JP2000181988A - Optical character reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform a highly accurate masking processing or a masking processing for not erasing anything other than an entered character by utilizing an optical character reader. SOLUTION: A slip 15 is optically scanned on the slip 15 in an image input part 1, image data converted to a binarized pattern are stored in an image pattern storage memory 2 and the coordinate information of a read area is generated through a slip extraction part 9 and a character extraction part 7 based on format information stored in a format information storage memory 4 beforehand. At the time, correction is performed based on correction information stored in a slip extension/contraction correction learning value storage memory 5. To the image data of the corrected masking row area, by turning the pixels of the entire masking row area or the character pixels of only secret characters to the same color as a background color, the masking processing of a secret term character area in an image pattern is performed by the information of the read area extension/contraction corrected from a slip image and the highly accurate masking processing is made possible.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical character reading apparatus having an image processing function for reading images and character information such as figures, and more particularly to an optical character reading apparatus for reading a form having a confidential item.

[0002]

2. Description of the Related Art In general, an optical character reader for reading a form in which characters and figures are mixed as an electronic file describes a confidential item such as a personal identification number as described in, for example, JP-A-6-290251. When automatically registering the created form image, in order to prevent the information of the confidential item from being read, a region in which the confidential item is described is calculated from format information of the confidential item specified in advance, and for this region, A device that generates and combines an image mask pattern and displays or prints image data has been proposed. The optical character reading device according to this prior art masks the confidential item during display and printing, but the registered image data object includes the confidential item.

[0003]

The above-described optical character reading apparatus according to the prior art described above, when masking a confidential item, replaces a specific area whose position is designated in advance with a mask pattern. For example, if the position of the confidential item is read out of the original position, the mask position may deviate from the confidential item position, and the confidential item may not be able to be hidden accurately.
In addition, there is a problem that only the confidential items that should be hidden cannot be hidden.

In particular, when a specific area is extracted from image information without a background captured by a scanner, a facsimile, or the like, since the expansion / contraction correction and skew correction of the form cannot be performed, the specific area on the form cannot be accurately masked. There was a problem that said.

Further, in the conventional apparatus, since the confidential item is also read as data, it is difficult to say that the confidential item is sufficiently protected.

SUMMARY OF THE INVENTION An object of the present invention is to eliminate the above-mentioned disadvantages of the prior art, and to provide an optical character reading device capable of masking and hiding confidential items of a form accurately. Another object of the present invention is to provide an optical character reading device capable of preventing data reading of confidential items.

[0007]

According to the present invention, there is provided an optical system for reading an image from a form having a mask area in which a confidential character is described on a background color and recognizing character data from the image. In the character reading device, the inclination of the form and the expansion and contraction of the form are corrected from the image image, and the entire mask area pixels on the corrected form are converted into pixels of the same color, thereby masking the entire mask area and keeping confidentiality. Hiding characters is a first feature.

Further, in the optical character reading apparatus according to the present invention, the inclination of the form and the expansion and contraction of the form are corrected from the image image, and only the pixels of the confidential character on the form after the correction are made the same as the background color. The second feature is to mask only the confidential characters and hide the confidential characters.

Further, according to the present invention, in the optical character reading device, an image mask pattern for correcting a tilt of a form and expansion and contraction of the form from the image image and masking all pixels in a mask area on the form after the correction is generated. The conversion of the image mask pattern into pixels in a mask area is performed in a third step.
The feature of.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an optical character reader according to an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a view for explaining an optical character reading apparatus according to the present embodiment, FIG. 2 is a view showing a form to be read in the present embodiment, and FIG. 3 is an image image when a confidential area is designated as an area mask. FIG. 4 is a diagram showing an image image when a confidential area is designated by a character mask, and FIG. 5 is a diagram for explaining a correction screen of an image image to be subjected to mask processing.

The optical character reading apparatus according to the present embodiment has, as its basic functions, an inherent difference in expansion and contraction correction amount due to a difference in feed speed when reading a form, and a tilt correction when the form is conveyed obliquely. After learning the amount, the image data of the form read using the stored expansion / contraction correction amount and inclination correction amount is subjected to expansion / contraction correction and inclination correction, and then character recognition and graphics are performed. And the like can be read.

