JP2001126052A - Optical character reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent partial removal of necessary entry information and insufficient removal of unnecessary preprinting at the time of removing preprinting for reducing an image quantity in an optical character reader. SOLUTION: Proper processing is executed respectively in areas 300 to 302 designated in advance, thereby a synthetic mask 203 and a subtraction mask image 204 are prepared previously from an unwritten document image 201. The mask 204 is image-subtracted from a written slip image 206 to prepare a subtracted image 208 to be used for storing and transferring. The image 208 and the image 203 are image-synthesized to prepare a synthesized image 209, thereby a desired image can be obtained. Thus, information which has to be erased and information which has to be restored on the image can be controlled. Further, efficiency of storing and transferring the image can be improved by remarkably reducing an image capacity.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical character reader capable of erasing and restoring a blank form image (preprint information image) from a completed form image.

[0002]

2. Description of the Related Art An optical character reader generally used for business uses a form on which a predetermined entry field (field) for a user to write a predetermined item or a handwritten character is printed. It is used for a process of optically reading characters, imprints and the like entered in the entry frame. Also, in recent years, in order to perform data transfer and confirmation between head office and branch office, in order to perform data transfer between head office and branch office, for example, it is required to store an entire filled-out form image input at a branch office or transfer it to a head office. I have.

[0003] The storage and transfer of the entire form image
There is a problem that the image capacity is large, the storage / transfer efficiency is poor, and there is a limit to the efficiency by compressing the image. In order to solve this problem, conventionally, as described in Japanese Patent Publication No. 4-27592, an unfilled form image is stored in advance, and the unfilled form image is subtracted from the filled-out form image. Then, the pre-printed information image (image information of the form including the printed predetermined equal items, etc. and the unfilled frame, etc.) is deleted, the subtracted image is compressed, and the compressed image and the pre-printed image are stored. It has been proposed that the transfer greatly reduces the capacity and increases the efficiency of image storage and transfer. In this prior art, when restoring data transferred from the branch, the original filled-out form image is restored by combining the subtracted image and the print information image at the head office.

[0004]

In the optical character reading apparatus according to the above-mentioned prior art, for example, when a character "mark" is preprinted in a stamp area of a form, and the character is stamped on the character, a mark " When the image portion of the mark was subtracted from the image, there was a problem that a part of the seal was missing. Further, in the conventional technology, for example, when a form has complicated and fine pre-printing, the shading state when a filled-in form image is read, and the shading density when a blank form image is read. Is different from the state of the above, a part of the shade remains in the subtracted image, the compression efficiency of the image deteriorates, and when the image is restored, the part of the shade remaining in the subtracted image is duplicated and unsightly There was a problem that it would be.

Further, in the conventional optical character reading apparatus, when information such as an unnecessary comment is written or printed on an image to be stored, transferred or referred to, the unnecessary comment written by image subtraction cannot be erased. Unnecessary preprinted comments will appear in the image after subtraction, and unnecessary comments that have been preprinted will be deleted from the image after subtraction by image subtraction, but will not be deleted from the blank form image, and will be deleted from the blank form image. There was a problem that unnecessary comments appeared when images were combined.

It is a first object of the present invention to provide an image processing method in which information on a filled-out form image appears on an image after subtraction without being subjected to image subtraction in a designated area, and image synthesis is performed in the designated area. An object of the present invention is to provide an optical character reading device that allows information on an image after subtraction to appear on an image after composition without subtraction.

[0007] A second object of the present invention is that the information in the filled-out form image does not appear on the image after subtraction without being subjected to image subtraction in the designated area. An object of the present invention is to provide an optical character reading apparatus that allows information on an unfilled form image to appear on a combined image without being targeted.

[0008] A third object of the present invention is that the information in the filled-out form image does not appear on the image after the subtraction because the specified area is not subjected to image subtraction. An object of the present invention is to provide an optical character reading apparatus that can prevent information on an image after subtraction and information on an unfilled form image from appearing on an image after composition without being targeted.

