JPH0728945A - Character recognition device - Google Patents

Abstract

PURPOSE:To provide a character recognition device which can correctly recognize the characters even when an original is set with its vertical direction reversed or the front and the back of a transparent original is reversed. CONSTITUTION:A character recognition device consists of an image reading part 1 which reads the image data, a page image storing part 2 which stores the image data in a single page, a character image segmenting part 3 which segments the image data on a single character from the image data on a single page, a character image storing part 4 which stores the image data on a single character, an image data vertically rotating part 5 which vertically rotates the image data by 180 deg., an image data upper/rear side turning part 6 which reverses the front and the back of the image data, a recognition dictionary 7 which stores the feature information on characters, a character recognition part 8 which obtains the candidate characters and the resemblance degree of the image data, and an original vertical and front/back direction deciding 9 for judging in the vertical and front/back direction of the original.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition device for obtaining character data from input image data, and more particularly to a preprocessing thereof.

[0002]

2. Description of the Related Art Conventionally, in this type of character recognition device, characters such as newspapers, magazines, papers, dot characters and handwritten characters are read as image data by an optical reading device such as a scanner, and the characters are separated into JIS characters. It was converted into code information and output.

[0003]

However, in the conventional character recognition device, when the document is set on the character recognition device, the characters cannot be correctly recognized unless the document is vertically oriented. Further, an original written on a transparent sheet such as an OHP sheet cannot be correctly recognized unless the front and back sides are aligned vertically.

The present invention solves such a conventional problem, and can correctly recognize a document set upside down in a character recognition device, and a transparent document sheet is set upside down. It is an object of the present invention to provide a character recognition device that can be correctly recognized even in case of failure.

[0005]

In order to achieve the above object, the present invention provides a conventional character recognition device with an image data vertical rotation unit for rotating image data up and down by 180 degrees, and an image in which the image data is turned upside down. A data front / back surface rotation unit and a document top / bottom front / back surface determination unit that determines the top / bottom front / back direction of the document are added.

[0006]

According to the present invention, since the user does not need to consider the up and down direction of the original set on the character recognition device and also the front and back sides of the original in the case of a transparent original sheet, the present invention does not require character recognition. It is possible to reduce the burden on the user during work.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a functional block diagram of a character recognition device according to an embodiment of the present invention. In FIG.
Reference numeral 1 is an image reading unit for reading image data, 2 is a page image storage unit for storing one page of image data input from the image reading unit 1, and 3 is a page image storage unit 2.
A character image cutout unit that cuts out one character image data from one page of image data stored in 4 is a character image storage unit that stores the one character image data obtained from the character image cutout unit 3, and 5 is a page image 1 stored in the storage unit 2
An image data vertical rotation unit that rotates the image data of a page and the image data of one character stored in the character image storage unit 4 up and down by 180 degrees, 6 is a 1 stored in the page image storage unit 2
An image data front / back rotation unit that turns the image data of a page and the image data of one character stored in the character image storage unit 4 upside down, 7 is a recognition dictionary storing character feature information, and 8 is a recognition dictionary 7. A character recognizing unit that refers to the character feature information and obtains the degree of similarity with the candidate character of the image data stored in the character image storage unit 9. Reference numeral 9 represents the upper and lower sides of the document based on the degree of similarity with the candidate character obtained by the character recognizing unit 8. A document top / bottom front / back determination unit that determines the direction.

FIG. 2 is a circuit block diagram of a character recognition device according to an embodiment of the present invention. In FIG. 2, 10 is a random access memory (hereinafter abbreviated as RAM),
Reference numeral 11 is a read only memory (hereinafter abbreviated as ROM), 12 is a central processing unit (hereinafter abbreviated as CPU), and 13 is a scanner.

The image reading unit 1 shown in FIG. 1 uses the scanner 13 to store a page image storage unit 2 and a character image storage unit 4.
The RAM 10 causes the character image cutout unit 3, the image data up / down rotation unit 5, the image data front / back rotation unit 6, and the document up / down front / back determination unit 9 to execute the program stored in the ROM 11 while the CPU 12 exchanges data with the RAM 10. As a result, the recognition dictionary 7 is executed by the ROM 11, and the character recognition unit 8 is executed by the CPU 12 in the ROM 11 and RA.
It is realized by executing the program stored in the ROM 11 while exchanging data with the M10.

The operation of the character recognition device according to the embodiment of the present invention will be described below with reference to the flow chart shown in FIG.

First, in step A, the user uses the image reading unit 1 to read image data. The input image data of one page is stored in the page image storage unit 2.

In step B, the character image cutout unit 3 cuts out image data of one character from the image data of one page stored in the page image storage unit 2. Got 1
The character data is stored in the character image storage unit 4.

In step C, the image data vertical rotation unit 5 creates vertically rotated character image data from the original text image data stored in the character image storage unit 4. The image data obverse / reverse rotation unit 6 creates obverse / reverse rotated character image data from the original text image data stored in the character image storage unit 4, and makes up / down / obverse / reverse rotated character image data from the vertically rotated character image data.

In step D, the character recognizing unit 8 obtains the similarity of each candidate character from the original sentence image data, the vertically rotated character image data, the obverse / reverse rotated character image data, and the up / down / obverse / reverse rotated character image data.

In step E, the document top / bottom front / back surface determination section 9 determines whether or not each candidate character having the highest similarity obtained in step D is an inappropriate character. If the character is inappropriate, the process returns to step B. If the character is not inappropriate, the process proceeds to step F.

