JP3176588B2 - Handwritten character input conversion device, document creation device, and computer-readable recording medium - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a handwritten character input conversion device for converting a handwritten character string input by a tablet or the like into a desired kanji or kanji mixed character string, and a document creation device such as a word processor having the same. The present invention relates to a computer-readable recording medium that records a program and data for causing a computer to realize the function of the handwritten character input conversion device.

[0002]

2. Description of the Related Art With the spread of personal computers, word processors, and portable information devices, there is an opportunity to perform a so-called handwriting input in which a character string is input to an electronic device by directly writing characters using a tablet or a pen mouse. Has increased.

[0003] However, the recognition accuracy of this handwritten input is inferior to human judgment, and it is still more difficult to accurately input fine handwritten Chinese characters by hand. Also, the user has become accustomed to using a word processor capable of converting kana-kanji characters, and thus cannot always write kanji accurately.

In order to efficiently input handwritten characters in such a situation, a method has been considered in which kanji with a small number of strokes are directly input and kanji with a large number of strokes are input with a kana. For example, when a character string "Saturday" is to be input, the character string is input as "Tsuchyo" in handwriting input, and converted into "Saturday" by kana-kanji conversion. Hereinafter, a character string in which a part of a kanji word or an idiom that the user wants to input is input in kana (hiragana or katakana) will be referred to as “spelling”.

As one method for realizing this, a kanji in an input character string is temporarily converted into its kana-kana character, and a normal kana-kanji conversion dictionary is searched based on the kana character string to obtain a desired kanji or kanji spelling. A method of obtaining a character string is disclosed in, for example, JP-A-62-209667 and JP-A-6-251203.

[0006]

However, in these prior arts, for example, for a spelled character string such as "Todoyo", the characters "Todoyo", "Tachiyo", "Zuchiyo" are used. Kana character strings are created for all the readings associated with the kanji, and whether or not the character strings are registered in the dictionary is sequentially searched.

[0007] There are many Japanese kanji whose readings are close to 10 due to sound facilities and muddy sounds. Reading 1
If there are two kanji that make up a idiom, there will be 100 readings that can correspond to this kanji, and searching for kanji candidates for all these readings will take a considerable amount of time. Finding the desired Kanji conversion from the candidates is difficult .

The present invention has been made in view of the above-mentioned problem, and narrows a spelled-out character string by handwriting input to a sufficiently reliable one at the stage of expanding the character string into a kana character string, thereby efficiently performing kana-kanji conversion. It is an object of the present invention to enable high-speed conversion to a desired kanji or kanji mixed character string.

[0009]

In order to achieve the above object, a handwritten character input conversion device according to the present invention has a handwritten character input means for inputting handwritten characters, and a character recognition device for recognizing characters input by the means. Means, and when the series of input character strings recognized by the means is a kanji word or a kanji character string in which a part of a idiom is input in kana, the kanji in the character string is read as a candidate kana character A connection conversion dictionary for registering connection conditions before and after each available kanji reading, and a connection candidate from the reading candidates converted by the reading conversion unit using the dictionary. Kana-kanji conversion means for converting a part to be a kanji in a kana character string including a kana character of the kana character selected by the means to a kanji, It includes those were.

[0010] By this, when expanding the kanji in the spelled-out character string into kana characters corresponding to the reading, it is possible to narrow down to kanji that are sufficiently reliable by selecting only reading candidates that match the connection conditions. , The subsequent kana-kanji conversion can be performed efficiently. The invention also provides
A document creation device such as a word processor provided with the above handwritten character input conversion device is also provided.

[0011] The present invention further relates to a character recognition means for recognizing an input handwritten character, and a series of input character strings recognized by the means for inputting a kanji word or a part of a phrasal word using a kana. When it is a written character string, using a reading conversion means that converts kanji in the character string into kana characters of reading candidates, using a connection condition dictionary that registers connection conditions before and after each available kanji reading A reading candidate selection unit that selects a reading candidate that matches a connection condition from the reading candidates converted by the reading conversion unit, and a kanji character in a kana character string that includes a kana character of the reading candidate selected by the unit. A program for functioning as a kana-kanji conversion means for converting a part into a kanji, and a computer-readable recording medium recording data of the connection condition dictionary are also provided. Subjected to.

[0012]

Preferred embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 2 is a block diagram showing a schematic configuration of an information processing apparatus such as a word processor (document creation apparatus) or a personal computer according to an embodiment of the present invention. This information processing device has the function of the handwritten character input conversion device according to the present invention from the beginning, or has the function by reading and loading an application program or the like recorded on a recording medium. .

In FIG. 2, a CPU 1 is a central processing unit for performing various processes including a handwritten character input conversion process according to the present invention. The ROM 2 is a program memory that stores various control programs executed by the CPU 1 and fixed data.

