JPH0528302A - Character reader - Google Patents

Abstract

PURPOSE:To read slip characters with a various kinds of printers having different fonts. CONSTITUTION:A scanning part 10 reads the characters of a slip printed in a prescribed font and stores the image data in an image memory 11. The data stored in the image memory 11 are cut out for each character by a character cut-out part 12, supplied to a decision part 13 and checked with a character dictionary of a dictionary memory part 14. The decision part 13 outputs the data which could be identified with a prescribed evaluation measure as read data. The identification is decided based on whether or not the reading of the characters printed on a learning field column provided on the slip ensures a prescribed identification precision and when the characters were not readable, a new character dictionary is prepared by utilizing the characters which were not readable and their image pattern in a dictionary preparation part 15 and added to the dictionary memory part 14.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character reading device, and more particularly to a character reading device capable of reading a character in a font which is not a reading target.

[0002]

2. Description of the Related Art As an example of a method for reading print characters,
There is a method of performing matching processing between an input character and a character dictionary registered in advance and extracting and identifying the one having the best degree of matching.

However, when reading characters in a font that is not the object of reading, it is necessary to create a dictionary corresponding to the font in advance and register it in the device.

[0004]

With the widespread use of personal computers in recent years, various types of printers connected to them and outputting a hard copy have been developed.

With the development of various printers, various types of printers have been developed, such as impact printers for dots and non-impact printers such as laser printers, and fonts for characters used in these printers. Is also in wide variety.

In the conventional character reading device, the font to be read is determined in advance. Therefore, when reading another font, it is necessary to create and register a dictionary for that, and it is possible to read characters of various printers. It had the drawback of being difficult.

An object of the present invention is to eliminate the above-mentioned drawbacks and to provide a character reading device capable of reading printed characters by various printers.

[0008]

The character reading apparatus of the present invention includes a learning field for printing a predetermined character in the same font as a printing character of a data field for printing data to be read and in a preset character order. Means and a first character dictionary for reading the form with the learning field by collation, and reading the form, based on the character patterns of all unreadable characters for the learning field and the character position. And a means for reading the form while incorporating a second character dictionary into which the unreadable characters can be read by inputting the character code data obtained by To be done.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

In FIG. 1, a scanning unit 10 scans a form or the like to be read and outputs its image pattern.

The image memory 11 is a memory for sequentially storing the image patterns output from the scanning section 10.

The character cutout unit 12 includes an image memory 11
The image pattern is cut out character by character based on the format information of the form.

The determination unit 13 identifies the cut-out character by character.

The dictionary memory unit 14 stores a character dictionary (standard pattern) for matching with the cut out characters (standard pattern). These dictionaries are created and stored in advance. And a portion for storing the second character dictionary newly created by the dictionary creating unit 15.

The dictionary creating unit 15 creates a second character dictionary from the character images output by the character cutting unit 12.

Next, the operation of this embodiment will be described.

FIG. 2 is a diagram showing an example of a form used in the embodiment of FIG.

In FIG. 2, a form identification code (hereinafter, I
A C code) 20 is used to specify the format of a form, that is, the number of lines to be read, their positions, character types, and the like.

The learning field 21 is a field for printing the characters used in the form, for example, the numbers 0 to 9 in order.

For example, if the font used in the form is the numbers of the new JIS-B font, the learning field 21 is printed with the numbers 0 to 9 of the new JIS-B in order.

In FIG. 1, the scanning unit 10 knows the form format such as the positions of the learning field 21 and the plurality of data fields 22 and the character type by reading the ID code 20 shown in FIG.

Next, the reading of the learning field 21 is started.

For example, as shown in FIG. 2, numbers 0 to 9 are printed in the learning field 21.

The characters read by the scanning unit 10 are
After being stored in the image memory 11, the character cutout unit 1
It is cut out by 2 and sent to the determination unit 13.

The determination unit 13 collates the cut out characters with the standard character pattern stored in the dictionary memory 14.

The characters printed in the learning field 21 are used for character reading, and when a character not to be read is input, it is used to create a new character dictionary necessary for reading the character not to be read. This means that the characters to be printed in the learning field are arranged in a preset character order, so if a certain character is unreadable, for example, from the position information where "2" exists in the above example, " This is possible by knowing the character code specifying 2 "and creating a new character dictionary using this and the character pattern that could not be read as a teacher signal.

When reading the form, the characters printed in the learning field 21 of the form are first identified.

At this time, if a highly accurate correct answer is obtained, it is determined that the dictionary of the character is already registered, and the subsequent reading of the data field 22 is executed.

On the other hand, if unreadable or erroneous reading occurs, or if the result is low in accuracy even if it is identified, it is determined that the character is not registered. If it is determined that the dictionary is not registered, the dictionary is automatically created from the printed character pattern of the learning field 21, and after the registration, the subsequent reading of the data field 22 is executed.

The characters printed in the learning field 21 are
Since the characters are printed in the determined character order, the correct answer can be easily understood, and if the identification result exceeds the predetermined judgment threshold value and is highly accurate, the subsequent reading of the data field 22 is immediately started.

However, if one of these characters cannot be identified, is misread, or has a low accuracy even if correctly read, it is determined that the character dictionary is not registered in the dictionary memory unit 14. To do.

At this time, for characters other than the correct answer with high accuracy, those character images are sent to the dictionary creating section 15,
Create those character dictionaries and register them in the dictionary memory unit 14,
It operates so as to proceed to the identification of the data field portion 22 thereafter.

A variety of algorithms for creating a dictionary have been published, but since they are not directly related to the present invention, their explanations are omitted.

The dictionary memory section 14 is divided into a section for constantly registering a dictionary of a font that is frequently used and a section for registering a dictionary newly created by the dictionary creating section 15.

When the area of the newly created dictionary in the dictionary memory unit 14 becomes full, it has a function of replacing the dictionary with a low frequency of use. If it is known in advance that the characters used in the form are fonts that are constantly registered in the dictionary memory unit 14, the learning field is left blank (blank), that is, blank. It is possible to control so as to omit the dictionary creation processing by detecting the.

Further, as shown in FIG. 3, even if there is a case that the lower part of the numeral 0 in FIG. The dictionary is automatically created and can be read.

[0038]

As described above, according to the present invention, the conventional character reading device can read only the characters of the predetermined font, but the dictionary corresponding to the characters of the font not to be read can be read. Can be created automatically, and it has the effect of enabling the reading of arbitrary fonts.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an exemplary embodiment of the present invention.

FIG. 2 is a diagram showing an example of a read form in the embodiment of FIG.

FIG. 3 is a diagram showing an example of a habit of print characters of a printer.

[Explanation of symbols]

 10 Scanning Section 11 Image Memory 12 Character Extraction Section 13 Judgment Section 14 Dictionary Memory Section 15 Dictionary Creation Section 20 Form Identification Code 21 Learning Field 22 Data Field

Claims (1)

Claims: 1. A means for providing a form with a learning field for printing predetermined characters in the same font as a printing character of a data field for printing data to be read and in a preset character order, and the learning field. While reading the form with a first character dictionary for reading the form by collation, the character patterns of all unreadable characters for the learning field, and the character code data obtained based on the character position. Characters comprising means for reading the form while creating a second character dictionary that makes it possible to read the unreadable characters by inputting and, and incorporating the second character dictionary with the first character dictionary. Reader.