JPH05128313A - Hand-written character recognition device - Google Patents

Abstract

PURPOSE:To improve the efficiency of hand-written input of a sentence including KANJI (Chinese character) together. CONSTITUTION:The reading or first some characters of a character string recognized by a tablet 1 unified with a display device and a hand-written character recognition part 2 are used as the index to store this character string in a KANA (Japanese syllabary)-word retrieval table 5, a KANJI-word retrieval table 6, and a KATAKANA (square form of Japanese syllabary)-word table 8. A KANJI-KANA conversion dictionary 4 is used for conversion from KANJI to reading. At the time of the next input, only the reading or first some characters of the character string are inputted to restore the corresponding character string from tables 5, 6, and 8 and the KANJI-KANA conversion dictionary 4; and thus, the character string including KANJI together is inputted with a small number of input characters.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a handwritten character recognition device, and more particularly to a technique for reducing an input load of an online handwritten character recognition device for immediately recognizing an input handwritten character.

[0002]

2. Description of the Related Art Online handwritten character recognition devices for recognizing handwritten characters immediately have become popular. An online handwritten character recognition device sequentially recognizes written characters such as kanji and hiragana based on handwriting information and captures them as an input character string. There is also an online handwritten character recognition device that recognizes only hiragana and converts it into kanji by using kana-kanji conversion technology to input a sentence.

[0003]

When inputting Japanese with such an online handwritten character recognition device, a function of inputting Chinese characters is indispensable. Inputting characters including Kanji by handwriting is as direct and convenient as writing a text by handwriting. However, in this case, it is necessary to fill in all the characters correctly, which causes physical fatigue. Especially when inputting Kanji, the input efficiency is not so high because there are characters with many strokes.

In order to solve such a problem, it has been attempted to input only hiragana by handwriting and use kana-kanji conversion technology. This method is advantageous in terms of fatigue because it is not necessary to input kanji. However, there is a problem in that it is not direct and somewhat inferior in usability because it takes the step of kana-kanji conversion.

Further, in the case of performing kana-kanji conversion, a dictionary for that purpose is required, and a considerable amount of capacity is required. This volume should be as small as possible. In particular, as for Katakana, etc., there are many that are newly started or not used, and it is not so useful to include them in the dictionary. However, it is preferable that the katakana language is input as efficiently as possible.
Further, in the conventional kana-kanji conversion, when there are a plurality of word candidates for the same reading, by pressing a specific key, the candidate words are sequentially switched and displayed, and the desired word is displayed. Sometimes I had to select by pressing another particular key. Such an operation is complicated in the handwritten character input device, and its improvement has been desired.

Therefore, an object of the invention described in claim 1 is to provide a handwritten character recognition device capable of inputting characters including Chinese characters with high efficiency.

An object of the invention described in claim 2 is to improve the efficiency of input and to reduce the capacity of the dictionary required for input.

A third object of the present invention is to provide a handwriting function which enables efficient input of words consisting only of katakana that are subject to particularly great fluctuation and are not good to include in the dictionary, without using the dictionary. It is to provide a character recognition device.

An object of the invention described in claim 4 is to provide a handwritten character recognition device capable of inputting characters including kanji with a small number of inputs without using kana-kanji conversion.

The object of the invention described in claim 5 is to efficiently input characters including kanji and katakana, and to input characters including kanji with a small number of input characters, among a plurality of candidates. It is to provide a handwritten character recognition device capable of easily selecting one of the above.

[0011]

A handwritten character recognizing device according to claim 1, wherein input means for inputting a handwritten character string and a predetermined signal, recognizing means for recognizing the input character string, and recognition. The converted character string to a kana character string,
A character string storage unit for storing the recognized character string in association with a predetermined number of Kana characters from the beginning of the converted Kana character string, and responding to the Kana character and a specific signal input by the input unit. Search the character string storage means,
The apparatus includes means for extracting a character string associated with a kana character that matches the input kana character, and display means for displaying the extracted character string.

A handwritten character recognition device according to a second aspect is
The handwritten character recognition device according to claim 1, wherein the character string storage unit stores the character string in association with each specific kana character string, and the given character string is associated with the corresponding kana character string. The maximum length of each kana character string of the dictionary is constant regardless of the length of each character string.

