JP2933178B2 - Character recognition method for optical character reader - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition method for an optical character reader for mainly recognizing first and last names of handwritten Chinese characters.

[0002]

2. Description of the Related Art Today, kana-kanji conversion using a keyboard is mainly used as a Japanese input method. However, this input method requires a certain level of skill in inputting, and it is not possible for an unfamiliar person to input easily. Therefore, an optical character reading device (OCR: Optical Character Reading Device) is a means for directly inputting characters.
Reader)).

For example, in a financial institution or the like, a customer inputs his / her name, address, and the like by hand into a predetermined field provided on an OCR form, and the kanji OCR optically reads the name to recognize the name. In such a case, in the first and last name recognition of the kanji OCR, a space (a space for one character) is often used as a separator between a first name and a last name. This is called solid writing. For example, in the case of "Ichiro Suzuki", the surname is "Suzuki" and the first name is "Ichiro", which is written as Ichiro Suzuki. Conventionally, such solid text recognition has been limited to four and five characters.
In the case of letters, two words are used, and in the case of five letters, two words and three letters (for example, "Kototaro Takahashi") Three letters and two letters (for example, "Tokudaiji Kazuo") I was going.

[0004]

However, in the above-described conventional recognition method, for example, in the case of five-character recognition, if the correct reading can be performed in any case such as "Ichitaro Kaneda", the optical The result is determined according to the priority determined by the specifications of the character reader, and the operator cannot freely select the result.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned conventional problems, and allows the operator to freely select the recognition result of the first and last names in the case of solid writing, thereby improving the recognition rate and operability. It is an object of the present invention to provide a character recognition method in an optical character reader that can be improved.

[0006]

According to the present invention, there is provided a character recognition method in an optical character reader for recognizing a first name and a last name from an input first and last name. If the input is in a character state and there are different methods for separating the first and last names, the data is separated by the different methods for separating the first and last names, and candidate data for selecting these separating methods is displayed. .

[0007]

According to the character recognition method of the present invention, when a kanji of a first and last name is input in a solid state, it is checked whether there is a different method for separating the first and last names of the first and last names. , And display the candidate data. The operator selects a desired first and last name delimiter from the displayed candidate data and determines this. Therefore, it is possible to improve the recognition rate of the first and last names and the operability when the kanji of the first and last names are written in solid.

[0008]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of an optical character reader for implementing the character recognition method of the present invention. The illustrated device includes a control unit 1, a storage unit 2, an operation unit 3, and a reading unit 4. The control unit 1 includes a processor or the like,
A candidate data output unit 11 is provided. The candidate data output unit 11 has a function of performing a predetermined process based on the candidate data output from the reading unit 4 and outputting the candidate data to a display 31 of the operation unit 3 described later. The storage unit 2 is composed of a magnetic disk device or the like, and stores delimiter (reading) data corresponding to solid written first and last name kanji, programs used for various data processing, and the like. The operation unit 3 includes a display 31 and a keyboard 32. The display 31 is a display device that includes a CRT or the like and displays candidate data in accordance with an instruction from the candidate data output unit 11. The keyboard 32 includes an independent keyboard in which candidate keys, an enter key, and the like are assigned to PF keys and other keys, or a so-called touch key provided on the display 31. Has a function of selecting a desired first and last name delimiter. The reading unit 4 includes a scanner unit 41 and a recognition unit 42. The scanner unit 41 is a device that reads characters handwritten on paper, and a recognition unit 42
Recognizes the first and last name kanji read by the scanner unit 41,
It has a function of referring to the candidate data in the storage unit 2 to determine whether or not there is other candidate data corresponding to the first and last name kanji.

Next, a description will be given of a character recognition method in the above-mentioned optical character reading apparatus. First, the operator sets a predetermined OC.
Write your first and last name in R form etc. with kanji and furigana. Thereby, the scanner 41 reads the written kanji of the given name and surname, and the recognizing unit 42 transmits the recognition result, that is, the input kanji and data indicating whether or not there is another candidate corresponding to the kanji. To send to. The control unit 1 processes and outputs candidate data from the input recognition result data of the first and last names and kanji.