A form 101 to which the present embodiment is applied
As shown in FIG. 2, a data line 103 in which black character data is described on a background color (for example, white), a graphic area 104 in which a graphic or the like is drawn, and a user's confidential character (item)
Is composed of a mask row 102 which is a mask area in which black is described.

As shown in FIG. 1, the optical character reading apparatus according to the present embodiment temporarily stores an image input unit 1 for reading an image of a form 15 and an image input from the image input unit 1. The image pattern storage memory 2 and format information corresponding to a plurality of forms to be read, for example, form width / height, form thickness, position and type of characters to be read, image cutout position and size, mask processing method (area Format information storage memory 4 for storing a mask target field position, an image pattern to be edited, a storage memory 5 for storing the expansion / contraction correction amount of the form, and the like. A form extracting unit 9 for extracting a form image image from the corrected image image in comparison with the background, and determining whether the extracted form image image is A coordinate extracting unit 8 for extracting coordinates of characters and the like; a character extracting unit 7 for extracting a character image from the coordinate position; a character recognizing unit 6 for recognizing characters from the image; Control unit 3
A recognition result information storage memory 10 for storing the character recognition result, a character pattern storage memory 11 for storing a character pattern, and a display / input unit for inputting an operator's command and displaying an edit screen and the like to be described later. 13, an image editing pattern memory 14 for storing image patterns generated for characters and area masks, an auxiliary storage unit 16 such as a hard disk for storing read data and image data after character correction, and a device control for controlling these. Part 1
2 is provided.

In the optical character reading apparatus thus configured, the operator can use the form information (form width / height, form thickness, position and type of characters to be read, image cutout position and size, mask processing method ( Area specification / character specification), mask target field position, image pattern to be edited)
Is input from the input / display unit 13 and stored in the format information storage memory, and it is checked that a field (mask line 102) to be masked is included in the target image cutout position.

Next, in response to a form reading instruction from the apparatus control section 12, the image input section 1 forms the form 15
The top is optically scanned and the form image (image data) is
Converts to a digitized pattern and stores it in image pattern storage memory 2.
To be stored.

Subsequently, the device control unit 12 issues a character reading instruction to the character recognition control unit 3, whereby the character recognition control unit 3 stores format information in advance from the form image stored in the image pattern storage memory 2. In accordance with the format stored in the memory 4, the form extracting unit 9 cuts out image data of the form from the image data including the background. The image data of this form is subjected to expansion / contraction correction with reference to the form expansion / contraction correction value stored in the form expansion / contraction correction learning value storage memory 5 previously stored.

Next, according to the present invention, the character extracting unit 8 comprises:
According to the format stored in the format information storage memory 4, characters and image patterns are cut out based on the coordinates of the form image cut out by the form extraction unit 9, and the coordinates of the character position and the image position are stored in the recognition result information storage memory 10. Output information. This data also performs expansion / contraction correction with reference to the form expansion / contraction correction value in the form expansion / contraction correction learning value storage memory 5 as described above.

Next, the character recognizing section 6 performs a character recognizing process from the character pattern cut out by the character extracting section 8 and stores the recognition result in the recognition result information storage memory 10. Output to the device control unit 12.

The apparatus control unit 12 reads from the image pattern storage memory 2 based on the read field coordinate information stored in the recognition result information storage memory 10 and the coordinate information of the image whose specific position is specified by the format specification. A field image pattern and an image pattern instructed to cut out the image are generated and stored in the image editing pattern storage memory 14, and if a mask instruction is present in the format stored in the format information storage memory 10, the mask type It is determined whether is a region specification or a character specification. This specification is specified in advance by the operator.

If the mask instruction is a mask instruction for specifying an area, that is, a mask instruction for masking the entire area in which the confidential item is described, the control unit 12 stores the coordinate information of the mask target field in the recognition result information storage memory 4. From the format information storage memory 10, read out the coordinates of the image pattern in the form from the format information storage memory 10, find the relative value of the mask area coordinates from the image cut out coordinates, and then read the image pattern storage memory 2 based on the image cut out coordinate information.
Then, the entire mask area is converted into white bits from the extracted image pattern and output to the image editing pattern storage memory 14.

