JP2013206141A - Character input device, character input method, and character input program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a character input device that can shorten the time required for input of an idiom.SOLUTION: The character input device is a character input device 1 for input of a character, includes a touch panel handwriting detection unit 11 for detecting handwriting, a handwritten character recognition control unit 12 for recognizing the detected handwriting as a character, and a prediction conversion control unit 13 for predicting an idiom including some of a recognized first character and a second character following the first character.

Description

  The present invention relates to a character input device for inputting characters, a character input method, and a character input program.

  Conventionally, there are character input devices having various character prediction functions and idiom prediction functions. For example, in Patent Document 1, when the first character is recognized, in order to predict a character that may be input next, a word, idiom, and phrase starting from the first character are extracted. The device to do is shown. The apparatus further refers to the frequency information, arranges the extracted words and phrases in the order of the possibility of appearing, and predicts the character following the first character.

  In addition, when the first character is input by handwriting, the device described in Patent Document 2 displays a candidate list of two-character kanji idioms including the first kanji having the radical.

  In addition, the device described in Patent Document 3 has means for designating all or part of an input kanji in handwriting input of a single kanji, and when a part (radical) is designated, the radical is designated. The correct character recognition result is made to be selected by the user by presenting the kanji candidate list possessed.

JP 07-152750 A JP 2000-200269 A Japanese Patent No. 3784467

  However, the conventional character input device that performs handwritten character recognition is not an effective means for improving the predicted hit rate when inputting two or more kanji idioms by handwriting at an early stage of input. For example, even if “kai” is entered as the first character and “goen” is entered as the radical as the second character, the second character is recognized as the kanji of “word” ”,“ Conversation ”,“ meeting ”,“ accounting ”, and“ journal ”did not appear. Therefore, in a conventional character input device that performs handwritten character recognition, when inputting kanji words, if the number of strokes is large, the input operation is troublesome. In particular, during the description of the second and subsequent characters of the idiom, it is desirable that the idiom candidate is recognized early even during the description because the time required for input is shortened.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a character input device, a character input method, and a character input program that can reduce the time required to input a idiom.

  The character input device of the present invention is a character input device for inputting characters, a handwriting detection unit for detecting a handwriting, a character recognition unit for recognizing a handwriting detected by the handwriting detection unit, and the character recognition A idiom prediction unit that predicts a idiom that includes a first character recognized by the unit and a part of a second character that follows the first character.

  According to this configuration, since the idiom including the first character and a part of the second character is predicted without recognizing the entire second character, the idiom narrowing time is shortened. Therefore, the time required for inputting the idiom can be shortened.

  The character input method of the present invention is a character input method for inputting characters, the step of detecting a handwriting, the step of recognizing the detected handwriting as a character, and the recognized first Predicting an idiom that includes a character and a portion of a second character that follows the first character.

  According to this method, the idiom including the first character and a part of the second character is predicted without recognizing the entire second character, so that the time for narrowing down the idiom is shortened. Therefore, the time required for inputting the idiom can be shortened.

  Moreover, the character input program of this invention is a program for making a computer perform each step of the said character input method.

  According to this program, the idiom including the first character and a part of the second character is predicted without recognizing the entire second character, so that the time for narrowing down the idiom is shortened. Therefore, the time required for inputting the idiom can be shortened.

  According to the present invention, the time required for inputting a idiom can be shortened.