A handwritten character recognition apparatus according to claim 3 is
The handwritten character recognition device according to claim 1, further comprising:
A means for determining whether the recognized character string is a katakana character string, and a recognized kana character instead of the kana character string when the recognized character string is determined to be a katakana character string. A predetermined number of Katakana characters from the beginning of the column,
Katakana character string storage means for storing the input katakana character string in association with each other.

A handwritten character recognition apparatus according to claim 4 is
An input means for inputting a handwritten character string and a predetermined signal; a recognition means for recognizing the input character string;
A character string decomposing means for decomposing the recognized character string into words by detecting a word boundary of the recognized character string, and each word output by the character string decomposing means is a predetermined number from the beginning of each word. And a word storage means for storing in association with the character, and a word associated with the character that matches the input character by searching the word storage means in response to the character input by the input means and a specific signal. Includes means for retrieving and display means for displaying the retrieved word.

A handwritten character recognition device according to a fifth aspect of the present invention is the handwritten character recognition device according to any one of the first to fourth aspects, wherein the display means is associated with a predetermined input area of the input means. And a window display means for displaying the retrieved character string or word in a window, and the input means includes means for selecting and inputting either the window-displayed character string or word by pen touch.

[0016]

In the handwritten character recognition device according to claim 1, the recognized character string is converted into a kana character string, and a character string is stored in association with a predetermined number of kana characters from the beginning of the converted kana character string. Stored in the means. When a kana character and a specific signal are input by the input means, the character string associated with the kana character that matches the input kana character is extracted from the character string storage means and displayed. Therefore, a character string including Kanji is searched and displayed by inputting only a predetermined number of kana characters.

In the handwritten character recognition device according to the second aspect of the present invention, the recognized character string is converted into a corresponding kana character string by a dictionary and is associated with a predetermined number of kana characters from the beginning of the character string. It is stored in the column storage means. At this time, the length of the kana character string converted by the dictionary is constant regardless of the length of the input character string, and compared with the case of converting all readings of the character string into the corresponding kana character string. The capacity of the dictionary is small. When a kana character and a specific signal are input by the input means, the character string storage means is searched with the input kana character as a key, and the corresponding character string is extracted and displayed.

In the handwritten character recognition device according to the third aspect, when the recognized character string is a katakana character string, the input katakana character string is replaced by the input katakana character string instead of the process of converting it into a kana character string. It is stored in the katakana character string storage means in association with a predetermined number of katakana characters from the beginning. When a katakana character and a specific signal are input by the input means, a katakana character string stored in association with the input katakana character is retrieved from the katakana character string storage means and displayed.

In the handwritten character recognition device according to the fourth aspect, the word boundary of the recognized character string is detected,
It is further broken down into words. The decomposed word is stored in the word storage means in association with a predetermined number of characters from the beginning. When a character and a specific signal are input by the input means, the word corresponding to the character is retrieved from the word storage means and displayed.

In the handwritten character recognition device according to the fifth aspect, the recognized character string is stored in any one of the character string storage means, the word storage means, and the katakana character string storage means in association with a predetermined number of characters. It When a character and a specific signal are input by the input means, the character string storage means,
Either the word storage means or the katakana character string storage means is used as a key to search, and the character string, word, or katakana character string corresponding to the character is displayed.
At this time, if there are a plurality of character strings, words, or katakana character strings, these are displayed in a window, and one of them can be input by selecting with a pen touch.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A handwritten character recognition device according to an embodiment of the present invention will be described below by taking as an example a device for sequentially inputting characters mixed with kana into a character frame on a display-integrated tablet. However, this device is merely an example of the present invention and should not be construed as limiting. For example, it can be applied to a handwritten character input device without a character frame for input. Further, in the following embodiment, the kanji / kana conversion is collectively performed after one set of character strings is input. However, the kanji / kana conversion process is sequentially performed every time the input characters are recognized one by one. Examples are possible.

