JPH0636069A - Character recognizing device - Google Patents

Abstract

PURPOSE:To enable detailed specification and improve the recognition rate by using normal expression to specify the character kind of format control information(FC). CONSTITUTION:This character recognizing device is equipped with an FC part 2 where the format control information referred to so as to read characters, etc., recorded on a form is stored and information specifying the character kind of the format control information is expressed in the normal expression, a normal expression analytic part 31 which analyzes the normal expression in the format control information stored in the FC part 2, a character kind control part 34 which selects a dictionary to be used among recognition dictionaries 33 according to the analytic result of the normal expression analytic part 31, and a matching part 35 which obtains a read result by matching the characters recorded on the form with patterns by using the dictionary selected by the character kind control part 34.

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 reading characters recorded on paper based on format control information (FC) including a reading position, a font, a character type and the like.

[0002]

2. Description of the Related Art A conventional character recognition device (OCR) reads a character or the like recorded on a sheet based on format control information (FC) including a reading position, a font, a character type, etc. on a form or the like. Format control information (FC) is
For each character entry field (field) provided on the form,
It is designated as alphabet (A), number (N), katakana (K), and so on. Therefore, all the characters recorded in the field are recognized as the designated character type.

[0003]

As described above, in the conventional character recognition device, all the characters recorded in the field are recognized as the character type designated in the field. Therefore, for example, like the amount column, the first character may be the symbol “¥”, and the first number may not be “0”, which makes it impossible to specify detailed control. In this case, depending on the quality of the characters recorded in the amount column, the recognition result for the first number may be erroneously recognized as "0", leading to a reduction in the recognition rate.

The present invention has been made in view of the above points. By using a regular expression to specify the character type of format control information (FC), detailed specification is possible and the recognition rate is improved. An object of the present invention is to provide a character recognition device capable of performing the above.

[0005]

SUMMARY OF THE INVENTION The present invention is for storing format control information that is referred to for reading characters and the like recorded on paper, and specifies the character type in the format control information. Format control information storage means in which information is represented by a regular expression, regular expression analysis means for analyzing a regular expression in the format control information stored in the format control information storage means, and an analysis result by the regular expression analysis means On the basis of,
And a reading unit that obtains a reading result of the characters and the like recorded on the sheet.

[0006]

With this configuration, since the information regarding the character type included in the format control information is represented by the regular expression, it is possible to finely control each character to be recognized.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. 2 is a block diagram showing a schematic configuration of the character recognition apparatus according to the embodiment, and FIG. 1 is a block diagram showing a detailed configuration of the recognition unit 3 in FIG.

As shown in FIG. 2, the character recognition device comprises a control unit 1, an FC unit 2, a recognition unit 3, an output unit 4, and an input unit 5. The control unit 1 controls the entire device. The FC unit 2 is for storing format control information (FC) including a reading position, a font, a character type, etc., which is referred to when performing character recognition. The information on the character type of the format control information (FC) can be specified by a regular expression for each field indicating a character entry position provided on a form, for example. Details of the regular expression will be described later. The recognition unit 3 performs character recognition by cutting out a character pattern from an image obtained by optically scanning a form or the like by an input unit 5 described later according to the format control information (FC) stored in the FC unit 2. By reading the character. The output unit 4 outputs the reading result obtained by the character recognition by the recognition unit 3. The input unit 5 optically scans a form or the like on which characters or the like to be read are recorded to detect a form image and outputs it to the recognition unit 3.

As shown in FIG. 1, the recognition unit 3 includes a regular expression analysis unit 31, a cutout unit 32, a recognition dictionary 33, and a character type control unit 3.
4 and the collation unit 35. The regular expression analysis unit 31 uses the format control information (F
Input C) and parse the regular expression specified by this FC. The cutout unit 32 cuts out a character pattern for each character from the form image output from the input unit 5 based on the reading position information designated by the format control information (FC). The recognition dictionary 33 is for storing a dictionary for all recognizable characters. The character type control unit 34 selects which character type dictionary to use for character recognition for each character according to the analysis result by the regular expression analysis unit 31.
The collation unit 35 recognizes the character by collating the dictionary selected by the character type control unit 34 with the character pattern cut out by the cutout unit 32, and outputs the read result. Next, the operation of the embodiment will be described.