The block diagram which shows the structural example of the character input device in the 1st Embodiment of this invention. The perspective view which shows the external appearance example of the character input device in the 1st Embodiment of this invention The flowchart which shows an example of the character input processing procedure in embodiment of this invention The flowchart which shows in detail an example of a part (steps S3-S5) of the character input processing procedure in the 1st Embodiment of this invention. (A), (B) The figure which shows an example of the input state of the handwritten character in the 1st Embodiment of this invention The block diagram which shows the structural example of the character input device in the 2nd Embodiment of this invention. The flowchart which shows in detail an example of a part (steps S3-S5A) of the character input processing procedure in the 2nd Embodiment of this invention. The block diagram which shows the structural example of the character input device in the 3rd Embodiment of this invention. The flowchart which shows in detail an example of a part (steps S3-S5B) of the character input processing procedure in the 3rd Embodiment of this invention. The block diagram which shows the structural example of the character input device in the 4th Embodiment of this invention. The flowchart which shows in detail an example of a part (steps S3A-S5C) of the character input processing procedure in the 4th Embodiment of this invention. The block diagram which shows the structural example of the character input device in the 5th Embodiment of this invention. The flowchart which shows in detail an example of some character input processing procedures (step S3A-S5D) in the 5th Embodiment of this invention. The block diagram which shows the structural example of the character input device in the 6th Embodiment of this invention. The flowchart which shows an example of a part of character input processing procedure (step S3A-S5E) in detail in the 6th Embodiment of this invention. The block diagram which shows the structural example of the character input device in the 7th Embodiment of this invention. The flowchart which shows in detail an example of a part (steps S3A-S5F) of the character input processing procedure in the 7th Embodiment of this invention. The figure which shows the example of a display of another recognition candidate character in the 1st Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram illustrating a configuration example of a character input device 1 according to the first embodiment. FIG. 2 is a perspective view showing an example of the appearance of the character input device 1. The character input device 1 includes a touch panel handwriting detection unit 11, a handwritten character recognition control unit 12, a prediction conversion control unit 13, a display control unit 14, a display unit 15, a character code database (DB) 16, and a vocabulary prediction dictionary database (DB). 17.

  The touch panel handwriting detection unit 11 includes a touch panel 11a arranged on the screen of the display unit 15 (see FIG. 2). The touch panel handwriting detection unit 11 detects handwriting input by the user touching or approaching a screen with a finger or stylus. The touch panel handwriting detection unit 11 outputs the handwriting data (coordinate information) to the handwritten character recognition control unit 12 and the display control unit 14.

  The handwritten character recognition control unit 12 sends handwriting coordinate information to the character code DB 16 to obtain a handwritten recognition character code and a radical code to be described later. The handwritten character recognition control unit 12 sends the obtained handwritten recognition character code to the display control unit 14, and sends the obtained handwritten recognition character code and radical code to the prediction conversion control unit 13. The handwritten character recognition control unit 12 also has a function as a character determination unit that determines a recognized character.

  A character code corresponding to handwriting data (coordinate information) is registered in the character code DB 16. The character code DB 16 has a radical code database (DB) 16a in which radical codes are registered. The radical code corresponding to the handwriting data is held in the radical code DB 16a. The character code DB 16 outputs a handwritten recognition character code and a radical code corresponding to the handwriting data.

  The prediction conversion control unit 13 sends the handwritten recognition character code and the radical code sent from the handwritten character recognition control unit 12 to the vocabulary prediction dictionary DB 17 to obtain a idiom of the prediction candidate.

  In the vocabulary prediction dictionary DB 17, idioms corresponding to character codes are registered. The vocabulary prediction dictionary DB 17 includes a radical vocabulary search database (DB) 17a corresponding to the character code and radical code. The lexical prediction dictionary DB 17 outputs the idiom candidate character codes narrowed down by the radical code among the prediction candidate idioms corresponding to the character codes.

  The predictive conversion control unit 13 outputs the idiom candidate character code corresponding to the handwritten recognition character code, and outputs the idiom candidate character code narrowed down by the radical code among the idiom candidates. Not limited to this, the predictive conversion control unit 13 may narrow down the idiom candidates by the radical code only when the idiom candidates corresponding to the character codes output from the vocabulary prediction dictionary DB 17 exceed a predetermined number. Good.

  The display control unit 14 is based on the handwriting data from the touch panel handwriting detection unit 11, the handwritten recognition character code from the handwritten character recognition control unit 12, and the narrowed idiom candidate character code from the prediction conversion control unit 13. Display control of the display unit 15 is performed.