The RAM 3 is a large-capacity data memory.
A hard disk device (HDD) is used as a working area when various processes are executed by the CPU 1.
4 is also used to load necessary programs, dictionary data, and the like, and to store document data before and after processing. In the case of a general-purpose device such as a personal computer, various application programs (including a handwritten character input conversion processing program for executing the present invention) loaded from the outside into the hard disk device 4 are loaded at the time of actual use. Also used to

The hard disk drive (HDD) 4 is a large-capacity nonvolatile storage device using a magnetic disk as a storage medium, and stores various application programs as described above, as well as a handwritten character input conversion process according to the present invention. It also stores data of various dictionaries such as a recognition dictionary, a single kanji dictionary, a connection condition dictionary, and a kana-kanji conversion dictionary to be used.

The keyboard 5 has a large number of keys, is connected to a pointing device 6 such as a mouse, and is input means for inputting various data and commands by an operation of an operator. Various characters from the keyboard, to input symbols, etc., to create a document, is also able to edit or calibration
Kill, but in the case of all input by filter Brett 12 to be described later can be omitted.

The tablet 12 includes a touch panel and a touch pen. By writing characters in a predetermined area of the touch panel with the touch pen, various characters such as kana, kanji, numbers, and alphabets can be input by hand. This is a known handwritten character input means.

The display 13 is a display means such as a CRT or an LCD, and displays an input character string, a character string of a conversion result, and various messages. This is one of the output means. Of course, as another output means, a communication means (not shown) for transmitting data such as a created document to the outside may be provided, or a printer may be connected. . The ODD 7 is an optical disk device that can read various application programs and data recorded on an optical disk 8 as a recording medium and load the application programs and data on the hard disk device 4.

FDD 9 is a floppy disk drive,
It is possible to read various application programs and document data recorded on the floppy disk 10 as a recording medium and load them on the hard disk device 4 or the RAM 3 or record (store) the created document data on the floppy disk 10. it can. The above components are mutually connected by a bus 11. The illustration of each interface circuit is omitted.

Next, the function of the information processing apparatus as the handwritten character input conversion apparatus according to the present invention will be described.
This will be described with reference to FIG. FIG. 1 is a block diagram showing a functional configuration as a handwritten character input conversion device.

This device is similar to the tablet 12 shown in FIG.
When a handwritten character is input by the handwritten character input unit 21 corresponding to the character string, the character recognition unit 22 recognizes the character using the recognition dictionary 31 and converts the character into a character code. This character recognition technique is known. Then, the spelling determining means 23 determines whether or not the recognized series of input character strings is a spelled character string in which a part of a kanji word or idiom is input with a kana.

Although this can be automatically discriminated from the recognized character string, it is difficult to distinguish this kanji from a kanji that has been input properly and a kanji kanji character string such as kanagana. There is a possibility that it will be determined to be a sword. In order to prevent this, when inputting a conversion instruction for an input character string from the keyboard 6 or the tablet 12 in FIG. Can be determined.

When it is determined that the recognized input character string is a spelling, the reading conversion means 24 stores a single kanji dictionary 3 in which all available kanji reading candidates are stored.
Using 2, the kanji in the character string is converted to a kana character as a reading candidate. Further, the reading candidate selection unit 25 selects only reading candidates that match the connection conditions from the plurality of reading candidates. At this time, a connection condition dictionary 33 in which connection conditions before and after each available kanji reading are registered is used. This connection condition will be described later.

Thereafter, the kana-kanji conversion means 26
Using the kana-kanji conversion dictionary 34, the part of the kana character string that includes the kana character of the selected reading candidate is converted into a kanji character to obtain a desired kanji mixed character string. This is based on a normal kana-kanji conversion technology and has a candidate selection function and a learning function.

The character string resulting from the conversion is output by the conversion result output means 27. This is, for example, the display 13 of FIG.
To be displayed. If it is determined that the input character string is not a kana character string but a kana character string, the kana character string is sent directly to the kana / kanji conversion means 26, and the kana character string Convert the part that should be kanji inside to kanji.

When the input character string is not a spelling, but is all kanji or characters that do not require conversion (such as kanji and its sending, numbers, alphabets, foreign words, katakana, etc.) (for example, it is determined by an instruction of no conversion) Is output by the conversion result output means 27 as it is.

Next, this handwritten character input conversion processing is shown in FIG.
3 to 5 show the flow of processing by the CPU 1 when the information processing apparatus such as a personal computer shown in FIG.
This will be described with reference to the flowchart of FIG. In these figures, steps are abbreviated as “S”. FIG. 3 is a flowchart showing a main routine of the handwritten character input conversion process.

When this process is started, first, in step 1, a character input from the tablet 12 shown in FIG. 2 is recognized, and the recognized character (character code) is stepped.
In step 2, the data is sequentially stored in a text buffer (an area in the RAM 3). Then, in step 3, it is determined whether or not there is a conversion instruction from the tablet 12, and the processing in steps 1 and 2 is repeated until there is a conversion instruction.