As shown in FIG. 3, in the image pattern 201 stored in the memory 14, the image of the data row 203 and the graphic area 204 is left visible, and the data in the mask row 202 is replaced with the mask row 202. Are converted into white bits to make the format invisible.
That is, according to the present embodiment, the entire area including the confidential character can be hidden by converting the entire area in which the confidential character whose background color is described in black on white into white pixels. The mask of the entire mask row may generate an image mask pattern for masking the mask row, and convert the image mask pattern to data in the mask row 202. This image mask pattern may be the same color or may be a pattern on which a predetermined character or pattern is applied.

When the mask instruction is a mask instruction for specifying a character, that is, a mask instruction for masking only characters,
The control unit 12 requests the character recognition control unit 3 to obtain the coordinate information of the character pattern, obtains the coordinate information of the character pattern of the field to be masked, and the coordinate extracting unit 8 uses the position of the field to be masked as input information for the character pattern. The coordinate information extraction process is performed in the field of the character pattern storage memory 11.
To be stored. Next, the control unit 12 reads the information in the character pattern storage memory 11 and the image pattern cutout coordinates in the form from the format information storage memory 10, and calculates the relative value of the mask area coordinates from the image cutout coordinates.
Based on the image cutout coordinate information, cut out from the image pattern in the image pattern storage memory 2, and subsequently convert all the coordinates of the character patterns in the mask area from the cutout image pattern into white bits and output them to the image edit pattern storage memory 14. I do.

As shown in FIG. 4, in the image pattern 301 stored in the memory 14, the data line 303 and the image of the graphic area 304 are left visible, and the mask line 302 is replaced with the characters in the mask line 302. All the pixels in the portion are converted into white bits, so that they are invisible. In FIG. 4, characters are indicated by broken lines to indicate that only the character portions have been deleted. That is, according to the present embodiment, only the pixels of the characters in the area in which the confidential character whose background color is described in black on white are described are converted into white pixels, so that only the confidential character can be hidden. Therefore, in the example of FIG. 4, the character of “ID” remains as compared with FIG.

When the read result contains unreadable characters, the device control unit 12 displays a correction screen on the display / input unit 13 to urge the operator to make correction. At this time,
The image pattern stored in the image editing pattern storage memory is also displayed. As shown in FIG. 5, the correction screen includes an image pattern image 401 including read image data and a masked mask row 402.
And a recognition result "ABC of a data line (reference numeral 103 in FIG. 2) in which character data in the image pattern 401 is described.
D ~ ", a read result image 404 displaying a recognition result of a mask row (reference numeral 102 in Fig. 2) in which personal recognition information is described, and a read image image of the mask row itself Is displayed by the operator with reference to the field image image 405 to determine that the unrecognizable character (indicated by “?”) In the read result image 404 is “4”. By correcting the recognition character of the image 404 as “4”, the correction can be performed without looking at the form even if the character cannot be recognized by the apparatus.

The optical character reading device that has made the correction stores the read data and the image data based on the form in the auxiliary storage device 16 such as a hard disk. At this time, the correction of the field designated by the mask is performed. In (2), an unmasked image is displayed for the corrected image of the read data, but the masked image is stored for the area image.

In the above embodiment, the optical character reading apparatus for reading a form in which black characters and the like are described on a white background color has been described. However, the present invention is not limited to this monochrome type. Mask processing so that characters cannot be recognized with respect to the background color of the form, for example, at the time of mask processing by specifying the area, collectively convert the pixels of the mask row to pixels of the same color, or at the time of mask processing by specifying characters, Converts only text pixels to background pixels,
The present invention can also be applied to a color optical character reading device.

As described above, the optical character reading apparatus according to the present embodiment corrects the inclination of the form and the expansion and contraction of the form from the image image, and converts the entire mask area pixels on the form after correction into pixels of the same color. By masking the confidential characters by masking the entire mask area, or by making only the pixels of the confidential characters on the form the same as the background color, the confidential items of the form can be accurately masked and hidden.

Note that the present invention can be represented as the following embodiments.

<Embodiment 1> In an optical character reading apparatus having a function of inputting an image, means for displaying and outputting the read image / character, and calculating the inclination of the form and expansion / contraction of the form from the input image. Means for accurately converting a character position on a form into coordinate information, means for calculating image information (the number of black pixels in the vertical / horizontal direction) of an input character, and means for inputting information on a character / field to be masked And image information (the number of white pixels in the vertical / horizontal direction) different from the specific image information (the number of black pixels in the vertical / horizontal direction)
Means for replacing a character / field designated in advance, and displaying or outputting an image in which only the number of black pixels of the image information of the input character is replaced with the number of white pixels. Optical character reading device.

