JPH0460887A - Optical character reader - Google Patents

Abstract

PURPOSE:To normally recognize a character by only giving simple correction by performing character recognition by excluding the character on a character entry field corresponding to a mark decided by a mark position deciding means from an object to be processed. CONSTITUTION:The character recognition is performed by excluding the character on the character entry field corresponding to the mark decided by the mark position deciding means from the object to be processed. The mark is entered at the position of the mark entry field 12 corresponding to the position of the mark entry field 12 corresponding to the position of the character desired to correct on the character entry field 11, namely the position of a mark entry field 12 corresponding to the position just above a wrong character, and the correct character is entered in the next frame of the character entry field 11. Thus, the character recognition can be normally performed by only re-entering the correct character in the character entry field, and in addition, entering the mark at the corresponding position of the mark entry field.

Description

[Detailed description of the invention] [[1st object] of the invention (Industrial application field) The present invention relates to an optical character reading device.

(Prior Art) When using an optical character reading device, sometimes characters are written incorrectly on a form, and since the characters were written with a hole bene or the like, it is impossible to correct the errors. In this case, the form entry 15 often involves writing a double line or the like over the incorrect letter □, and then writing the correct letter in the next square L1.

But why correct the characters in this way? In the case of -i, even if character recognition could be performed for each character, there were many cases in which no results were obtained from knowledge processing, or completely different results were output.

For example, if "1m tree" (zasaki) is written in the form, even though the word "sasaki" exists in the knowledge dictionary, the knowledge processing will give priority to the number of characters in the judgment. In the end, the word candidate "Sasaki" is output.

(Problems to be Solved by the Invention) As described above, conventional optical character reading devices have a problem in that they lack a countermeasure when a character is written incorrectly on a form.

The present invention is intended to solve these problems, and aims to provide an optical character reading device that can correctly recognize characters by making simple corrections even if characters are written incorrectly on a form. It is said that

[Configuration of the Invention (Means for Solving the Problems) In order to achieve the above object, the optical character reading device of the present invention includes a character entry field in which characters are written and a screen that indicates typographical errors on this character entry field. an image reading means for reading an image of a form having a mark entry field in which a mark is entered; a mark detection means for detecting a mark from the result of reading the image of the mark entry field; and a mark detection means for detecting a mark detected by the mark detection means. Mark position determination means for determining the position of the mark on the mark entry field, and character recognition by excluding from processing the characters on the character entry field corresponding to the position of the mark on the mark entry field determined by the mark position determination means. Character recognition means that performs

(Function) In the optical character reading device of the present invention, the mark detection means detects the mark from the image reading result of the mark entry field, and then the mark position determination means determines the position of the mark on the mark entry field. . Then, the character recognition means performs character recognition by excluding from processing the character on the character entry field corresponding to the position of the mark on the mark entry field determined by the mark position determination means.

Therefore, according to the present invention, even if a character is written incorrectly on a form, character recognition can be performed normally by simply rewriting the correct character in the character entry field and writing a mark at the corresponding position in the mark entry field. be able to.

(Example) Hereinafter, embodiments of the present invention will be described based on the drawings.

FIG. 2 is a block diagram for explaining the configuration of an optical character reading device according to an embodiment of the present invention.

In the figure, 1 is an image reading unit that optically reads the image on the form P, 2 is a character cutting unit that cuts out an image for each character from the image read from the form P, and 3 is 1
A character recognition unit performs character recognition on character images cut out for each character; 4 is a knowledge dictionary storing knowledge information such as words related to names, addresses, etc.; and 5 is a knowledge dictionary that stores information obtained by the character recognition unit 3. This is a knowledge processing unit that determines the final character recognition result by performing knowledge processing on combinations of character candidates using knowledge information in the knowledge dictionary 4.

FIG. 1 shows the format of a form adopted in this optical character reading device.

In the figure, numeral 11 is a character entry field in which characters are entered, and 12 is a mark entry field in which a mark for indicating a typographical error on the character entry field 11 is entered. That is, in this form P, each mark entry field 12 is arranged in parallel with the character entry field 11 at a position near one of the corresponding character entry fields 1.

