out a character area surrounding each of one or more character patterns included in the input image data captured by the image input unit, recognize the character pattern, and output a recognition result Data recognizing means, character frame extracting means for extracting the character frame from the input image data captured by the image input means, and all or any part of the extracted character frame having a color different from the character frame A character frame conversion / deletion means for converting or deleting the part to output a conversion / deletion result, and a display for displaying characters in the input image data, conversion / deletion result, and recognition result hand When the character in the input image data displayed on the display means and the recognition result do not match, a correction input means for inputting a correction signal based on an operation of an operator, and when the correction signal is input, And control means for performing a correction process for matching the recognition result displayed on the display means with characters in the input image data.

By adopting such a configuration, input image data of a document in which a character is described in a character frame is taken into the image input means. A character area is cut out from the input image data by an image data recognizing means, a character pattern is recognized, and a recognition result is output. Further, a character frame is extracted from the input image data captured by the image input means by a character frame extracting means. The extracted character frame is converted by the character frame conversion / deletion means into a collation character frame of a color different from that of the character frame in its entirety or an arbitrary part, or the part is deleted. -The conversion / deletion result is output from the deletion means. The characters in the input image data, the conversion / deletion result, and the recognition result are displayed on a display unit. The operator compares the character in the input image data displayed on the display means with the recognition result, and when they do not match, operates the correction input means to input a correction signal to the control means. When the correction signal is input, the control unit performs a correction process so that the recognition result displayed on the display unit matches a character in the input image data.

In the invention according to claim 2, image input means for inputting input image data of a document in which characters are described in a character frame, first storage means for storing the input image data, Image data recognizing means for extracting a character area surrounding one or more character patterns included in the input image data stored in the storage means, recognizing the character pattern and outputting a recognition result, and recognizing the recognition result. Second storage means for storing; character frame extraction means for extracting the character frame from the input image data stored in the first storage means; and all or any part of the extracted character frame being represented by the character Convert to a collation character frame of a color different from the frame, or delete the part and output the conversion / deletion result.
Deletion means, display means for displaying characters in the input image data stored in the first storage means, conversion / deletion results, and recognition results stored in the second storage means, and display by the display means A correction input unit for inputting a correction signal based on an operation of an operator when the character in the input image data does not match the recognition result, and a recognition unit displayed on the display unit when the correction signal is input. And control means for performing a correction process for matching the result with characters in the input image data.

By adopting such a configuration, input image data of a document in which characters are described in a character frame is taken into the image input means. This input image data is stored in the first storage means. From the input image data stored in the first storage means, a character area is cut out by the image data recognition means, the character pattern is recognized, and the recognition result is output. This recognition result is stored in the second storage means. A character frame is extracted from the input image data stored in the first storage means by the character frame extraction means. The extracted character frame is converted by the character frame conversion / deletion means into a collation character frame of a color different from that of the character frame in its entirety or an arbitrary part, or the part is deleted. -The conversion / deletion result is output from the deletion means. Characters in the input image data stored in the first storage means, conversion / deletion results, and recognition results stored in the second storage means are displayed on the display means. The operator compares the character in the input image data displayed on the display means with the recognition result, and when they do not match, operates the correction input means to input a correction signal to the control means. When the correction signal is input, the control unit performs a correction process so that the recognition result displayed on the display unit matches a character in the input image data. According to a third aspect of the present invention, the image input means of the first or second aspect of the present invention is configured to take in input image data from a scanner that scans a form in which characters are described in a character frame.

By adopting such a configuration, the input image data is output from the scanner and taken into the image input means, and the same processing as in the invention according to claim 1 or 2 is performed. According to a fourth aspect of the present invention, the image input means according to the first or second aspect of the present invention includes a facsimile apparatus (hereinafter, referred to as a facsimile) connected to a communication network for writing characters in a character frame via the communication network. It is configured to take in input image data of a written document. By adopting such a configuration, input image data is output from a facsimile and taken into an image input unit via a communication network. After that, the same processing as in the invention according to claim 1 or 2 is performed.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First Embodiment FIG. 1 is a block diagram of a document reading apparatus according to a first embodiment of the present invention. This document reading device includes a scanner unit 1. The scanner unit 1 includes, for example, a charge-coupled device (C
CD) a photoelectric conversion unit including a sensor, an analog / digital conversion circuit, and the like, irradiates and scans a form P on which characters are described in a character frame, and uses the reflected light to input image data S1 of an electric signal. It has the function of converting to and outputting.
The output side of the scanner unit 1 is connected to a reading apparatus main body 10 that analyzes input image data S1 and reads characters. The reading device main body 10 includes a processor 20 for performing a recognition process on the input image data S1,
0 and a first storage means (for example, a memory) 40 connected via a data bus 30.