  The display unit 15 is a liquid crystal display arranged in the casing of the character input device 1. In FIG. 2, a recognized character display area 15 a and a prediction candidate display area 15 b of the character input application are arranged at the upper part of the screen of the display unit 15. In FIG. 2, a handwriting area 15 c and an operation key screen area 15 d are arranged below the screen of the display unit 15.

  In the operation key screen area 15d, a confirmation key 25a, arrow keys 25b and 25c, a full-width / half-width key 25d, a character candidate key 25e, and a delete key 25f are arranged.

  The handwriting area 15c is an area in which the handwriting of two squares (in the example of FIG. 2, the left side is the first square and the right side is the second square) can be displayed. The touch panel 11a described above is arranged in front of the handwriting area 15c. The portion of the touch panel 11a corresponding to the left first cell corresponds to the first section, and the portion of the touch panel 11a corresponding to the second right cell corresponds to the second section.

  In addition, the function of the handwritten character recognition control part 12 and the prediction conversion control part 13 is implement | achieved when the CPU in the character input device 1 runs the character input program mentioned later stored in ROM.

  The handwritten character recognition control unit 12 and the character code DB 16 constitute a character recognition engine unit 30 (an example of a character recognition unit). The prediction conversion control unit 13 and the vocabulary prediction dictionary DB 17 constitute a prediction conversion engine unit 40 (an example of a idiom prediction unit).

Next, an operation example of the character input device 1 will be described.
FIG. 3 is a flowchart showing a character input processing procedure. This character input program is stored in the ROM in the character input device 1 and is executed by the CPU in the character input device 1.

  When the user starts character input, the touch panel handwriting detector 11 detects a handwritten character (handwriting) input to the handwriting region 15c (step S1). Here, at the start of character input, the handwritten character recognition control unit 12 initializes a variable C representing the number of characters in an unconfirmed state to a value of 1.

  Subsequently, the touch panel handwriting detection unit 11 sends the coordinate information (handwriting data, ink data) of the point sequence of the handwritten character detected by the touch panel handwriting detection unit 11 to the handwritten character recognition control unit 12 (step S2).

  Subsequently, the handwritten character recognition control unit 12 extracts a character code recognized using the character code DB 16 (step S3).

  Subsequently, the display control unit 14 displays the unrecognized recognized characters in the recognized character display area 15a of the display unit 15 (step S4). For example, an underbar is added to the unrecognized recognition character (see FIG. 2).

  Subsequently, the prediction conversion control unit 13 extracts idiom candidates predicted using the vocabulary prediction dictionary DB 17 from the characters in an undefined state. The display control unit 14 displays the idiom candidates in the prediction candidate display area 15b (step S5). Further, as described above, the predictive conversion control unit 13 narrows down the prediction candidates from the radical after the second character in the unconfirmed state (C ≧ 2).

  Subsequently, the handwritten character recognition control unit 12 determines whether or not the recognized character in the undetermined state displayed in the recognized character display area 15a in step S4 is correct (step S6). In the present embodiment, it is determined that the recognized character of the first cell is correct when the input to the first cell in the handwritten region 15c is completed (may be halfway) and the input to the second cell is detected.

  If it is determined that the unrecognized recognized character is correct, the handwritten character recognition control unit 12 increases the variable C by 1 (step S7).

  Subsequently, the predictive conversion control unit 13 determines whether or not the recognized character is to be confirmed (step S8). Here, the recognition character is confirmed by pressing the confirmation key 25a. The handwritten character recognition control unit 12 detects pressing of the confirmation key 25a.

  When it is determined that the recognized character is to be confirmed, the character input device 1 completes the character input (step S9) and ends the operation.

  On the other hand, if it is not determined that the recognized character is to be confirmed, the character input device 1 proceeds to the process of step S14 described later.

  When the user determines that the recognized character is incorrect in step S6, the user performs a predetermined touch. Therefore, the handwritten character recognition control unit 12 detects that the user touches the character candidate key 25e or directly touches the underlined “kai” character that is the current recognized character. When a touch is detected, the display control unit 14 displays a list including other recognized character candidates in the recognized character candidate display area 15e on the screen as shown in FIG.

