JPH08123903A - Character processing device - Google Patents

Abstract

PURPOSE: To easily retrieve KANJI (Chinese character) to be looked up by retrieving KANJIs having a partial pattern by the common partial pattern and displaying these KANJIs to make a user select them. CONSTITUTION: Writing information of the partial pattern is inputted from an input means 1 by the control or a control means 7 and is stored in a storage means 4. Thereafter, a character code conversion means 6 uses a stored character code conversion dictionary to convert the character code of the sent partial pattern to a proper character code. The converted character code is stored in the storage means 4. A retrieval means 3 uses a partial pattern dictionary stored in the storage means 4 to retrieve KANJIS including the partial pattern by the converted character code, and they are displayed by an output means 5. The user selects desired one of displayed KANJIs by pointing of a pen or the like, and this KANJI is outputted to the output means 5 in accordance with a prescribed format.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device having a function of recognizing and displaying handwritten characters input with a pen or the like and an electronic dictionary function (for example, word processor, electronic notebook, electronic dictionary).

[0002]

2. Description of the Related Art In a kanji search device, there are the following techniques for searching kanji from a partial pattern (a part of kanji, that is, radical). JP-A-4-3558
80 In the “Kanji Search Method”, the reading of the partial pattern (called “part” in the “Kanji Search Method”) is input from the keyboard, the Kanji group containing the partial pattern is searched, and the Kanji group is searched from the Kanji group. A device for selecting the kanji to be searched and searching the semantic information of the kanji is described.

However, since there are many Kanji whose readings are the same and have different notations, generally the number of Kanji searched for by reading the partial patterns is very large, and the capacity of the partial pattern dictionary becomes large. For example, if you want to look up "Ka" and put the reading "ko" of the partial pattern "Kou", kanji including the different lines "row", "ko", "ko" etc. will also be in the partial pattern dictionary. Therefore, there are only two Kanji characters that contain the partial pattern "Kou", "Kakan", but 1481
Individual kanji are registered for "Kou". Not only does this increase the capacity of the pattern dictionary, but it also requires a great deal of effort to search the Kanji character intended by the user from this search result because it must be searched in combination with other partial patterns.

Further, in Japanese Patent Laid-Open No. 58-158777 "Online Handwritten Character Recognition Device", the first kana reading kana is used.
Handwriting recognition of the sound or radical of Kanji on the tablet,
A device for searching a kanji group including the first phonetic kana of the reading kana or the radical of the kanji, displaying them, and inputting the correct kanji font again by handwriting is described.

Also in this method, the above-mentioned Japanese Patent Laid-Open No. 4-3 is used.
55880 “Kanji Search Method”, the capacity of the dictionary for storing the information of the kanji group including the kana of the reading kana or the radical of the kanji becomes large, and the kanji group searched by the user becomes large. It takes a lot of effort to find the kanji you want to look up.

In handwriting recognition of a partial pattern, some partial patterns have almost the same pattern as katakana and the like, and therefore, some of them may be erroneously recognized (for example, "Ninben" and "A" in katakana). Occurs. Since there are such characters that are similar to the characters that should be recognized, the user does not notice the misrecognition, continues the search process, and does not find the desired kanji, and continues handwriting input until the correct character is recognized. The work is repeated and the user selects a correct character from the recognition candidates.

[0007]

The above problem, that is, in the kanji search, the number of kanji including these for the partial pattern reading, the first phonetic writing of the kanji reading, or the kanji radical notation Therefore, in order to solve the problem that the capacity of the dictionary (partial pattern dictionary) for storing such information becomes large, the present invention does not register all the kanji belonging to the partial pattern in the partial pattern dictionary. , Replace a Kanji group that can be replaced with a partial pattern with a partial pattern. This makes it possible to reduce the dictionary capacity of partial patterns.

Further, when the apparatus erroneously recognizes a character input by the user, if the erroneously recognized character is similar to the correctly recognized character, the handwriting input operation is repeated until the correct character is recognized. There is also a problem that the user has to select the correct character from the recognition candidates. Therefore, in the present invention, by using a character code conversion dictionary that converts a character code that is erroneously recognized as a partial pattern of Kanji into a correct character code, even if the device is erroneously recognized, a partial pattern search can be performed smoothly. Will be easier.

It is an object of the present invention to reduce the dictionary capacity of partial patterns and to search for Kanji characters that include partial patterns even if the characters are similar to the misrecognized partial patterns.