FIG. 1 is a block diagram of a handwritten character input device which is an example of a handwritten character recognition device according to the present invention.
With reference to FIG. 1, this device recognizes a display-integrated tablet 1 as an input unit and a display unit and handwritten character information given from the display-integrated tablet 1 according to a predetermined method, and also a display-integrated tablet. Based on the information such as the handwritten character recognition unit 2 for determining whether or not there is a specific control button input from 1 and outputting it, and the information such as the recognized character or control button, the entire device is controlled according to a predetermined procedure. A CPU (Central Processing Uni) 3 that operates according to a predetermined program, and a Kanji / Kana conversion dictionary 4 that is connected to the CPU 3 and that converts a word included in an input character string from Kanji to Kana. C
Connected to PU3, using Hiragana or Katakana 1 character as an index, memorizing the word associated with that Hiragana or 1 Katakana, and searching for and outputting the corresponding word when Hiragana or 1 Katakana is given. Kana / word search table 5 and to attach to the kana / word search table 5 to store the corresponding katakana word with one katakana character as an index and to output the katakana word corresponding to a given katakana character The katakana word table 8 and the kanji / word search table 6 which is connected to the CPU 3 and stores the corresponding word by using 1 kanji character as an index and outputs the word corresponding to the given 1 kanji character, and the CPU 3 CP for storing necessary information such as executed programs and text data, display data And a memory 7.

The display-integrated tablet 1 is a panel-shaped coordinate input device. The display-integrated tablet 1 outputs information such as pen-up, pen-down, and pen coordinates to the handwritten character recognition unit. The display-integrated tablet 1 also displays text data, handwriting data, recognition results, and the like under the control of the CPU 3.

The handwritten character recognition unit 2 recognizes Chinese characters and kana characters based on the pen information and coordinate information provided from the display-integrated tablet 1, and determines whether or not the control button of the display-integrated tablet 1 is pressed. Check. The handwritten character recognition unit 2 gives the recognition result and button information to the CPU 3.

The kana / kana conversion dictionary 4 is referred to by the CPU 3 when converting a word from kanji into kana characters. The Kanji / Kana conversion dictionary 4 comprises a Kanji / Kana conversion dictionary table and a Kanji / Kana conversion dictionary index, as shown in FIG. The kanji-kana conversion dictionary table includes a set of words and readings stored in order from address 00000 with reference to FIG. Word numbers 00001 to are given to each word in ascending order. The reading assigned to each word is one character in this embodiment. In FIG. 2B, it is assumed that the word “accept” is stored in the address X and the word number is J, as an example. Furthermore, the word "recognition" referred to later is stored two times after "recognition",
The word number is J + 2.

The Kanji-Kana conversion dictionary index is a table for speeding up access to the Kanji-Kana conversion dictionary table. As shown in FIG. 2A, a Kanji-Kana conversion dictionary index has a start address indicating which Kanji and a word in the Kanji-Kana conversion dictionary table where a word starting with that Kanji is stored, and its start. A set consisting of the word number given to the word stored in the address is created in advance for each Chinese character. By using this kanji-kana conversion dictionary index, when a certain kanji is given, the starting address and word number of the word starting with that kanji can be known. The kanji-kana conversion dictionary table is read from the start address, the target word is searched, and the reading given to the word can be known. In this way, the input character string can be converted at high speed into reading one kana character.

As shown in FIG. 8A, the kana / word search table is composed of the word number of the input word given in the kanji / kana conversion dictionary table and the first character of the reading of the word. It stores multiple sets of indexes. At this time, in the case of Katakana, the index is also the first Katakana character. Further, Katakana is not stored in the Kanji / Kana conversion dictionary 4, so that a word number different from the word number in the Kanji / Kana conversion dictionary table is given. The word number given to this Katakana language is the number on the Katakana language table 8. It can be easily discriminated from the character code that the index is katakana, and therefore it is easily understood that the corresponding character string is stored in the katakana word table 8.

With reference to FIG. 8B, the Katakana language table 8 can store a plurality of character strings composed of Katakana. A number from 0 is given to each character string in ascending order.

Referring to FIG. 8 (c), the kanji / word search table 6 stores one kanji character as an index and the word number of the word associated with that kanji in the kanji / kana conversion dictionary table. Includes multiple sets. An END mark is added to the end of this set of the kanji / word search table.