  The handwritten character recognition control unit 12 determines whether or not there is a correct character among the recognized character candidates displayed on the screen (step S10). Whether there is a correct character is determined based on, for example, whether a direct touch by a user on a character button in the list is detected. When there is a correct character, the handwritten character recognition control unit 12 selects a desired recognition candidate character based on a touch on the character button. The touch as described above is detected by the touch panel 11a.

  If there is a correct character among the recognition character candidates, the character input device 1 returns to the process of step S4.

  On the other hand, when there is no correct character among the recognized character candidates, the handwritten character recognition control unit 12 determines whether or not all strokes have been written (input) (step S11).

  When the total number of strokes has been written, the handwritten character recognition control unit 12 deletes the character input to the handwriting region 15c in order to rewrite the handwritten character (step S12). And the character input device 1 returns to the process of step S1.

  On the other hand, when the total number of strokes is not written, the handwritten character recognition control unit 12 determines whether or not the variable C is 2 or more (step S13).

  When the variable C is 1 or less, the character input device 1 returns to the process of step S1.

  On the other hand, when the variable C is 2 or more, the prediction conversion control unit 13 is displayed in the prediction candidate display area 15b by the subsequent second or subsequent radical prediction even if a correct recognized character is not obtained. It is determined whether or not the prediction candidate has a desired idiom (step S14). Whether there is a desired idiom is determined by, for example, directly touching a idiom in the prediction candidate.

  When there is no desired idiom, the character input device 1 returns to the process of step S1.

  On the other hand, when there is a desired idiom, the touch panel handwriting detection unit 11 detects selection by touching the desired idiom. In response to this detection, the predictive conversion control unit 13 sets the character in a finalized state with a desired idiom (step S15). Thereafter, the character input device 1 completes character input in step S9 and ends this operation.

  As described above, the character input device 1 narrows down the idiom prediction candidates from a part of characters after the second character in the unconfirmed state (C ≧ 2). Thereby, even in a state where correct handwritten recognition characters are not obtained, there is a possibility that a desired idiom can be quickly selected by prediction based on a radical or the like after the second character.

  FIG. 4 is a flowchart showing in detail the processing procedure of steps S3 to S5 in the character input processing procedure.

  In step S3, when the handwritten character recognition control unit 12 recognizes the character using the character code DB 16 and extracts the character code, if there is a radical after the second character, the handwritten character recognition control unit 12 uses the radical code DB 16a to Recognize and extract radical codes after the first character.

  Subsequently, in step S4, as described above, the display control unit 14 outputs the unrecognized recognized character to the recognized character display area 15a.

  5A and 5B are diagrams showing the input state of handwritten characters. As shown in FIG. 5 (A), the handwritten character recognition control unit 12 makes a “meeting” to be displayed in the recognized character display area 15a based on the handwritten character entered in the first square of the handwritten area 15c. Recognize the characters. Further, as shown in FIG. 5B, the handwritten character recognition control unit 12 is a radical that is displayed on the recognized character display area 15 a based on the handwritten character entered in the second square. Recognize words.

  In step S5, the handwritten character recognition control unit 12 determines whether or not the variable C is 2 or more (step S51). When the variable C is 1 or less, the handwritten character recognition control unit 12 sends the extracted character code to the prediction conversion control unit 13. The predictive conversion control unit 13 acquires idiom candidates from the vocabulary prediction dictionary DB 17 using the character code from the handwritten character recognition control unit 12 (step S52).

  Subsequently, the display control unit 14 displays the acquired idiom candidates in the prediction candidate display area 15b (step S53).

  On the other hand, when the variable C is 2 or more in step S51, the handwritten character recognition control unit 12 sends the extracted character code and radical code to the predictive conversion control unit 13 (step S54).

  Subsequently, the predictive conversion control unit 13 determines the character code of the 1st to (C-1) th characters (also simply referred to as (C-1)) of the indeterminate and completed (recognized) characters and the radical of the C character. From the code, idiom candidates narrowed down using the radical vocabulary search DB 17a are acquired (step S55). Then, the character input device 1 proceeds to the process of step S53.