In step F, the document top / bottom front / back determination unit 9 determines the image data having the highest degree of similarity obtained in step D. In the case of original text image data, the process proceeds to step K. In the case of vertically rotated character image data, the process proceeds to step G.
In the case of front and back rotated character image data, go to Step I. In the case of up / down / front / back rotated character image data, the process proceeds to steps H and J.

In steps G and H, the image data vertical rotation unit 5 rotates the image data of one page stored in the page image storage unit 2 up and down 180 degrees.

In step I and step J, the image data front / back rotation unit 6 turns over the front and back of the image data of one page stored in the page image storage unit 2.

In step K, the character recognition unit 8 character-recognizes one page of image data stored in the page image storage unit 2.

Hereinafter, how the characters are recognized in one embodiment of the present invention will be described with reference to FIGS. 4, 6, 7, 8, 9 and 1.
This will be described in more detail using 1. Fig. 4 shows the original up and down 1
It shows a state of rotating by 80 degrees. FIG. 5 shows a state in which the front and back sides of a transparent original are turned upside down. FIG. 6 shows a state in which a transparent document is turned upside down and then turned upside down. FIG. 7 shows the scanner 1.
An example of a character image obtained by cutting out the character image from the page image data read from No. 3 is shown. FIG. 8 shows a vertically rotated character image obtained by rotating the original text image shown in FIG. 7 up and down 180 degrees. FIG. 9 shows a front and back rotated character image obtained by turning the original text image shown in FIG. 7 upside down. Indicates. FIG. 10 shows the upside-down / upside-down rotated character image obtained by turning the upside-down rotated character image shown in FIG. The results of character recognition of the image data shown in FIGS. 7, 8, 9, and 10 are shown in FIG.

As a result of character recognition from FIG. 11, the candidate character having the highest similarity is "B". When the candidate character obtained here is a character such as "A" or "M" whose left and right are almost symmetrical, "C", or when it is a character such as "E" which is almost vertically symmetrical "Z", " When there is a character that is a point object such as "N", when there is a character that is a point object such as "n", it is almost object such as "b", "d", "p" and "q" In the case where there is a proper character, these characters are inappropriate for determining the orientation of the document, so the process returns to the process of cutting out the character from the page image data again.

From FIG. 11, the image data having the highest degree of similarity as a result of character recognition is the "vertical / upside-down rotated character image". For this reason, it is determined that the original set on the scanner 13 is set upside down and the front and back are opposite. If the character recognition of the page image data is executed as it is, a normal result cannot be obtained.
After rotating once, the front and back sides are turned over and character recognition is performed.

[0023]

As is apparent from the above embodiment, the present invention has an image data vertical rotation unit that rotates image data 180 degrees up and down, an image data front and back rotation unit that reverses the front and back of the image data, and the upper and lower front and back sides of the document. Since it has an original top / bottom front / back determination unit that determines the direction, it is not necessary for the user to consider the vertical orientation of the original to be set in the character recognition device, and in the case of a transparent original sheet, it is also necessary to consider the front and back of the original. Since it is eliminated, the burden on the user at the time of character recognition work can be reduced.

[Brief description of drawings]

FIG. 1 is a functional block diagram of a character recognition device according to an embodiment of the present invention.

FIG. 2 is a circuit block diagram of a character recognition device according to an embodiment of the present invention.

FIG. 3 is a flowchart of a character recognition device according to an embodiment of the present invention.

FIG. 4 is a schematic diagram showing a state in which a document is rotated up and down according to an embodiment of the present invention.

FIG. 5 is a schematic diagram of a state in which the front and back of a document are rotated according to an embodiment of the present invention.

FIG. 6 is a schematic view showing a state in which the upper and lower sides and the front and back sides of a document are rotated in an embodiment of the present invention.

FIG. 7 is an original character image diagram according to an embodiment of the present invention.

FIG. 8 is a vertically rotated character image diagram according to an embodiment of the present invention.

FIG. 9 is a front and back rotated character image diagram in one embodiment of the present invention.

FIG. 10 is a top and bottom and front and back rotated character image diagram in one embodiment of the present invention.

FIG. 11 is a list diagram showing a relationship between a candidate character and similarity in one embodiment of the present invention.

[Explanation of symbols]

 1 image reading unit 2 page image storage unit 3 character image cutout unit 4 character image storage unit 5 image data vertical rotation unit 6 image data front / back rotation unit 7 recognition dictionary 8 character recognition unit 9 original top / bottom front / back determination unit 10 RAM 11 ROM 12 CPU 13 Scanner

Claims (1)

[Claims] 1. An image reading unit for reading image data, a page image storage unit for storing image data of one page input from the image reading unit, and a page image storage unit for storing one page of image data stored in the page image storage unit. A character image cutout unit that cuts out image data of one character from image data, a character image storage unit that stores image data of one character obtained from the character image cutout unit, and one page stored in the page image storage unit Image data and the image data vertical rotation unit that rotates the image data of one character stored in the character image storage unit up and down 180 degrees, the image data of one page stored in the page image storage unit, and the character image storage Image data front / back rotation unit that reverses the front and back of one character of image data stored in a copy, a recognition dictionary that stores characteristic information of characters, and the recognition word From the character recognition unit that refers to the characteristic information of the characters stored in the text and obtains the degree of similarity with the candidate character of the image data stored in the character image storage unit, and from the degree of similarity between the candidate character obtained by the character recognition unit and the degree of similarity A character recognition device including a document top / bottom front / back determination unit that determines the top / bottom / front side of a document.