Character recognition processing of the handwritten character input device according to this embodiment will be described below. What is used for the method for recognizing handwritten characters is not directly related to the present invention, and any method such as a pattern matching method or a basic stroke method may be used.

The handwritten character recognition unit 2 performs a recognition process, a recognition result display, a candidate selection, etc. using the display-integrated tablet 1 independently of the CPU 3. The processing at this time is interactively performed using a display screen as shown in FIG. Referring to FIG. 3, the display screen of the display-integrated tablet 1 displays a body display screen 9 in which a body of an input sentence is displayed and handwritten characters are input one by one, and recognized characters are displayed. Six character frames 10a to 10f and "recognition"
It is used by being divided into the function keys 11 of “shortening”, “cancel”, and “confirm”. The handwritten character recognizing unit 2 is activated by the CPU 3 and then "determined" and "shortened" shown in FIG.
Until the user touches the three types of buttons “end” (not shown) with the pen (referred to as pen touch), the CPU 3 performs the processing without involvement.

The function key 11 will be described below.
The "recognition" key is a key that is pressed when the handwritten character is written in any of the character frames 10a to 10f and then the character is recognized. The "shortening" key is a key used when converting an input character into a character string. By inputting one character in the character frame 10a and pressing the abbreviated key, the word associated with that character is searched and displayed. When the abbreviated key is pressed, the recognition results of up to 6 characters input so far are returned to the CPU 3, and the handwritten character recognition unit 2 is stopped. When the handwritten character recognition unit 2 is stopped, the tablet data and the display data of the display-integrated tablet 1 are exchanged with the CPU 3.

The "cancel" key is used to cancel the input character string and input again. The “confirm” key is a key used to select one of a plurality of result candidates obtained by searching the character string associated with the input character by the abbreviated key. Even when the enter key is pressed, the recognition result is returned to the CPU 3 as in the case of the shortening key, and the handwritten character recognition unit 2 is stopped.

Hereinafter, in the case of inputting a sentence "a recognition method in an online character recognition device. What is online recognition ..." with a mixture of kanji and kana using the handwritten character input device of this embodiment, the operation and the device. The operation will be described.

(1) With reference to FIG. 4, first, the display-integrated tablet 1 is displayed with “O / N / La / I / N” and a character frame 10.
Fill in a to 10e character by character. Handwritten character recognition unit 2
Performs recognition processing for each character frame and displays the recognition result in the corresponding character frame. In FIG. 3, “a” is recognized, and the last “n” is being recognized.

(2) After inputting and recognizing the five characters and confirming the display of the recognition result, the operator touches the "confirm" key with the pen. In response to this, the recognition result is given from the handwritten character recognition unit 2 to the CPU 3. At this time,
CPU3 also has information about the pressing of the "confirm" key.
Given to. The handwritten character recognition unit 2 erases the recognition result and enters a stopped state.

(3) The CPU 3 has a handwritten character recognition unit 2
The recognition result given by is captured and displayed in the text. The display is shown in FIG. At the same time, the CPU 3 creates a kana-word search table and a katakana table related to the present invention based on the recognition result. In this case, it is assumed that no word is stored in the Katakana table before this. The character "online" is stored in the number 0 portion of the Katakana language table, as shown in FIG. 6 (b). In the kana / word search table, as shown in FIG. 6A, the first character "o" of the word "online" and the number of the katakana word table storing the word "online" are stored. "0" is stored as a pair. This is because Katakana is not stored in the Kanji-Kana conversion dictionary 4 in order to improve the efficiency of the dictionary, and information for searching the Kanji-Kana conversion dictionary cannot be obtained. That is, the actual character string is stored in the Katakana word table 8, and the kana / word search table 5 is filled with the index and the number of the Katakana word table 8 containing the character string corresponding to the index. In the present embodiment, the maximum length of the character string that can be stored in the Katakana language table 8 is a fixed length of 6 characters because the maximum number of characters that can be fixedly input at one time is 6.