Here, let us assume that when attempting to enter the first and last name "Kae Okazaki, Hanae Okazaki" in the character entry field 11 of the form P, the user mistakenly enters "Karyo Okazaki, Hanako Okazaki." In this case, enter a mark in the mark entry field 12 at the position of the mark entry field 12 corresponding to the position of the character you want to correct in the character entry field 11, that is, the position directly above the typo, and place the correct character next to the character entry field 11. Fill in one box. Note that any shape of the mark may be used. In other words, it is sufficient if the number of dots in the area corresponding to one mark of the mark entry field 12 is large enough to be distinguished from adhesion of dust or the like.

Next, the operation of this optical character reading device will be explained.

FIG. 3 is a flowchart showing the flow of this operation.

First, the image reading section 1 reads an image of the form P having the corrected typographical errors as shown in FIG. 1 (step A).
). As a result, images of the character entry field 11 and the mark entry field 12 are respectively input. Next, the character cutting section 2 cuts out the image in each field 11, 12 into areas corresponding to one character (step B).

Next, the character recognition section 3 performs character recognition, or in this embodiment, the following processing is performed prior to character recognition processing.

That is, a mark is detected from the image input from the mark entry field 12 (step C), and if a mark is detected, the position of this mark on the mark entry field 12 is determined (step D). Next, on the screen of a display device connected to the outside of this optical character reader,
A character image at the corresponding position on the mark entry field 12 is displayed to prompt the operator to confirm whether the character is actually a typo (step E).

As a result of the confirmation, if it is determined that the character is correct or a typo, the process is interrupted and the positional relationship between the mark position on the form P and the typo is confirmed. In addition, if it is confirmed that the determination of typographical errors is correct, the process is continued as is.

Thereafter, in the character recognition unit 3, the character entry field 11 corresponding to the mark position on the mark entry field 12 is
Character recognition is performed with the upper character excluded from processing (step F).

Then, from the combination of character complements written as a result of this character recognition,
The knowledge processing unit 5 determines the final character using the knowledge information in the knowledge dictionary 4 (step G).

Thus, according to the optical character reading device of this embodiment, even if a character is written incorrectly on the form P, a mark is written in the corresponding position of the mark entry field]2, and the correct character is entered in the character entry field l'11. By filling in the next square 1], character recognition can be performed normally.

In the above embodiment, the form P is provided with a character entry field 1 and a mark entry field 12 each having a square opening for each character, or a free pitch field having no square openings is used. The present invention can be applied even in such cases.

Furthermore, the positional relationship between a character entry field and a corresponding mark entry field is not limited to the above-mentioned embodiment; for example, a mark entry field may be set below a character entry field, or a mark entry field may be set below a character entry field. Fields and mark entry fields may be arranged in a cat. In addition, mark witness fields may be gathered in a specific area, and furthermore, virtual mark entry areas are set above and below the text entry field without actually displaying the mark entry field with lines on the form. As shown in Figure 4,
There is also a method of filling in marks by penetrating the character entry field 11 up and down.

[Effects of the Invention] As explained above, according to the optical character reading device of the present invention, even if a character is written incorrectly on a form, the correct character can be rewritten in the character entry field and the mark entry field can be corrected. Character recognition can be performed normally just by placing a mark at the desired position.

[Brief explanation of the drawing]

FIG. 1 is a diagram for explaining a form format adopted in an optical character reading device according to an embodiment of the present invention, and FIG. 2 is a diagram for explaining the configuration of the optical character reading device in FIG. 1. FIG. 3 is a block diagram showing a flowchart I for explaining the operation of the optical character reading device shown in FIG. 1, and FIG. 4 is a diagram showing another mark writing method according to the present invention. 1... Image reading section, 2. Character cutting section, 3. Character recognition section, 4. Knowledge dictionary, 5. Knowledge processing section. Applicant: Toshiba Corporation

Claims (1)

[Scope of Claims] Image reading means for reading an image of a form having a character entry field in which characters are written and a mark entry field in which a mark for indicating a typographical error on the character entry field is written, and the mark entry field. mark detection means for detecting the mark from the result of reading an image of the field; mark position determination means for determining the position of the mark detected by the mark detection means on the mark entry field; 2. An optical character reading device comprising: character recognition means for performing character recognition while excluding from processing a character on the character entry field corresponding to a position of the mark on the mark entry field.