If there is a conversion instruction, the process proceeds to step 4 to read an unconfirmed character string input so far from the text buffer, and in step 5 whether the input character string is a "seriously written" character string or not. Is determined. If it is a character string, the process proceeds to step 6 to perform a reading conversion process, and converts kanji in the character string into kana characters as reading candidates. The details will be described later.

Subsequently, in step 7, ordinary kana-kanji conversion is performed, and a portion to be converted into a kanji in a kana character string including the kana character as a reading candidate is converted into a kanji. And step
In step 8, a validity determination process is performed to obtain a desired kanji mixed character string. This legitimacy determination process is a process of excluding kanji candidates that cannot be converted from the input spelled out character string from conversion candidates, the details of which will be described later. In step 9, the result of the conversion is output. At this time, if there are a plurality of kana-kanji-converted kanji candidates, the first candidate is displayed so that it can be sequentially changed to another candidate according to a user's instruction, and the user selects the intended kanji.

If it is determined in step 5 that the input character string is not skewed, the process proceeds to step 10 in which all the character strings are kanji characters or characters that do not need to be converted (kanji characters and their characters, numbers, alphabets, foreign words, etc.). of and discrimination between katakana, etc.) to determine (e.g., by an instruction of the non-conversion), and if so, it the string stearate
In step 9, the data is output as a conversion result. Otherwise, the character string is determined to be all kana character strings, and step 7
Proceed to perform normal kana-kanji conversion.

Next, the contents of the subroutine of the reading conversion process in step 6 will be described with reference to the flowchart of FIG. In this subroutine, first, at step 11, it is determined whether or not all the kanji in the input "magic character string" have been converted. If all the kanji have been converted, the procedure shown in FIG. Return to the main routine. Otherwise, the process proceeds to step 12, where the kanji to be converted is set to the next kanji, but the first kanji in the input order of the characters is set to be the first kanji.

Then, in step 13, the single kanji dictionary 32 shown in FIG. 1 is searched, and if there is any unconverted reading of the kanji to be converted, the process proceeds to step 14.
If not, return to step 11. In step 14, FIG.
The connection condition dictionary 33 shown in (1) is searched to determine whether there is information on connection conditions corresponding to the reading. As a result, if there is information on the connection condition corresponding to the reading, the process proceeds to step 15 to add the reading (one or a plurality of kana characters) to the reading candidate, and then returns to step 13. If not, the process returns to step S13 .

In step 13, steps 14 and 1 are repeated until there are no unconverted readings for the kanji to be converted.
5 is repeated, and when there is no more, the process returns to step 11, and
If there is any kanji that has not been converted, the conversion target is set to the next kanji in step 12, and the processing in steps 13 to 15 is repeated. When all the kanji in the spelled character string have been converted to kana characters as reading candidates, the process returns to the main routine.

[0035] Next, stearyl in the main routine shown in FIG. 3
The contents of the subroutine of the validity determination processing of-up. 8, and FIG. 5
This will be described with reference to the flowchart of FIG. In this routine, first, at step 21, it is checked whether or not there is a conversion candidate kanji for which the validity has not yet been determined. If not, we have determined the validity of all conversion candidates,
It returns to the main routine of FIG.

[0036] If there is a conversion candidate that is not to determine the validity, but the check subject to the following conversion candidate proceeds to step 22, first time will be the first of the conversion candidates. And
In step 23, the original character string (the input spelling character string) is compared with the conversion candidate. As a result, if there is a character (kanji) inconsistent with the original character string, it is excluded from the candidates in step 26, and if not, the candidates (kanji) are sorted in character string order in step 25. And in each case
Returning to step 21, if there is a next conversion candidate, the above processing is repeated.

[0037] Next, a description will be given of the connection conditions dictionary. The connection condition dictionary stores connection conditions in the form of “(connection condition) reading (connection condition)” for all the readings of all registered kanji. For example, FIG. 6 shows a connection condition for the kanji “pay”. Each symbol indicating the connection condition has the following meaning. “/” Indicates a reading break.

+ Any one character [] Only specific characters in [] [] can be connected. * Characters may or may not be present. 読 み Reading always comes at the top. In addition, under the connection conditions for each reading in FIG. 6, examples of idioms that are read by readings that correspond to the connection conditions are also shown.

When an input such as "payout" is given by this connection condition dictionary, for example,
"Barashoku" (there is no character in front of the "payment"), (there is no "Tsu" before the "payment") "Parashoku", "Harashoku"
There is no need to create a reading for (there is no "Iueo" after "Payment").