(4) Next, "sentence / character / recognition / knowledge / apparatus /
"Oki" and enter one character in each box. The handwritten character recognition unit 2 performs the character recognition process for each character frame as in (1), and displays the recognition result on the display-integrated tablet 1. The operator confirms the displayed recognition result of the 6 characters and touches the “OK” key with the pen. In response to this, the recognition result and the “fixed” information are passed to the CPU 3.

(5) The CPU 3 incorporates the recognition result into the text. At the same time, the CPU 3 causes the kanji / kana conversion dictionary 4 (FIG. 1).
Is used to analyze the character string of the recognition result into smaller words. In this case, as a technique for converting kanji / kana, a general technique used in a “document composition system”, a “text voice synthesis system” or the like is used. In this case, the "character recognition device" is divided into three words "character", "recognition", and "device". Each of these words is associated with the corresponding reading and word number by the Kanji / Kana conversion dictionary 4. For example, in order to know the reading and word number of the word “recognition”, first, the kanji-kana conversion dictionary index shown in FIG. 2A is referred to. From the Kanji to Kana conversion dictionary index, a word starting with the letter "Kana" is stored in the area starting from the address X of the Kanji to Kana conversion dictionary table, and the word number given to the first word is J. Recognize. Based on the obtained start address, the Kanji-Kana conversion dictionary table shown in FIG. 2B is referred to. In this case, the word “accept” is stored in the address X. By sequentially checking the kanji-kana conversion dictionary table from this start address, it is found that the word "recognition" is two words after the word "recognition". Therefore, it can be seen that the word number given to the word "recognition" is J + 2. Further, since the reading associated with the word “recognition” in the kanji-kana conversion dictionary table is “ni”, it is understood that “ni” may be used as the reading of this character string. The above results are summarized in FIG. 7.

(6) In FIG. 7, it is understood that the reading given to the character string "character" is "mo" and the word number in the kanji-kana conversion dictionary is L. Similarly, the reading of the word "apparatus" is "so", and the word number is "N".
Is.

(7) Using the analysis results shown in FIG.
The CPU 3 creates a kana / word search table 5 and a kanji / word search table 6. The created result is shown in FIG.
It is shown in (c). That is, "character""recognition"
In the kana / word search table, the character string “apparatus” is represented as a pair of the reading and the word number as shown in FIG. 8A, and as shown in FIG.
The word "sentence" at the beginning of the word search table
It is represented as a pair of the characters "acknowledgement" and "dressing" and the word number of each character string.

(8) Next, "ni / o / ke / ru" is entered in each character frame, one character at a time. The handwritten character recognition unit 2 similarly recognizes and displays input characters one by one. The user confirms the displayed recognition result and then touches “OK” with the pen. The recognition result and the information indicating that the “confirm” key has been pressed are passed to the CPU 3. Since the given recognition result includes only kana characters, the CPU 3 only takes in the text and does not perform the process of creating each table.

(9) Next, in order to input the character string "recognition", "ni" and one character are entered in the character frame 10a.
The handwritten character recognition unit 2 recognizes the input character "ni" and displays the recognition result. After confirming the displayed recognition result, the operator pen-touches the "shortening" key. The handwritten character recognition unit 2 passes the recognition result “ni” and information indicating that the “shortening” key has been pressed to the CPU 3, and the handwritten character recognition unit 2 enters the stopped state. In response to the pressing of the "shortening" key, the CPU 3 searches the kana / word search table 5 (see FIGS. 1 and 8) for all the words having the recognition result "ni" as an index, and takes them out. .. All the extracted words are displayed on the display-integrated tablet 1. At this time, as shown in FIG. 9, the window 12 opened in the text display screen 9 displays all the words of the search result.

Extraction of the corresponding word from the kana / word search table 5 is performed as follows. Referring to FIG. 8A, M is used as the word number corresponding to the index "ni".
Is obtained. The address range in which the word number M exists can be easily known by the kanji-kana conversion dictionary index shown in FIG. In this case the start address X
It is known that there is a word with this word number within a predetermined range thereafter. Therefore, the word "recognition" having the corresponding word number can be obtained by searching the starting address X and subsequent addresses in the Kanji / Kana conversion dictionary table shown in FIG. 2B.

The operator selects a desired one from the displayed word candidates by touching it with a pen and presses the "confirm" key.