  Thus, in the character input device 1, the predictive conversion control unit 13 uses the first character recognized by the handwritten character recognition control unit 12 and the idiom including a part of the second character subsequent to the first character. Predict. The first character is, for example, the first to (C-1) th characters. The second character is the C-th character, for example. A part of the character is, for example, a radical.

  According to the character input device 1, since the radical code is recognized from the recognized handwritten radical and the idiom is extracted from the character code and the radical code, the idiom candidates can be narrowed down with high accuracy. Therefore, the idiom can be input promptly. Further, even when the handwritten character of the second cell is being input, the idiom candidate is displayed when the radical is recognized, thereby further reducing the time required for inputting the idiom.

(Second Embodiment)
In the second embodiment, unlike the first embodiment, the vocabulary prediction dictionary DB is not provided with a radical vocabulary search DB. Instead, the lexical prediction dictionary DB registers character codes of idiom candidates corresponding to handwritten recognition character codes and radical information of kanji included in the idioms.

  FIG. 6 is a block diagram illustrating a configuration example of the character input device 1A according to the second embodiment. As shown in FIG. 6, the character input device 1A of the second embodiment has almost the same configuration as the character input device 1 of the first embodiment shown in FIG. About the same component as 1st Embodiment, the description is abbreviate | omitted by using the same code | symbol, A different structure and operation | movement are mainly demonstrated.

  In the vocabulary prediction dictionary DB 17A, the idiom candidate character code corresponding to the handwritten recognition character code and the radical information (for example, radical code) of the kanji included in the word (idiom candidate) are registered.

  Further, a radical filter unit 18 is added to the character input device 1A. The radical filter unit 18 acquires the idiom candidate character code and the kanji radical information included in the idiom candidate from the prediction conversion control unit 13. The radical filter unit 18 acquires a radical code from the handwritten character recognition control unit 12.

  The radical filter unit 18 searches the radical information of the K-th character of the idiom candidate, extracts the idiom candidates that match the radical code from the prediction conversion control unit 13, and narrows down the idiom The candidate character code is sent to the display control unit 14.

  The prediction conversion control unit 13 of the present embodiment is an example of a first idiom prediction unit, and the radical filter unit 18 is an example of a second idiom prediction unit.

Next, an operation example of the character input device 1A will be described.
About the same step process as FIG. 3, FIG. 4 shown in 1st Embodiment, the description is abbreviate | omitted by attaching | subjecting the same step number. Here, only processing different from the first embodiment will be described. In the second embodiment, the process of step S5A is performed instead of the process of step S5.

  FIG. 7 is a flowchart showing in detail the processing procedure of steps S3 to S5A in the character input processing procedure.

  In step S3, when the handwritten character recognition control unit 12 recognizes the character using the character code DB 16 and extracts the character code, if there is a radical after the second character, the handwritten character recognition control unit 12 uses the radical code DB 16a to Recognize and extract radical codes after the first character.

  Subsequently, in step S4, the display control unit 14 causes the recognized character display area 15a to output an unconfirmed recognized character.

  Subsequently, in step S5A, when the variable C is equal to or greater than 2 in step S51, the predictive conversion control unit 13 includes the idiom candidate and the idiom candidate from the character code of the unconfirmed and completed (recognized) character. The radical information of the Chinese character to be acquired is acquired (step S54A). When the variable C is 2 or more, the prediction conversion control unit 13 predicts using the first to (C-1) th characters. For example, as shown in FIG. 5A, a idiom is predicted using a “kai” as a completed character.

  Subsequently, the radical filter unit 18 searches the radical information of the K character of the C character of the idiom candidate obtained from the predictive conversion control unit 13, and is obtained from the handwritten character recognition control unit 12 among the idiom candidates. Those matching the radical code are extracted (step S55A). For example, as shown in FIG. 5 (B), the phrase whose second character is “word” is extracted from the idiom candidates.

  In step S53, the display control unit 14 outputs the extracted idiom candidates to the prediction candidate display area 15b.