<Embodiment 2> An optical character reader having means for inputting a color image, means for displaying and outputting a read image / character, and means for converting color image information into binary image information The apparatus calculates the inclination of the form and the expansion and contraction of the form from the input binary image, and converts the character position on the form into coordinate information accurately. The character pattern information (vertical / vertical) is obtained from the binary image information. Means for calculating the number of black pixels in the horizontal direction, means for inputting information on characters / fields to be masked, and image information different from specific image information (number of black pixels in the vertical / horizontal direction) (number of white pixels in the vertical / horizontal direction) ) Is provided, and for a previously specified character / field, only the number of black pixels in the image information of the input character is replaced with the number of white pixels. An optical character reading device for displaying or outputting

<Third Embodiment> In an optical character reading apparatus capable of inputting an image, a means for displaying and outputting the read image / character and a means for inputting information of a character / field to be masked are input. It has means for calculating the inclination of the form and expansion / contraction of the form from the image and accurately converting the character position on the form into coordinate information, and means for generating an image mask pattern of a specific area in the read image. An optical character reading apparatus for generating and displaying or outputting an image mask pattern for an accurate area on a form in an input image for a predetermined character / field.

[0032]

As described above, the optical character reading apparatus according to the present invention corrects the inclination of the form and the expansion and contraction of the form from the image image, and converts the entire mask area pixels on the form after correction into pixels of the same color. By doing so, the confidential character is masked to hide the confidential character, or only the pixels of the confidential character on the form are made the same as the background color, whereby the confidential item of the form can be masked and hidden accurately.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining an optical character reading device according to an embodiment of the present invention.

FIG. 2 is a view showing a form to be read according to the embodiment;

FIG. 3 is a diagram showing an image image when a confidential area is designated as an area mask.

FIG. 4 is a diagram showing an image image when a confidential area is designated by a character mask.

FIG. 5 is a view for explaining a correction screen of an image image to be subjected to mask processing.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Image input part, 2 ... Image pattern storage memory, 3 ... Character recognition control part, 4 ... Format information storage memory, 5
... Form expansion / contraction correction learning value storage memory, 6.
... Character extraction unit, 8 ... Coordinate extraction unit, 9 ... Form extraction unit, 10
... Recognition result information storage memory, 11 ... Character pattern storage memory, 12 ... Device control unit, 13 ... Display / input device, 14
... Image editing pattern storage memory, 15 ... Form, 16
... Auxiliary storage device 101 ... Form, 102 ... Mask line, 10
3 data line, 201 image pattern, 202 mask line, 203 data line, 301 image pattern, 302 mask line, 303 data line, 401 image pattern, 402 mask line, 403 correction screen, 404: read result 1, 405: field image, 406: read result 2.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Hideharu Hattori 2880 Kozu, Odawara-shi, Kanagawa Prefecture Storage Systems Division, Hitachi, Ltd. (72) Masatoshi Kurata 2880 Kozu, Kozu, Odawara-shi, Kanagawa Hitachi, Ltd. F term in the storage system division (reference) 5B029 AA01 BB02 CC27 EE04 EE15 5B064 AA01 BA01 CA07 CA08 EA05 FA05

Claims (3)

[Claims] 1. An optical character reader for reading an image from a form having a mask area in which a confidential character is described on a background color and recognizing character data from the image, wherein the inclination of the form and the form An optical character reader which corrects expansion and contraction and converts all the pixels of the mask area on the form after the correction into pixels of the same color. 2. An optical character reading apparatus for reading an image from a form having a mask area in which a confidential character is described on a background color and recognizing character data from the image, the inclination of the form and the form An optical character reading device, wherein expansion and contraction is corrected, and only pixels of confidential characters on the form after the correction are made the same as the background color. 3. An optical character reader for reading an image from a form having a mask area in which a confidential character is described on a background color and recognizing character data from the image, wherein the inclination of the form and the form An optical character reading device for correcting an expansion / contraction, generating an image mask pattern for masking pixels in the entire mask area on the form after the correction, and converting the image mask pattern into pixels in the mask area.