(10) Next, similarly input "method / method /." And press the "confirm" key. CPU3 is (5) ~
Each table is created by the same procedure as (7).

(11) Next, consider the case of inputting "online" again. This character string has already been entered once, and the corresponding Katakana table and Kana / Word search table have been created. Therefore, the operator first writes only "o". The handwritten character recognition unit 2
"O" is recognized, and the recognized result is displayed in the corresponding character frame. The operator confirms the displayed recognition result and then "shortens"
Press the key. The recognition result “o” and the information indicating that the “shortening” key has been pressed are passed to the CPU 3. Handwritten character recognition unit 2
Is stopped. The CPU 3 performs the same process as (9). However, in this case, the index is "o".
Since it is Katakana, the character string with the number 0 in the Katakana word table is taken out as shown in FIGS. The extracted character string is displayed on the display-integrated tablet 1.

(12) Next, consider the case of inputting "recognition". In this case, if the operator remembers that the word “recognition” has already been entered once, the operator
By inputting, it is possible to input this word as described in (9). However, if you typed "OK" without being aware that you have already entered this word,
There are also cases in which "Kana" is entered because the user dislikes disturbing thoughts by performing Kanji and Kana conversion in his head. In this case, in the handwritten character input device according to the present invention, after inputting the first character in Kanji, "shortening" can be input to input the already input word. That is, first, the handwritten character recognition unit 2 recognizes the input word "recognition" and displays it in the corresponding character frame. After confirming the displayed recognition result, the operator presses the "shortening" key.
The handwritten character recognition unit 2 passes the recognition result “recognition” and the information that the “shortening” key is pressed to the CPU 3. CPU3
Performs the following processing in response to the pressing of the "shortening" key. In this case, the recognition result is Kanji,
The CPU 3 uses the kanji / word search table 6 (see FIGS. 1 and 8).
(See (c)) to search for and display all corresponding words. In the example shown in FIG. 8 (c), the word number M corresponding to the Chinese character “Kana” is obtained, and the word “recognition” is extracted from the Kanji / Kana conversion dictionary shown in FIG. The extracted words are displayed on the display-integrated tablet 1.

(13) The input is repeated according to the above procedure to input the kanji / kana mixed sentence by handwriting.

As described above, in the handwritten character input device of this embodiment, the kana / word search table, the kanji / word search table, and the katakana word table are automatically created from the input words. As the index in that case, the first character when each word is converted into a kana, the first character including the kanji of each word, and the like are used. So when you want to retype a word that you have already typed,
The same word can be input by inputting the first character of the reading of the word and pressing the "shortening" key, or by inputting the first kanji character and pressing the "shortening" key. Therefore, the input speed of handwritten characters is significantly improved as compared with the conventional case. In this case, so-called kana-kanji conversion is not performed, but it is a more direct input method, and the operability of input is improved. In the kanji-kana conversion dictionary, not all the readings of the word are used as the readings assigned to each word, but the reading is limited to only the first one character or two characters. Therefore, the capacity of the kanji-kana conversion dictionary can be saved. When there are a plurality of words obtained as a result of the search, all of the words are displayed in the window, and the desired word can be selected simply by touching the pen. Therefore, even when there are a plurality of candidates, the character string can be input easily and surely.

[0051]

As described above, according to the invention described in claim 1, the character string once input is
By inputting a predetermined number of kana characters at the beginning and operating the input means to generate a specific signal, the original character string can be converted and displayed / input. Since it is not necessary to input all the character strings mixed with kanji by handwriting, the input efficiency is improved. In addition, the input efficiency is higher than that in the case of performing kana-kanji conversion, and the conversion operation is small, so that the operability is improved.

According to the handwritten character recognition device of the second aspect, the character string once input is displayed with a small number of inputs from the next time by inputting a predetermined number of kana characters from the beginning of the reading. You can enter. Therefore, in a dictionary for storing a character string and a kana character string that is a key in association with each other, the length of the kana character string is limited to a certain value. Therefore, the efficiency of input can be improved, and the capacity of the dictionary for that can be reduced.