The processor 20 includes a control unit 21 for controlling the entire processor 20 according to a document reading program,
Image input means (for example, an image input unit) 22 which takes in the input image data S1 and sends it to the image storage area of the memory 40 via the data bus 30, and is included in the input image data S1 taken in by the image input unit 22 A character area surrounding the character pattern to be extracted is cut out, and image data recognizing means (for example, an image data recognizing unit) for recognizing the character pattern and sending the coded recognition result to the recognition result storage area of the memory 40 via the data bus 30 ) 23. Further, the processor 20 includes a character frame extracting unit (for example, a character frame extracting unit) 24 for extracting a character frame from the input image data S1 captured by the image input unit 22, and a character frame extracting unit 24. A character frame conversion / deletion unit (for example, a character frame conversion unit) 25 that deletes all or an arbitrary part of the character frame and outputs a conversion / deletion result, and a character in the input image data S1 and a character frame conversion unit 25 A display unit (for example, an image display unit) 2 for generating a display signal for displaying the output conversion / deletion result and the recognition result output from the image data recognition unit 23
6 are provided.

A monitor 51 for displaying an image based on the display signal generated by the image display unit 26 to the processor 20 and the memory 40 via the data bus 30, and an input signal is sent to the control unit 21 by an operator's operation. Mouse 52 and keyboard 53, and a floppy disk 5 for storing the recognition results stored in the recognition result storage area of the memory 40.
4. A hard disk 55 for storing a program for operating the control unit 26 is connected.

Next, a reading operation (1) and a correction processing operation (2) in the document reading apparatus of FIG. 1 will be described. (1) Reading Operation FIG. 4 describes a reading operation in the document reading apparatus of FIG. It is a flowchart for the.

In the reading operation, first, the form P is irradiated with light by the scanner unit 1, an image of the form P is read from the reflected light, converted into input image data S 1, and
0 (step ST1). In the processor 20, the input image data S1 is taken into the image input unit 22, and stored in the image storage area of the memory 40 (step ST2). Input image data S1 stored in memory 40
Indicates that a character area is cut out by the image data recognizing unit 23, character recognition processing is performed, and a recognition result is output (step S).
T3). After the end of the character recognition process, a correction process for the recognition result is performed (step ST4). After the correction processing is completed, it is determined whether the reading of the image of the form P is completed. If the reading is completed, the processing operation in the document reading apparatus is completed.
Return to 1.

(2) Correction processing operation FIG. 5 shows the recognition result correction processing operation (step ST) in FIG.
It is a flowchart for demonstrating 4). FIG.
FIG. 3 is a diagram illustrating an example of a display screen of a monitor 51 in FIG. 1, showing an input image V2 when the document in FIG. 2 is captured by the document reading device in FIG. 1, and a recognition result R2 of characters in the input image V2; It is shown.

In the correction processing operation, in FIG. 5, the recognition result output from the image data
6 and a display signal is generated by the image display unit 26. This display signal is sent to the monitor 51, and the recognition result is displayed (step ST11). The operator operates the input image data S1 as a guide for correcting the recognition result.
Is displayed on the monitor 51. Alternatively, it is set in advance so that the input image data S1 is automatically displayed on the monitor 51 (step ST12). When displaying the input image data S1 on the monitor 51, the input image data S1
Extract a character frame from 1. In this case, the character frame extracting unit 24
In, for example, a histogram of the number of pixels in the vertical direction and the horizontal direction of the input image data S1 is calculated, and when the histogram is equal to or larger than a preset threshold, it is extracted as a character frame (step ST13). Character frame conversion unit 25
In, the character frame F in the input image data S1 is deleted. In this case, for example, the image data of the extracted character frame is not sent to the monitor 51 (step ST1).
4). As shown in FIG. 6, only each character is displayed in the input image V2, and no character frame is displayed. Recognition result R
During cursor 2, place the cursor C at the position of the character to be corrected.
Is displayed.

An image in which the character frame has been removed from the input image data S1 (ie, the characters in the input image data S1) is displayed on the monitor 51 (step ST15). The operator compares the character in the input image V2 displayed on the monitor 51 with the recognition result R2, and inputs a correction signal to the control unit 21 by operating the mouse 52 or the keyboard 53 if they do not match. When the correction signal is input, the control unit 21 performs a correction process so that the recognition result R2 matches a character in the input image V2 (step ST16). The result of the correction process is sent to the monitor 51 via the image display unit 26. The operator determines whether or not the correction process has been completed by looking at the display screen of the monitor 51 (step ST17).
Returning to step S6, if the processing has been completed, the processing moves to step S5 in FIG. In step ST12, the input image data S
If 1 is not displayed on the monitor 51, step ST16
Move to As described above, in the first embodiment, the character frame conversion unit 25 converts the input image data S1 into the character frame F.
Is deleted and the input image V2 without the character frame F is displayed on the monitor 51, so that the burden on the operator in the correction work can be reduced.