[0009]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a method for preprinting information by subtracting an image from a form image including an imprint using a subtraction mask created based on an unfilled form image. An optical character reading apparatus for restoring a form image including a seal by creating an erased image after subtraction and combining the subtracted image with a synthetic mask created based on an unfilled form image, Subtraction is performed using a subtraction mask that does not perform image subtraction on the specified area including the seal imprint, and a subtracted image including the imprint image data is created in the fixed area, and the subtracted image and the preprint information are not included. The first feature is that the mask and the mask are combined.

The present invention also provides a subtracted image in which preprinted information is deleted by subtracting an image from a form image including hatching using a subtraction mask created based on an unfilled form image, and forming the subtracted image. An optical character reading device for restoring a form image including hatching by synthesizing a post-image and a composite mask created based on an unfilled form image, for a designated area including the hatching, Subtracting using a subtraction mask that performs image subtraction, creating a subtracted image by subtracting the shaded image data, and synthesizing the subtracted image with a formed mask including preprinted shaded information. 2.

Further, according to the present invention, a subtracted image in which preprinted information is erased is created by subtracting an image from a form image containing characters that do not need recognition using a subtraction mask created based on an unfilled form image. An optical character reading device for combining the subtracted image and a combined mask created based on an unfilled form image, wherein the subtraction mask performs image subtraction on a designated area including the unnecessary characters. And subtracting the image data in the fixed area to create a subtracted image, and synthesizing the subtracted image with a synthetic mask that does not include the image data in the fixed area. And

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the optical character reading device according to the present invention will be described below in detail with reference to the drawings.
FIG. 1 is a diagram for explaining each process of the optical character reading apparatus according to the present embodiment, FIG. 2 is a diagram showing a form or an image appearing before and after the process according to the present embodiment, and FIG. 3 is a designation according to the present embodiment. FIG. 4 is a diagram for explaining an area, and FIG. 4 is a diagram showing a change in color of each pixel during image subtraction and image synthesis.

First, the configuration of the optical character reading apparatus according to the present embodiment is a general configuration. For example, an image scanner for optically scanning a form and inputting image data, and reading by the image scanner. A / A that converts analog image data into digital data
A D converter, a recognition unit that cuts out image data in an entry frame of a character or the like at a designated position from the converted image data, recognizes the data as a character or a symbol, and converts it into a character code, and a control unit that controls these. And a storage unit for storing a program for controlling the control unit, read image data and recognition results, a display, a keyboard, and the like.

The form applied to the present invention has a predetermined wording on paper, an entry frame for a user to enter characters and numbers by hand, and an authentication seal. It is assumed that the user writes a character or the like by hand in the entry frame and uses the form stamped on the stamp frame.

This form, for example, an unfilled form is shown in FIG.
As shown in (a), an entry frame A on which characters "yen" are printed in advance to enter the amount of money, a seal frame B on which characters "symbol" for stamping are printed, and a check box for input. It is assumed that the form includes a confirmation frame C and a hatched frame D on which fine hatching is printed.
As shown in (f), it is assumed that the amount "12" is entered in the entry frame A, "Hitachi" is stamped in the stamp frame B, and a check mark indicating that the confirmation has been made is written in the confirmation frame C. It should be noted that the actual form has a different form from that shown in the figure, but in the present specification, for ease of understanding, a form in which the entry frame A and the like related to the present invention are extracted will be described as an example.

First, the optical character reading apparatus according to the present embodiment creates a subtraction mask 204 and a composite mask 203 for performing processing. This subtraction mask 204 corresponds to FIG.
As shown in (e), the completed form image (FIG. 2f)
2 are used to remove character symbols and the like (frame lines, characters of “circle” and “confirmation” and hatched portions in FIG. 2A) printed in advance from FIG. As shown in (d), this is a mask image for restoring the original image by combining it with the input characters and the imprint part (subtracted image) subtracted using the subtraction mask 204.

The combination mask 203 and the subtraction mask 2
04 (mask image), as shown in FIG.
Optically scanning the blank form 200 shown in FIG.
The blank form image 201 of (b) is input (step 100), and distortion correction and size correction are performed based on the edge coordinates of the form image 201, and the blank form after the distortion correction shown in FIG. An image 202 is created (step 101), and a process such as filling each area is performed on each of the pre-designated areas in the form image 202 (step 10).
2).