[0010]

[Means for Solving the Problems] Input means for inputting handwritten information with a pen, a tablet, etc., for recognizing the input information as characters, a storage means for storing intermediate results of processing, etc., and an output means for outputting characters. In a character processing device including: a recognition unit for recognizing a partial pattern input by the input unit; and a character code conversion unit for converting the recognized partial pattern into a character code of an appropriate partial pattern. It is equipped with a pattern dictionary in which a kanji group that can be replaced with is replaced by a partial pattern and stored, and a search means for searching for kanji having that partial pattern from the partial pattern. A desired kanji can be input by displaying and selecting the kanji including the pattern.

[0011]

First, the writing information of the partial pattern is input from the input means under the control of the control means and stored in the storage means. Then, the recognition unit recognizes the character code of the corresponding character pattern from the handwriting information input using the basic stroke dictionary and the character dictionary stored in the handwritten character recognition means. The character code is stored in the storage means.

Next, the character code conversion means converts the character code of the sent partial pattern into an appropriate character code by the character code conversion dictionary stored in the storage means. The converted character code is stored in the storage means. Next, the search means searches the converted character code for a kanji character including the partial pattern using the partial pattern dictionary stored in the storage means, and the output means displays the searched kanji character. Then, the user selects one desired kanji from the displayed kanji with a point such as a pen, and the kanji is output to the output means in a predetermined format.

According to the present invention, in a partial pattern dictionary composed of partial patterns and Chinese characters including the partial patterns, instead of registering all the Chinese characters including the partial patterns for each partial pattern, a group of Kanji characters that can be replaced with the partial patterns. The above problem is solved by substituting for a partial pattern. In addition, even if the user fails to recognize the partial pattern of the Kanji that he / she wants to search, the user can convert the partial pattern into the correct partial pattern and perform the search, thereby solving the problem caused by erroneous recognition.

[0014]

The present invention will be described in detail below with reference to the drawings.
FIG. 1 is a block diagram of a kanji search device according to the present invention.

In FIG. 1, reference numeral 1 is an input means for inputting writing information of a partial pattern of Chinese characters and for inputting information for selecting a Chinese character from searched Chinese characters. Reference numeral 2 is a recognition means for recognizing a character from the input handwriting information and generating a character code. Reference numeral 3 is a search means for searching for kanji having a partial pattern and for searching the meaning information of kanji. Reference numeral 4 is a storage means such as a floppy disk / ROM / RAM / hard disk for temporarily storing the partial pattern dictionary and the intermediate results of the processing. Reference numeral 5 is an output means such as a printer. 6
Is a character code conversion means for converting the character code of the transmitted partial pattern into an appropriate character code. If the sent character code is correct (if the partial pattern dictionary has a partial pattern character code), this character code conversion dictionary does nothing. 7 is C for controlling each of these means
It is a control means such as a PU.

FIG. 2 is a plan view of the input / display means of the present invention. Reference numeral 8 is an entry frame for performing character recognition, and one character is handwritten in one frame. You can write from any frame, and regardless of the position of the input frame, it will be recognized in the order written and displayed on the input line. Reference numeral 9 denotes a candidate character display window in which the recognized character candidates are displayed. 10 is an input line, and the cursor is underlined. Then, in this window, the character input / recognized in the entry box 8 enters the position of the cursor. Reference numeral 11 is a search character display window, in which the contents of 10 input lines are stored by pointing the character string used for search to the 16 "input" icon with a pen or the like. Reference numeral 12 is an icon for moving the cursor on the input line. Reference numeral 13 is an icon for deleting the character immediately before the cursor position on the input line. Reference numeral 14 is an icon for deleting the character at the cursor. Reference numeral 15 is an icon for ending the entire search processing. Reference numeral 16 is an icon for setting the character string of the input line in the 11 search character display window. Reference numeral 17 is an icon for performing a partial pattern search of Chinese characters. 1
Reference numeral 8 is an icon for performing kana-kanji conversion.

An example in which the user searches for a Chinese character from the partial pattern "2" is shown. In order to input "2" which is the partial pattern of this Kanji, the user inputs a stroke in the entry box of 8 with the pen in the order of "→", "→"("→" is stroke information) (Fig. 3 ), The recognition means recognizes the stroke information input with the movement of the pen by collating with the basic stroke dictionary having the information of FIG. 4, and further, a character dictionary in which characters are described in advance in the basic stroke in the stroke order. And the characters that satisfy the contents of the dictionary are recognized. An example of the character dictionary is shown in Table 1 below.

Table 1 Example of character dictionary. ． ． 01 01: d. ． ． 01 04 07 01 04 04 01: Sub. ． ． 01 01: Two. ． ． In this example, in order to make the explanation easier to understand, characters are used instead of codes, and the delimiters are shown as ":", but in the dictionary, they are actually expressed as code information such as JIS code, The delimiter is also represented by another code or without a delimiter. The same applies to the other dictionaries described below.