Second Embodiment In the document reading apparatus of this embodiment, a character frame conversion unit 2 having a different configuration (not shown) is used instead of the character frame conversion unit 25 shown in FIG.
5A is provided. The character frame conversion unit 25A is configured to convert all or any part of the character frame extracted by the character frame extraction unit 24 into a collation character frame of a color different from the character frame and output a conversion / deletion result. It has become.

FIG. 7 is a flowchart of a correction processing operation according to the second embodiment of the present invention. Elements common to those in FIG. 5 according to the first embodiment are denoted by the same reference numerals. I have. FIG. 8 is a diagram showing an example of a display screen of the monitor 51 of FIG. In this document reading apparatus, the reading operation is performed in the same manner as in the first embodiment. In the correction processing operation, as shown in FIG. 7, in step ST14A, in the character frame conversion unit 25A, all or any part of the character frame extracted by the character frame extraction unit 24 is converted to a color conversion unit such as a color palette. Is converted to a collation character frame of a different color from the character frame. The image data of the collation character frame is transmitted to the monitor 51 via the image display unit 26. As shown in FIG. 8, in the input image V3, a collation character frame FA whose color has been changed is displayed together with each character. Recognition result R
In 3, the cursor C is displayed at the position of the character to be corrected. Otherwise, the same processing as in FIG. 5 is performed.

As described above, in the second embodiment,
Since the color of the character frame in the input image data S1 is changed in the character frame conversion unit 25A and the input image V3 including the character frame is displayed on the monitor 51, it is easier to see than the input image V2 in FIG. Thus, the burden on the operator in the correction work can be reduced.

Third Embodiment In the document reading apparatus of this embodiment, a character frame conversion unit 2 having a different configuration (not shown) is used instead of the character frame conversion unit 25 shown in FIG.
5B is provided. The character frame conversion unit 25B converts all or any part of the character frame extracted by the character frame extraction unit 24 into a collation character frame of a color different from the character frame, deletes the part, Alternatively, whether to output without changing is selected by an operation of the operator.

FIG. 9 is a flowchart of a correction processing operation according to the third embodiment of the present invention. Elements common to those in FIGS. 5 and 7 are denoted by the same reference numerals. In this document reading apparatus, the reading operation is performed in the same manner as in the first embodiment. In the correction processing operation, as shown in FIG. 9, in step ST21, it is determined whether or not the character frame is hidden by the operator. The frame is deleted. Thereafter, the process proceeds to step ST15. When displaying the character frame, the process proceeds to step ST22, and it is determined whether or not to change the color of the character frame. When changing the color of the character frame, the process proceeds to step ST22a, where the character frame conversion unit 25
B changes the color of the character frame. Then, step ST1
Go to 5. Step ST when not changing the color of the character frame
Proceeding to 23, the character frame is displayed as it is. The other is FIG.
Is performed.

As described above, in the third embodiment,
Since the character frame conversion unit 25B can select whether to change or delete the color of the character frame in the input image data S1, the burden on the operator in the correction work can be further reduced.

Fourth Embodiment FIG. 10 is a block diagram of a document reading apparatus showing a fourth embodiment of the present invention, and is common to the elements in FIG. 1 showing the first embodiment and common elements. Are given. In this document reading apparatus, a reading apparatus main body 10A having a different configuration is provided instead of the reading apparatus main body 10 in FIG. In the main body 10A of the reading apparatus, a processor 20A in which the image input section 22 in FIG. 1 is replaced with a FAX receiving section 22A is provided. The FAX 60 is connected to the FAX receiving unit 22A via a communication network (for example, a public network) NW. Others
The configuration is similar to that of FIG.

In this document reading apparatus, the transmission data S60 transmitted from the FAX 60 is transmitted to the FA via the public network NW.
The image is taken into the X receiving unit 22A and stored in the image storage area of the memory 40. After that, the same reading operation and correction processing operation as in the first, second, or third embodiment are performed. As described above, in the fourth embodiment, the first, second, or third
As in the embodiment, the burden on the operator in correcting the recognition result of the transmission data S60 transmitted from the FAX 60 can be reduced.