In this process, as shown in FIG. 3, a process of filling the composite mask 203 and the subtraction mask 204 with white is performed on the area 300 for which the input through is specified, and the area 301 for which the synthetic mask image through is specified on the lower stage. 2 (d) and (e) by performing a process of painting the subtraction mask 204 black and performing a process of painting the composite mask 203 white and painting the subtraction mask 204 black in the area 302 designated as a complete mask. ) Are for forming the masks 203 and 204 shown in FIG. In other words, step 102 is a subtraction mask 2 for subtracting, from the filled-out form image, a frame portion, a portion with a margin of a character “circle”, and a portion that does not need to be read (a confirmation frame C and a hatched frame D).
04, and the combined mask 203 for restoring the original image by combining the original input character and the imprint part (subtracted image 208) subtracted by using the subtraction mask 204 is created. is there. This step 1
The combined mask 203 and the subtraction mask 204 obtained in step 02 are output as files to a storage unit or the like for use in processing described later (step 103).

Next, a method of creating and transferring an image after subtraction, which efficiently stores and transfers an image using the above-described combination mask 203 and subtraction mask 204, will be described with reference to FIG. In this process, first, the subtraction mask 2 shown in FIG.
04 is input from the storage unit (step 110), and then the filled-out form 205 is scanned to take in the filled-out form image 206 of FIG. 2 (g) (step 11).
1), the form image 206 is corrected for distortion, and FIG.
A completed form image 207 after the distortion correction of (h) is obtained (step 112).

Next, in this processing, the subtraction mask image 204 is image-subtracted from the corrected filled-out form image 207, and a post-subtraction image 208 shown in FIG. 2 (i) is created (step 113). The subtracted image 208 is obtained by subtracting a frame portion, a portion with a margin of a character “circle”, and a portion that does not need to be read (a confirmation frame C and a hatched frame D) from the image of the completed form. .

Here, the change of the color of each pixel in the image subtraction will be described with reference to FIG. In FIG. 4 (a), when the value of each pixel is 0, it is white, and when it is 1, it is black. In each pixel, only when both the filled-out form image and the subtraction mask image are black, The color becomes white, and in other cases, the color of the completed form image appears on the image after the subtraction.

Therefore, in the input through area 300 of FIG. 3, since all the pixels of the subtraction mask image are white, the input through area does not change by image subtraction, and the completed form image appears as it is in the subtracted image. Therefore, the character "mark" in the area 300 is not subjected to the image subtraction, so that a part of the image is not lost.

On the other hand, in the mask-through area 301 and the complete mask area 302, all the pixels of the subtraction mask image are black. It can be seen that the painting appears.
Further, in areas other than the area in which the above-mentioned processing is designated, pre-print image subtraction is performed on an unfilled form image.

The subtracted image 208 of FIG. 2 (i) to which the image subtraction has been performed in step 113 is compressed and output to a storage unit as a file for storage, and the stored subtracted image 208 is necessary. If there is, the transfer is made to the head office and the like (step 115).

According to the above-described series of processing, the present embodiment provides:
By compressing the created subtracted image and storing / transferring it, the efficiency can be greatly improved as compared with the case where the filled-out form image with the preprint is compressed as it is.

Next, a method of creating a combined image for referring to a completed form image will be described with reference to FIG. In this process, the stored combined mask image 203 is input from a file in the storage unit (step 120), and similarly, the subtracted image 208 is input from the file (step 12).
1), the image 208 after subtraction and the composite mask image 20
3 are combined to create a combined image 209 shown in FIG.

Here, the change in the color of each pixel in image synthesis will be described with reference to FIG. In FIG. 4B, as in the case described above, white when the value of each pixel is 0 and black when it is 1, and when each of the post-subtraction image and the composite mask image is black, the composite is performed. This indicates that the color on the post-image is black, and in other cases the color on the post-combination image is black. In other words, this figure shows that the result of the logical sum of the subtracted image and the combined mask image is the combined image.