First, the FC unit 2 registers format control information (FC) relating to a form to be processed. The format control information (FC) includes information on a field position (reading position) which is a target area for reading characters provided on a form, a font of characters recorded in the field, a character type, and the like. The information about the character types is represented by regular expressions.

The regular expression is a method of expressing a pattern of a character string, and is expressed according to the rule of the regular expression as shown in FIG. 3, for example. Here, description will be given using an example of a regular expression for a character string entered in the amount column provided on the form.

For example, in the amount column, the first character may be the symbol "\", and the first number is a character other than "0". [1-9] [0-9] * (a). That is, in order from the left of the character string, "\" is 0 or 1, "1" to "9" is 1 and "0" to
This means that any one of “9” continues for 0 or more.

The form image detected by the input unit 5 is recognized by the recognition unit 3.
Format control information (FC) about the form to be processed, which is input to the cutout unit 32 of and stored in the FC unit 2.
Is input to the regular expression analysis unit 31 and the cutout unit 32 of the recognition unit.

The cutout unit 32 cuts out the character pattern based on the information indicating the field position and outputs the cut-out character pattern to the collation unit 35. Further, the regular expression analysis unit 31 analyzes the format control information (FC) regarding the field to be processed. The regular expression analysis unit 31
By analyzing the regular expression as shown in (a) above, it can be seen that the character entered in the first digit is either "\" or "1" to "9". In some cases, the character type control unit 34 is instructed to use only the dictionary for these characters. The collation unit 35 collates the dictionary selected by the character type control unit 34 with the character pattern of the first digit cut out by the cutout unit 32, and outputs the read result.

Next, the regular expression analysis unit 31 analyzes the above (a) again, and when the read result of the first digit is "\", the character entered in the second digit is " It can be seen that it is any of "1" to "9".

The reading result of the first digit is "1".
When it is any of "9", it is understood that the character entered in the second digit is any of "0" to "9".
In this way, the dictionary used for reading the second digit can be narrowed down depending on the reading result of the first digit. Further, it can be seen from the above (a) that a dictionary of "0" to "9" may be used for the third digit and thereafter.

As described above, since the detailed character type can be specified by the regular expression for the character string entered in the field, fine control can be performed on a character-by-character basis and the dictionary is limited for each character. Therefore, the recognition result can be obtained more reliably, and the processing time can be shortened by limiting the dictionary.

In the above embodiment, the character recognition is performed by narrowing down the dictionary based on the character type specified by the regular expression, but the same character is used for all the character strings recorded in the field. Read as a seed,
It may be referred to when checking the validity of the read result (post-processing).

Although the designation of the character type in the field has been described, it also has a function as an alternative to the designation of the character set. For example, when only "Y" and "N" are entered, it is sufficient to specify "NY" using a regular expression. Furthermore, even if only a limited number of words are entered, it can be specified by a regular expression. For example, Tokyo | Saitama |
Chiba | Kanagawa can be specified and used as a simple post-processing word dictionary.

[0020]

As described above, according to the present invention, since a detailed expression can be specified by using a regular expression for specifying the character type of format control information (FC), the recognition rate can be improved. It is possible.

[Brief description of drawings]

FIG. 1 is a block diagram showing a detailed configuration of a recognition unit 3 in FIG.

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

FIG. 3 is a diagram showing an example of a regular expression rule used in this embodiment.

[Explanation of symbols]

1 ... Control unit, 2 ... FC unit, 3 ... Recognition unit, 4 ... Output unit, 5
... input part, 31 ... regular expression analysis part, 32 ... cutout part, 33
... recognition dictionary, 34 ... character type control unit, 35 ... collation unit.

Claims (1)

[Claims] 1. A storage medium for storing format control information that is referred to for reading a character or the like recorded on a sheet, wherein information for designating a character type in the format control information is represented by a regular expression. A format control information storage means, a regular expression analysis means for analyzing a regular expression in the format control information stored in the format control information storage means, and a record on the sheet based on an analysis result by the regular expression analysis means. A character recognition device comprising: a reading unit that obtains a reading result of the written characters and the like.