The first example of this character dictionary is the stroke information of "0101", which is "ni" in katakana, and the second example is the stroke information of "01040707040401", which indicates "sub". The third indicates that the stroke information of "0101" is the Chinese character "two." Since the stroke information cannot identify the katakana "ni" and the kanji "two", it is erroneously recognized as katakana "ni" (FIG. 5) from the order of the character codes. Even if the user erroneously recognizes, it is not necessary to return the cursor position to the position of the input katakana "ni" and select the kanji "two" from the correct candidates displayed in the candidate character display window of 9. You can move to the next process.

Next, by pointing the 16 icon with the pen, the katakana "ni" is set in the search character display window as shown in FIG.
When you point with the pen, the search for the partial pattern starts, and the character code of katakana "ni" is sent to the character code conversion means. The character code conversion means uses a character code conversion dictionary as shown in Table 2 below to collate the sent character code with the character code conversion dictionary and replace them with appropriate character codes.

Table 2 Example of character code conversion dictionary. ． ． Ni: Nino: Rice bowl. ． ． The example on the first line shows that the katakana "ni" is replaced with the kanji "ni", and the second example is the katakana "no".
Is replaced with the partial pattern "Bill".
The character code conversion dictionary contains information for converting a character code that is erroneously recognized as a partial pattern of Kanji into a correct character code. This process can relieve the user from the trouble of re-inserting the partial pattern again due to erroneous recognition and selecting another recognition candidate in the candidate character display window.

Next, the searching means searches the partial pattern dictionary (Table 3) for a Chinese character containing "two." The partial pattern dictionary is made up of partial patterns and information on Chinese characters belonging to the partial patterns. In the partial pattern dictionary, the following 55 Chinese characters are registered for the partial pattern "2".

Table 3 Example of partial pattern dictionary. ． ． Two: Aiun Yumaki Gyakukyo Gokaku Gansai Hakken Kengen Gogo Gotoku Jinjing Jinyi Shunjou Nana 2nd half-table cheek spirit crossing 卍 庞 亟 冑 冕 劵 卷 咢 夸 佞 晉 滕 灁 眷 縢虧 豢 鞞 韵 齊． ． ． In the present invention, in the partial pattern dictionary, instead of registering all the Chinese characters including the partial pattern for each partial pattern, the Chinese character group that can be replaced with the partial pattern is replaced with the partial pattern. This is because it is wasteful in capacity to simply register all the kanji belonging to the partial pattern in the partial pattern dictionary.

For this reason, in the search means, when a partial pattern is included in the searched Chinese character, the repeated partial pattern dictionary is searched. In other words, the kanji containing the partial pattern "2" can be found in the partial pattern dictionary (above) from the Ai Yun Shimaki Gyōkyo Gokaku Ganza Hikken Kengen 5亞 亟 冑 冕 劵 卷 咢 夸 佞 晉 滕 瑁 眷 縢 冐 乕 虧 豢 鈞 韵 齊, but "A" is a partial pattern, so "A"
Is replaced by "Atsubo". Similarly, the partial patterns such as "i" and "un" are similarly repeated, and the partial pattern dictionary is searched. That is, among the kanji searched for by "2", the kanji that can be a partial pattern displays the kanji searched by the kanji. Therefore, for example, there are 215 Chinese characters including “two” as described below.

蔖 卖國 蔋 虭 in Yinyu Deungunun Dirty Meeting Egaki Jinyakugyokan Complete 桓 莞 莫 碓 厓 要 纓 胠 轠 侠Honorable Gogogogogogogo Hengko soul Triage Sai agent Katsushin Jinin Qi Spin Guoju tsubo tsutsuden Biography Tōhō Tōmei Kasumi Nagi于亞 亟儕 helmet 冕冦 冱刳 劑劵 gourd 卷吁 咢唔 囈圄 圈堊 Bintsubo 壼夸 佞娜 寇寤 恆惡 惓愕 愃拌 舉擠 晉晤 暄桍 Nagi 椏椦 沍浣 滕濟 炬牾 珸瑁 Avery 畊瘠皖盂 眷齋 秬穽 rattan 籘紆 紜絆 navel 縉縢 繧纃 緕罎 耘冐 Yutaka 胯膩 umbilical 芫苣 calyx 蕚蒄 Nazunagei 乕虧 Nina 衙袢 諤諠 豢貳 齎蹐 躋鄂 chun 鉅錏 鐚阮 霽靆 靉齏韲 韵 鶚 鶺 齊 齬 齶 As you can see from this example, if it is a partial pattern for "A" or "I" which is the search result of "2", it is searched for as a search target as in the past. To do. Therefore, the capacity of the partial pattern dictionary can be significantly reduced. The search result will be displayed as shown in FIG. 19 is a window (8 to 1 described in FIG. 2) displaying a part of the searched Kanji (because 215 cannot be displayed at one time).
It is displayed on the window icon up to 8) and captures the information of Kanji pointed by a pen or the like. Reference numeral 20 is an icon for returning to the screen of FIG. Reference numeral 21 is an icon for viewing other retrieved Kanji display screens and the previous display screen.