Therefore, in the input through area 300 (FIG. 3)
Since the corresponding portion of the composite mask image 203 is white, the subtracted image appears as it is in the composite image, and the “mark” character on the blank form image is superimposed and composited. To prevent that. Similarly, since the subtracted image is white in the mask through area 301, the combined mask image (shaded) appears as it is in the combined image, and the complete mask area 302 is white because both the subtracted image and the combined mask image are white. In addition, a white fill appears in the image after combination (characters “confirmation” and the like are deleted). In areas other than the areas where these processes are designated, preprinted image synthesis on an image of an unfilled form is performed, for example, as a frame. By this series of processes, the present embodiment is configured as shown in FIG.
As shown in (1), a combined image 209 that matches the input form image can be obtained. As described above, according to the present embodiment, the image in the seal frame B (imprint image) uses the subtraction mask that does not subtract the image in the specified area including the imprint.
A post-subtraction image including the image data of the seal imprint is created in the fixed area, and the post-subtraction image is pre-printed on the seal frame B by combining the post-subtraction image with a synthesis mask not including the pre-print information. However, even if it is performed, the seal imprinted on the character “seal” appears in the image after subtraction, that is, the seal imprinted from above the “seal” without setting the seal frame B as the image subtraction target. Since it appears in the image after the subtraction together with the “mark” of the preprint, it is possible to prevent a part of the imprint from being lost. In addition, since the information on the image after subtraction appears on the image after composition without using the sealing frame B as a target of image composition, the “mark” of the preprinted form image of the blank form image appears on the image after composition on the image after composition. Can be prevented from overlapping.

Further, according to the present embodiment, in the hatched frame D, a subtracted image is created by subtracting the shaded image data by using a subtraction mask for performing subtraction in a designated area including the shaded area. The subtracted image and the formed mask including preprinted shaded information are combined. As a result, when there is a complicated and fine hatching of the pre-print, the area of the hatching frame D does not appear in the subtracted image because the hatching does not appear, so that the compression efficiency of the image is not deteriorated and the image is not subjected to image synthesis. For this reason, shading on an unfilled form image appears on the combined image, so that it is possible to prevent a part of the shading from being duplicated and becoming unsightly.

Further, according to the present embodiment, the subtraction mask for subtracting the image data in the fixed area is used in the confirmation frame C by using a subtraction mask for performing subtraction in a designated area including characters that do not need to be recognized in this frame. By creating an image and synthesizing the subtracted image and a synthetic mask not including the image data in the fixed area, an unnecessary comment such as the character “confirmation” is added to the image to be stored or the like in the above example. Even if the information is entered or preprinted, it is possible to prevent unnecessary comments from appearing on the image after the subtraction in the area of the confirmation frame C. Also, it is possible to prevent unnecessary comments from being stored, transferred, or referred to.

In other words, the first feature of this embodiment is that the designated area of the seal imprint on the blank form image is filled in white before subtraction of the image, so that the blank form image is not filled out. Even if the image of the form image is subtracted, the filled form image appears in the designated area as it is in the designated area. Also, when combining images, the designated area is previously filled with white, and an unfilled form image and the subtracted image are combined to form an image.
The pre-print in the designated area is not overlapped with the combined image.

Further, a second feature of the present embodiment is that, before subtraction of an image, a designated area of a stitch on an unfilled form image is filled in black in advance, so that an unfilled form is When the image is subtracted from the image, the information on the form image already written in the designated area does not appear in the image after the subtraction. Also, when combining images, pre-printing on the blank form image is performed in the designated area of the combined image by combining the subtracted image and the blank form image without any information in the designated area. It is to appear.

Further, a third feature of the present embodiment is that, before subtraction of an image, a designated area on an unfilled form image is painted in black in advance, so that an unfilled form image can be converted from a completed form image. When the image is subtracted, the information on the completed form image does not appear in the designated area at all in the designated area. In addition, when combining images, an unfilled form image in which the designated area is painted in white in advance and a subtracted image having no information in the designated area are combined to form an image in the designated area of the combined image. No characters or pre-printing appear at all.

The present invention can also be represented by the following embodiments. <Embodiment 1> An image of an unfilled form image is subtracted from a filled-out form image to create a subtracted image in which pre-print information is deleted. In an optical character reading device capable of restoring pre-print information by combining, in a designated area, information on a filled-out form image appears on the image after subtraction without being subjected to image subtraction, and the designated area An optical character reading device characterized in that information on an image after subtraction appears on an image after composition without being used as an object for image composition.