The present invention is not limited to the above embodiment,
Various modifications are possible. For example, there are the following modifications. (A) In FIG. 1, the input image data S1 is the scanner unit 1
The input image data generated by another device is stored in a storage device such as a floppy disk, and the stored input image data is read out by a driving device of the storage device. May be output. (B) In FIG. 6, the extracted image data of the character frame is not sent to the monitor 51. However, the image data of the character frame may have the same color as the background of the display screen. Thereby, the character frame is not displayed on the display screen, and the same effect as in the first embodiment can be obtained.

(C) Although the collating character frame FA in FIG. 8 is displayed by a solid line, it may be displayed by, for example, a dashed line or a dashed line. In this case, a process of deleting a part of the collation character frame FA is performed. (D) In the embodiment, the first and second storage units are configured by the memory 40, but may be configured by independent memories. (E) The public network NW in FIG. 10 may be configured by another communication network.

[0028]

As described in detail above, according to the first aspect of the present invention, the character frame conversion / deletion means makes all or any part of the character frame in the input image data different from the character frame. Since the result is converted into a character frame for color comparison or the part is deleted and the result is displayed on the display means, the burden on the operator in the correction work can be reduced.
According to the invention of claim 2, the input image data captured by the image input means is stored in the first storage means, and the recognition result output from the image data recognition means is stored in the second storage means. Thus, the correction process can be performed at an arbitrary time. Therefore, in addition to the effect of the invention according to claim 1, the burden on the operator in the correction work can be further reduced.

According to the third aspect of the present invention, since the image input means is constituted by a scanner, the same effect as that of the first or second aspect of the present invention can be obtained for input image data taken in by the scanner. . According to the fourth aspect of the present invention, since the image input means is constituted by a facsimile, the same effect as that of the first or second aspect of the invention can be obtained for the input image data captured by the facsimile.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a document reading device according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating an example of a document in which characters are described in a character frame.

FIG. 3 is a view showing a display screen corresponding to FIG. 2;

FIG. 4 is a flowchart of a reading operation of FIG. 1;

FIG. 5 is a flowchart of a correction processing operation in FIG. 4;

FIG. 6 is a diagram showing an example of a display screen in FIG. 1;

FIG. 7 is a flowchart of a correction processing operation according to the second embodiment of the present invention.

FIG. 8 is a diagram showing an example of the display screen of FIG.

FIG. 9 is a flowchart of a correction processing operation according to the third embodiment of the present invention.

FIG. 10 is a configuration diagram of a document reading device according to a fourth embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Scanner part 10, 10A Reading device main body 21 Control part 22 Image input part 22A FAX receiving part 23 Image data recognition part 24 Character frame extraction part 25 Character frame conversion part 26 Image display part 40 Memory 51 Monitor 52 Mouse 53 Keyboard 60 FAX

Claims (4)

[Claims] 1. An image input unit for inputting input image data of a document in which characters are described in a character frame, and one or more character patterns included in the input image data input by the image input unit. An image data recognition unit that cuts out each of the surrounding character regions, recognizes the character pattern and outputs a recognition result, and a character frame extraction unit that extracts the character frame from the input image data captured by the image input unit. A character frame conversion / deletion means for converting all or any part of the extracted character frame to a collation character frame of a color different from the character frame or deleting the part and outputting a conversion / deletion result; A display unit for displaying characters in the input image data, a conversion / deletion result, and a recognition result; and an operation by an operator when the characters in the input image data displayed by the display unit do not match the recognition result. Correction input means for inputting a correction signal based on the correction signal, and when the correction signal is input, a control means for performing correction processing for matching the recognition result displayed on the display means to characters in the input image data, A document reading device comprising: 2. An image input unit for receiving input image data of a document in which characters are described in a character frame, a first storage unit for storing the input image data, and a first storage unit for storing the input image data. Image data recognizing means for extracting a character area surrounding each of one or a plurality of character patterns included in the input image data, recognizing the character pattern and outputting a recognition result, and a second storage for storing the recognition result Means, character frame extracting means for extracting the character frame from the input image data stored in the first storage means, all or any part of the extracted character frame having a different color from the character frame. Character frame conversion / deletion means for converting to a collation character frame or deleting the part and outputting a conversion / deletion result; characters in the input image data stored in the first storage means; Results and said Display means for displaying the recognition result stored in the second storage means; and if a character in the input image data displayed by the display means does not match the recognition result, a correction signal is inputted based on an operation of the operator. Correction input means, and when the correction signal is input, control means for performing correction processing for matching the recognition result displayed on the display means to characters in the input image data, Document reading device. 3. A document reading apparatus according to claim 1, wherein said image input means takes in input image data from a scanner that scans a form in which characters are described in a character frame. . 4. The apparatus according to claim 1, wherein said image input means takes in input image data of a document in which a character is written in a character frame from a facsimile apparatus connected to a communication network via the communication network. The document reading device according to claim 1 or 2, wherein