In FIG. 7, when the Chinese character "Yin" is pointed by a pen or the like, the character code of the Chinese character "Yin" is sent to the output means, and the Chinese character is displayed according to a predetermined format.

FIG. 8 shows an example of searching for "A", which is a screen in which a stroke is entered in the entry box 8 with a pen. FIG. 9 shows the result of recognition using the basic stroke dictionary of FIG. 4 and the character dictionary of Table 1. Then, by pointing the 16 icon in FIG. 2 with the pen, "A" is set in the search character display window as shown in FIG.
The partial pattern is searched by pointing to the icon of 7. Since "A" is not registered in the character code conversion dictionary, no processing is performed by the character code conversion means. The search means searches the partial pattern dictionary as it is, and the search result is displayed as shown in FIG.

Then, when a desired Chinese character is selected from the displayed Chinese characters with a pen or the like, the output is performed to the output means.

[0029]

According to the present invention, even if the recognition of the input handwritten character is not successful, the handwriting input operation is repeated until the correct character is recognized, or the correct character is selected from the recognition candidates by the user. You can easily search for the kanji you want to search without making any selections, and you can reduce the capacity of the partial pattern dictionary.

Further, since the number of Chinese characters including a certain partial pattern is generally small, it is not necessary to search in combination with other partial patterns, and the Chinese character to be searched can be quickly searched.

Further, even if one does not understand the reading of a kanji or a part of a kanji, the kanji desired to be searched can be searched for only by inputting a partial pattern of the kanji.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an apparatus according to an embodiment of the present invention.

FIG. 2 is a plan view of an input / display unit according to an embodiment of the present invention.

FIG. 3 is a diagram showing an example of inputting a Chinese character “2” by handwriting in an embodiment of the present invention.

FIG. 4 is a diagram showing an example of a basic stroke in handwriting input.

[Fig. 5] Kanji "ni" by handwriting input is katakana "ni"
It is the figure which showed the example erroneously recognized.

FIG. 6 is a diagram showing an example in which katakana “D” is set in the search character display window according to the embodiment of the present invention.

FIG. 7 is a partial pattern “d” in one embodiment of the present invention.
It is the figure which showed the example of the search result of.

FIG. 8 is a diagram showing an input example of a Chinese character “A” by handwriting input.

FIG. 9 is a diagram showing an example in which a Chinese character “A” is recognized by handwriting input.

FIG. 10 is a diagram showing an example in which “A” is set in the search character display window in the embodiment of the present invention.

FIG. 11 is a diagram showing an example of a search result of a partial pattern “A”.

[Explanation of symbols]

1 input / display means 2 recognition means 3 search means 4 storage means 5 output means 6 character code conversion means 7 control means 8 entry frame for character recognition 9 candidate character display window 10 input line window 11 search character display window 12 input Icon to move the cursor on the line 13 Icon to delete the character just before the cursor position on the input line 14 Icon to delete the character at the cursor 15 Icon to end the whole search process 16 Search the character string on the input line Character display window Set icon 17 Icon for Kanji pattern search 18 Icon for Kana-Kanji conversion 19 Window for displaying a part of searched Kanji 20 Icon for returning to the screen of FIG. 4 21 Kanji display screen Icon to switch

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G06K 9/68 A 9061-5H

Claims (2)

[Claims] 1. A character having an input means for inputting handwritten information with a pen, a tablet, etc., for recognizing the input information, a storage means for storing an intermediate result of processing, etc., and an output means for outputting a character. In the processing device, a recognition unit that recognizes the partial pattern input by the input unit, a pattern dictionary that stores a Chinese character group that can be replaced by the partial pattern by replacing it with the partial pattern, and a partial pattern from the partial pattern A character processing device comprising: a search means for searching for kanji, wherein when a user inputs a partial pattern, a kanji including the partial pattern is displayed and selected so that a desired kanji can be input. 2. The character processing device according to claim 1, further comprising character code conversion means for converting the input and recognized partial pattern into a character code of an appropriate partial pattern, wherein the partial pattern erroneously recognized by the recognition means is appropriate. A character processing device characterized by converting to a different character code.