<Second Embodiment> An image of a blank form image is subtracted from a filled-out form image, thereby making it possible to create a subtracted image in which pre-print information is deleted. In an optical character reading device that can restore pre-print information by combining images, the information on the filled-out form image does not appear on the image after subtraction without specifying the image subtraction in the designated area ,Also,
An optical character reading device, wherein information in a blank form image appears in a combined image without being subjected to image combining in a designated area.

<Embodiment 3> An image of a blank form image is subtracted from a completed form image to create a subtracted image in which preprinted information is deleted. Further, the subtracted image and an unfilled form can be created. In an optical character reading device that can restore pre-print information by combining images, the information on the filled-out form image does not appear on the image after subtraction without specifying the image subtraction in the designated area ,Also,
An optical character reading device, wherein a designated area is not subjected to image composition, and neither information on a subtracted image nor information on an unfilled form image does not appear on the composed image.

[0037]

According to the present invention, when a pre-print information is erased by subtracting an unfilled form image from a filled-in form image to create a subtracted image, the erased pre-print information is also used. In order to restore, when combining the image after subtraction and the blank form image, by specifying the area in advance, the information on the filled-out form image in the designated area of the subtracted image and the combined image, Alternatively, it is possible to select whether information on an unfilled form image or information without any information appears, so that information to be deleted or information to be restored on the image can be controlled. Further, according to the present invention, it is possible to greatly reduce the image capacity and increase the efficiency of image storage and transfer.

[Brief description of the drawings]

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

FIG. 2 is a view for explaining a form or an image appearing before and after each processing in FIG. 1;

FIG. 3 is a view for explaining area designation according to the embodiment of the present invention;

FIG. 4 is a diagram showing a change in color of each pixel during image subtraction and image synthesis according to the embodiment.

[Explanation of symbols]

100 to 103; each processing step when creating a mask image; 110 to 115; each processing step when creating and transferring an image after subtraction; 120 to 123; each processing step when creating and referencing an image after synthesis; 200 Blank form, 201;
Blank form image, 202; Blank form image after distortion correction, 203; Composite mask image, 204; Subtraction mask image, 205; Filled form, 206; Filled form image, 207; Filled form after distortion correction Image, 208; image after subtraction, 209; image after synthesis, 30
0: input through area, 301: mask through area,
302: Complete mask area.

Claims (3)

[Claims] An image is subtracted from a form image including a seal image using a subtraction mask created based on an unfilled form image to create a subtracted image in which preprinted information is erased, and the subtracted image and an unprocessed image are created. An optical character reader for restoring a form image including an imprint by synthesizing a synthetic mask created based on a filled-in form image, wherein a subtraction mask that does not perform image subtraction is applied to a designated area including the imprint. Optical character reading, wherein a subtracted image including image data of a seal imprint is created in the fixed area by subtraction using the subtracted image, and the subtracted image is combined with a composite mask not including preprint information. apparatus. 2. A subtracted image in which preprint information is deleted by subtracting an image from a form image including a hatched form using a subtraction mask created based on an unfilled form image, to generate a subtracted image, An optical character reading device for restoring a form image including hatching by combining a composite mask created based on an unfilled form image,
A subtracted image is subtracted by using a subtraction mask for performing image subtraction on the designated area including the shaded area, thereby creating a subtracted image by subtracting the shaded image data, and preprinting the subtracted image and the shaded area. An optical character reading device for synthesizing a synthetic mask containing information. 3. A subtracted image in which preprint information is deleted by subtracting an image from a form image containing characters that do not require recognition using a subtraction mask created based on an unfilled form image, and creating a subtracted image. An optical character reading device that combines a post-image and a composite mask created based on an unfilled form image, using a subtraction mask that performs image subtraction on a designated area including the unnecessary characters. An optical character, wherein a subtracted image is created by subtracting the image data in the fixed area by subtraction, and the combined image and the combined mask not including the image data in the fixed area are created. Reader.