FIG. 2 is a flowchart showing candidate data processing by the control unit 1 and the reading unit 4. That is, the recognizing unit 42 of the reading unit 4 refers to the data in the storage unit 2 and determines whether or not there are a plurality of different names (reading methods) for the recognized data (step S1). In this step S1, if there is no candidate, the display 31
The first and last name fields displayed above are displayed in normal colors (step S2), and the process proceeds to normal processing. When there are a plurality of candidates, the first and last name fields on the display 31 are displayed in a warning color (step S3). The operator determines from the warning color display whether the recognition result displayed in the first and last name fields is sufficient, and presses a candidate key on the keyboard 32 when another candidate is to be displayed (step S4).
If the displayed recognition result is satisfactory, the process directly proceeds to the normal processing. When the candidate key is pressed, the candidate data output means 11 displays other candidates on the display 31 (step S5). In this case, there are two types of display formats: a method of replacing the recognition result displayed in the same field on the screen with the recognition result that has been displayed so far, and a method of displaying the recognition result in another place on the screen.

FIG. 3 shows two display methods. That is,
FIG. 3A shows a method of switching and displaying another candidate in the same field each time a candidate key is pressed, and FIG.
This is a method of displaying other candidates at independent positions on the screen.
When the candidates are displayed in step S5, the operator checks these candidates and determines which candidate is selected or whether none is selected (step S6). Here, if no candidate is selected, the process returns to step S3 and the warning color display of the character string is performed again. When the operator determines that the candidates on the screen are correct divisions, the operator presses the enter key (step S8). Thereby, the control unit 1 determines the character string of the first and last names, and displays the character string in the same field on the display 31 in a normal color (step S9).

For example, it is assumed that the kanji of the first and last names handwritten by the operator is "Ichitaro Kaneda". As a result, the control unit 1 performs the above-described processing, and as shown in FIG.
As a result of the recognition, the name is displayed as “Ichitaro Kaneda” and the reading as “Kaneda Ichitaro”. In this case, another candidate "Ichitaro Kaneda = Kindaiichi Taro"
This is displayed by pressing the candidate key. If the character string determined by the operator is “Ichitaro Kaneda = Kaneda Ichitaro”, the character string is displayed on the display 31 in a normal color.

FIG. 4 shows an example of such a recognition result. That is, the character string read by the reading unit 4 is “Ichitaro Kaneda”, and the recognition result of the control unit 1 is two types, that is, Ichitaro Kaneda and Ichitaro Kaneda. The character string determined by the operator is "Ichitaro Kaneda".

In the above embodiment, handwritten characters are recognized as a character recognition method. However, the present invention is not limited to this. For example, the same effect can be obtained even for solid printed characters. Can play. Further, in the above embodiment, the case where the first and last names are five characters has been described as an example.
Other than this, the same effect can be obtained if the first and last name kanji are three or more characters. Further, although not specifically mentioned in the above embodiment, it goes without saying that not only the display of the candidate data or the recognition result on the display 31 but also the transmission of the candidate data or the recognition result to a host device or the printing by a printer may be performed. It is.

[0015]

As described above, according to the character recognition method in the optical character reading device of the present invention, when the first and last name and kanji are written in solid, if the last name and the first name are different from each other, there is a difference between the last name and the first name. Since these candidate data are displayed and a desired delimiter of the first and last names is selected from these candidate data, the recognition rate of the first and last names and the operability can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram of an optical character reading device for implementing a character recognition method of the present invention.

FIG. 2 is a flowchart of a character recognition method according to the present invention.

FIG. 3 is an explanatory diagram of a screen display according to the present invention.

FIG. 4 is an explanatory diagram of a recognition result of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Control part 2 Storage part 3 Operation part 4 Reading part

Claims (1)

(57) [Claims] 1. A character recognition method in an optical character reader for recognizing first and last names from input first and last name characters,
If the kanji of the first and last names are entered in a continuous character state and there are different methods of separating the first and last names, they are separated by the different methods of separating the first and last names and display candidate data for selecting these separating methods. Character recognition method in an optical character reader.