According to the handwritten character recognition device of the third aspect, the katakana character string once input can be input as a whole by inputting the leading characters from the next input. .. Therefore, the amount of operation at the time of inputting is reduced, and it is not necessary to include the Katakana word, which has a large variation, in the dictionary, and the dictionary capacity can be suppressed to be small. Moreover, the user can abbreviate the input of katakana by the same operation as when inputting a character string other than katakana, and it is possible to prevent the user from feeling unnatural when using.

According to the handwritten character recognition device of the fourth aspect, a word which has been input in the past can be easily input by inputting only the first character.
Moreover, when inputting a Chinese character, the conversion work of remembering the reading in the head is not necessary, and the efficiency at the time of inputting is improved. Furthermore, if you remember that you entered the word in the past when you entered the first few characters, you can immediately enter the entire string without entering the subsequent characters. Efficiency is improved.

According to the handwritten character recognition device of the fifth aspect, the character string once input can be entirely read by inputting only the leading character of the reading or the leading character of the character string. You can enter the character string of. Therefore, the efficiency in inputting is improved. Further, when there are a plurality of character strings corresponding to the input character, they are displayed collectively in the window, and a desired word can be easily specified by selecting with a pen touch. Therefore, the input can be surely and easily performed.

[Brief description of drawings]

FIG. 1 is a block diagram of a handwritten character input device according to an embodiment of the present invention.

FIG. 2 is a schematic diagram showing a configuration of a kanji / kana conversion dictionary.

FIG. 3 is a schematic diagram showing a display screen of a display-integrated tablet.

FIG. 4 is a schematic diagram showing a display screen of a display-integrated tablet.

FIG. 5 is a schematic diagram showing an input screen of a display-integrated tablet.

FIG. 6 is a schematic diagram showing a kana / word search table and a katakana word table.

FIG. 7 is a diagram showing a result of character string analysis.

FIG. 8 is a schematic diagram showing configurations of a kana / word search table, a katakana table, and a kanji / word search table.

FIG. 9 is a schematic diagram of a display screen of a display-integrated tablet showing a mode of window display.

[Explanation of symbols]

 1 Display-integrated tablet 2 Handwritten character recognition unit 3 CPU 4 Kanji / kana conversion dictionary 5 Kana / word search table 6 Kanji / word search table 7 CPU memory 8 Katakana table 9 Body display screen 10a to 10f Character frame 11 Function key 12 window

Claims (5)

[Claims] 1. An input means for inputting a handwritten character string and a predetermined signal, a recognition means for recognizing the input character string, and a conversion of the recognized character string into a kana character string, which is converted. And a character string storage means for storing the recognized character string in association with a predetermined number of kana characters from the beginning of the kana character string, and responding to a kana character and a specific signal input by the input means. And a means for retrieving the character string storage means to retrieve a character string associated with the kana character that matches the input kana character, and a display means for displaying the retrieved character string. Handwritten character recognition device including. 2. The character string storage means includes a dictionary for storing each character string in association with a specific kana character string and converting the given character string into a corresponding kana character string. The handwritten character recognition device according to claim 1, wherein the maximum value of the length of each kana character string in the dictionary is constant regardless of the length of each character string. 3. A means for determining whether the recognized character string is a katakana character string, and when the recognized character string is determined to be a katakana character string, the kana character string is replaced. And a predetermined number of katakana characters from the beginning of the recognized katakana character string,
The handwritten character recognition device according to claim 1, further comprising: katakana character string storage means for storing the recognized katakana character string in association with each other. 4. An input means for inputting a handwritten character string and a predetermined signal, a recognition means for recognizing the input character string, a word boundary of the recognized character string, and the recognition. A character string decomposing means for decomposing the formed character string into words, and a word memory for storing each of the words output by the character string decomposing means in association with a predetermined number of characters from the beginning of each word. Means for responding to a character and a specific signal input by the input means, searching the word storage means, and extracting a word associated with a character matching the input character, A handwriting character recognition device including a display unit for displaying the retrieved word. 5. The display means includes window display means for displaying the searched character string or word in a window in association with a predetermined input area on the input means, and the input means includes the window display. The handwritten character recognition device according to any one of claims 1 to 4, further comprising means for selecting and inputting, with a pen touch, any of the written character strings or words.