That is, when "payout" is read and converted to be expanded into "kana character string", the connection information shown in FIG.
We can narrow down to three of "Futsuku". If there is no connection information, the expansion result is seven. When this expanded “kana character string” is converted to kana-kanji, the result is as follows. (1) “Fishoku” → “wiping” 1 phrase (2) “Haraishoku” → “payment / color” 2 phrases (3) “Fushishoku” → “Buddha / color” 2 phrases

The validity of this conversion result is checked. In other words, the original input character string “Passho” is compared with the kanji conversion result, and the correct one is selected. In this example, “wiping” is selected. When it is not possible to judge only by the character face, the one with less number of phrases is given priority.

Although a tablet is used as the handwritten character input means in this embodiment, other devices may be used as long as they can input kana and kanji. Scan a handwritten character string handwritten on paper with a scanner,
The handwritten character input conversion processing according to the present invention can also be performed using character string data that has been character-recognized by the R processing or a character string that has already been partially converted to kanji as input data.

Further, the present invention can be applied not only to a document creation apparatus but also to input of handwritten characters in various portable information devices and terminal devices. In addition, the following computer-readable recording medium is provided so that a general-purpose personal computer or the like can be used as the handwritten character input conversion device according to the present invention or as a document creation device having the function.

That is, a character recognizing means for recognizing an input handwritten character, and when a series of input character strings recognized by the means is a spelled character string, kanji in the character string is read as a kana character as a candidate for reading. Reading conversion means for converting to
Using a connection condition dictionary in which connection conditions before and after each available kanji reading are registered, and selecting a reading candidate that matches the connection condition from the reading candidates converted by the reading conversion device; and A program for causing a computer to function as a kana-kanji conversion means for converting a portion to be a kanji in a kana character string including a kana character of a reading candidate selected by the means into a kanji, and a data for the connection condition dictionary, The information is provided by being recorded on an optical disk or a floppy disk which is a readable recording medium.

[0045]

According to the present invention, a kanji character string obtained by handwriting input is narrowed down to a sufficiently reliable one at the stage of developing it into a kana character string, and kana-kanji conversion is performed efficiently to achieve a desired kanji or kanji mixing. Can be converted to character strings at high speed. Therefore, the input accuracy and input efficiency of the handwritten kanji mixed character string can be greatly improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a functional configuration as a handwritten character input conversion processing device by the information processing device shown in FIG. 2;

FIG. 2 is a block diagram illustrating a schematic configuration of an information processing apparatus such as a word processor (document creation apparatus) or a personal computer according to an embodiment of the present invention.

By the information processing apparatus shown in FIG. 3 FIG. 2 a flow chart showing a main routine of the processing performed by the CPU1 when performing handwriting input converter according to the present invention Oh
You.

FIG. 4 is a flowchart showing a subroutine of a reading conversion process in FIG. 3;

FIG. 5 is a flowchart illustrating a subroutine of a validity determination process in FIG. 3;

FIG. 6 is a diagram showing an example of connection conditions in a connection condition dictionary used in step 14 of the reading conversion process of FIG.

[Explanation of symbols]

1: CPU 2: ROM 3: RAM 4: Hard disk drive (HDD) 5: Keyboard 6: Pointing device 7: Optical disk drive (ODD) 8: Optical disk (recording medium) 9: Floppy disk drive (FDD) 10: Floppy disk (Recording medium) 11: Bus 12: Tablet (handwritten character input means) 13; Display (conversion result output means)

Continuation of the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) G06F 17/22 G06F 3/03 380 G06K 9/62, 9/72

Claims (3)

(57) [Claims] 1. A handwritten character input means for inputting a handwritten character; a character recognition means for recognizing a character input by the means; and a series of input character strings recognized by the means being one of kanji words or idioms. If the part is a kanji character string entered in kana, read conversion means for converting kanji in the character string to kana characters of reading candidates, and connection conditions before and after each available kanji reading A registered connection condition dictionary, reading candidate selection means for selecting a reading candidate that matches a connection condition from the reading candidates converted by the reading conversion means using the dictionary, and a reading candidate selected by the means. A handwritten character input conversion device, comprising: a kana-kanji conversion means for converting a part to be a kanji in a kana character string including a character into a kanji. 2. A document creation device comprising the handwritten character input conversion device according to claim 1. 3. A computer, comprising: a character recognizing means for recognizing an input handwritten character; a series of input character strings recognized by the means being a kanji character string in which a part of a kanji word or a idiom is input in kana; , A reading conversion means for converting kanji in the character string into kana characters of reading candidates,
Using a connection condition dictionary in which connection conditions before and after each available kanji reading are registered, a reading candidate selection unit that selects a reading candidate that matches the connection condition from the reading candidates converted by the reading conversion unit, A program for functioning as a kana-kanji conversion means for converting a portion to be a kanji in a kana character string containing a kana character of a reading candidate selected by the means into a kanji, and a computer recording the data of the connection condition dictionary A readable recording medium.