JPH01270183A - Character recognition device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a character recognition device that searches for a character area from a character image by inputting characters written on paper as an image, and converts it into a code number. .

[Conventional technology]

In recent years, rapid advances in character recognition devices have made it possible to recognize characters from various document images and automatically create document files. It is now possible to accurately recognize characters.

Document images include, for example, vertically written documents such as Japanese, and horizontally written documents such as English.

In order to recognize these documents, the user provides document information in advance, but the format of the document to be recognized is limited.

[Problem to be solved by the invention]

However, if the user has to provide the format information of the document image to the character recognition device each time it is used, or if the recognition formats are limited, the operability is very poor when actually using the recognition device. .

SUMMARY OF THE INVENTION The present invention is intended to solve these problems, and an object of the present invention is to provide a character recognition device that automatically determines the format from a document image of unknown format and recognizes the document.

[Means to solve the problem]

The present invention provides (1) a format determining means in a character recognition device that reads a character image written on a paper surface etc. using an optical image input means, recognizes the written character on the paper surface, places it in a code number, and calculates the IQ; Further, the format determining means is characterized in that the format determination means determines the format of the input image by determining the horizontal and vertical peripheral distribution of the input image.

It is also characterized in that the existence and size of blank areas in a plurality of peripheral distributions of one character image are used as a means of judgment.

〔Example〕

The present invention will be described in detail below based on examples.

The character recognition device of the present invention includes a CP tJ 1 , an image input device 2, a J2 character display device 3, a ROM 4, and a RAM 5, as shown in the block diagram of FIG.

The operation of the character recognition device of the present invention is as shown in the block diagram of FIG. It consists of a character string extraction means 8 for extracting a character string, a character extraction means 9 for extracting characters from the character string, a character recognition means 10 for recognizing the extracted characters, and a recognition result display means 11 for displaying the recognized characters. There is.

Hereinafter, a series of operations for determining a format from an input image will be explained in detail based on the flowchart shown in FIG.

In the image input means 6, the image input device 2
Character images 12 and 13 are read into RAM 5. It is unclear whether the data read into the RAM 5 is a vertically written document 12 such as Japanese or a horizontally written document 13 such as English, and it is impossible for the character string extraction means 8 to extract the character string as it is. Here, the present invention provides a format determining means 7 (18). The format judgment means 7 (
In 18), the horizontal peripheral distributions 14 and 15 and the vertical peripheral distribution 1 of the input character images 12 and 13 are
If the input character image is 12 shown in Figure 3(b), the horizontal peripheral distribution 14 and the vertical peripheral distribution 16 are calculated as shown in Figure 3(C) and 6.17. The result is as shown in FIG. 3(a). Here, we will look at the vertical peripheral distribution 16.

If the input character image is a letter-writing document, there must be a blank area (corresponding to the space between lines) in the vertical peripheral distribution 16, and if there is no such blank area, it should be regarded as a horizontally-written document. I can do it.

The marginal distribution 16 includes blank areas that may be between lines. Therefore, there is a possibility that the input character image 12 is a fir writing document. However, with this alone, it is impossible to regard the input character stroke @12 as an f7 writing document. Next, look at the horizontal peripheral distribution 14. If there is no blank space in the horizontal peripheral distribution 14, it can be determined that the input character image is a vertically written document,
Further, if a blank portion exists in the peripheral distribution 14, it can be determined that there is a possibility that the input character image is a horizontally written document. In FIG. 3(b), the vertical peripheral distribution 16
, a blank portion exists in both of the horizontal peripheral distributions 14. Therefore, both possibilities exist. Therefore, the sizes of the blank areas are compared.

In a typical document, when blank areas exist in both the horizontal and vertical peripheral distributions, the blank areas in the row direction are larger than the blank areas in other directions. Comparing the sizes of the blank parts of marginal distributions 14 and 16 in Figure 3(b), the blank part of the vertical marginal distribution is larger than the blank part of the vertical marginal distribution.
Therefore, it can be determined that the input character image 12 is a fir writing document.

Furthermore, even if the input image is 13, which is not shown in FIG. 3(e), the format may be determined in the same manner. Fortunately, there is no blank area in the vertical peripheral distribution 17 of the input image 13 shown in FIG. 3(d). If it is a vertically written document, there will definitely be a blank space, so it can be determined that the input character image 13 is a horizontally written document.

As described above, since the peripheral distribution of the input image in the horizontal and vertical directions is calculated and the judgment is made by focusing on the blank areas, it is possible to easily and reliably judge the format of the input image.

Furthermore, since the format is automatically determined from the input character image, the trouble of specifying the format when inputting the image is eliminated, and operability is improved.

Furthermore, although the present invention has described the case where the format of the character image is horizontal and vertical, the method is not limited to horizontal and vertical formats, and if characters are written in parallel, the format is diagonal. It is possible to judge the format by finding the peripheral distribution of characters in all directions (within a range of 180 degrees).

〔Effect of the invention〕

As described above, according to the present invention, the character recognition device includes a format determination means for determining whether the input character image is a horizontally written document or a Japanese-style document. There is no need to specify format before or after, and
As a format determination means, the peripheral distribution in the horizontal and vertical directions is taken, and the determination is made by focusing on blank areas, making it possible to easily and reliably determine the format of the input image.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a character recognition device of the present invention. FIG. 2 is a block diagram of the recognition means of the present invention. FIGS. 3(a), 3(b), 3(c), 3(d), 3(e), and 3(f) are views showing the format determining means of the present invention. FIG. 4 shows a flowchart of format determination according to the present invention. 1... CPU 2... Image input device 3... Recognition result display device 4... ROM 5... RAM 6... Image input means 7...Format determination means 8...Character string extraction means 9...Character extraction means 10...Character recognition means 11...Recognized character display means 12.13・・Character image■・1.15・・Horizontal peripheral distribution 16.17・・Vertical peripheral distribution 18・・・Format judgment means and above Applicant Seiko Epson Corporation Representative Patent attorney Masayoshi Kamiyanagi (and others) 1 person) Figure 3 (I
l) Figure 3 (j)

Claims (2)

[Claims] (1) A character recognition device that reads a character image written on paper etc. using an optical image input means, recognizes the written character on the paper, and replaces it with a code number, characterized by having a format judgment means. character recognition device. (2) Claim 1 characterized in that the presence or absence and size of blank areas in a plurality of peripheral distributions of a character image are used as a means of determination.
The character recognition device described.