  As described above, the character input device 1A includes the prediction conversion control unit 13 that predicts the first character, and one of the second characters that follow the first character among the idioms predicted by the prediction conversion control unit 13. And a radical filter unit 18 for extracting idioms including the part. Further, the predictive conversion control unit 13 acquires information on a part of unrecognized characters included in the predicted idiom, and the radical filter unit 18 recognizes the second information recognized as part of information on the unrecognized characters. Based on the part of the character, the idiom including the part of the second character is extracted. The unrecognized character is, for example, one in which the second character in the idiom candidate is “word”.

  According to the character input device 1A, even if the radical prediction vocabulary search DB is not provided in the vocabulary prediction dictionary DB 17A, the idiom candidates can be narrowed down by providing the radical filter unit 18 outside. Therefore, the idiom can be input promptly. Moreover, since radical information included in the idiom is registered in the vocabulary prediction dictionary DB 17A, it is not necessary to provide a radical vocabulary search DB, and the configuration of the vocabulary prediction dictionary DB 17A can be simplified.

(Third embodiment)
In the third embodiment, unlike the first and second embodiments, radical information is not registered in the vocabulary prediction dictionary DB, and character codes corresponding to handwritten recognition character codes are registered. In addition, a radical search database (DB) is added to the character input device.

  FIG. 8 is a block diagram illustrating a configuration example of the character input device 1B according to the third embodiment. The character input device 1B of the third embodiment has substantially the same configuration as the character input devices 1 and 1A of the first and second embodiments. About the same component as 1st, 2nd embodiment, the description is abbreviate | omitted by using the same code | symbol, and a different structure and operation | movement are mainly demonstrated.

  In the vocabulary prediction dictionary DB 17B, idiom candidate character codes corresponding to handwritten recognition character codes are registered. The vocabulary prediction dictionary DB 17B does not hold information related to radicals. Further, a radical search DB 19 is added to the character input device 1B.

  The radical search DB 19 receives the character code of the C character from the radical filter unit 18, searches for radical information (for example, radical code) of the K character of the K character, and searches the radical information for the radical filter unit. Return to 18.

  The radical filter unit 18 obtains the character code of the idiom candidate from the prediction conversion control unit 13 and the radical code from the handwritten character recognition control unit 12, and sends the C character code to the radical search DB 19. Then, the radical filter unit 18 extracts from the idiom candidates those that match the radical information obtained from the radical search DB 19 and sends the narrowed idiom candidate character codes to the display control unit 14.

Next, an operation example of the character input device 1B will be described.
The same processes as those in FIGS. 3 and 4 shown in the first embodiment are denoted by the same step numbers, and the description thereof is omitted. Here, only processing different from the first embodiment will be described. In the third embodiment, the process of step S5B is performed instead of the process of step S5.

  FIG. 9 is a flowchart showing in detail the processing procedure of steps S3 to S5B in the character input processing procedure.

  In step S3, when the handwritten character recognition control unit 12 recognizes the character using the character code DB 16 and extracts the character code, if there is a radical after the second character, the handwritten character recognition control unit 12 uses the radical code DB 16a to Recognize and extract radical codes after the first character.

  Subsequently, in step S4, the display control unit 14 causes the recognized character display area 15a to output an unconfirmed recognized character.

  Subsequently, in step S5B, the prediction conversion control unit 13 predicts and acquires a idiom candidate from the character code of the unconfirmed and completed character (step S50). At this time, when the variable C is 2 or more, the prediction is performed using the 1st to (C-1) th characters. For example, as shown in FIG. 5A, a idiom is predicted using a “kai” as a completed character.

  Subsequently, the handwritten character recognition control unit 12 determines whether or not the variable C is 2 or more (step S51). When the variable C is 1 or less, the character input device 1B performs the process of step S53 described above.

  On the other hand, if the variable C is equal to or greater than 2 in step S51, the radical filter unit 18 extracts the K character of the C character of the idiom candidate obtained from the predictive conversion control unit 13, and sends it to the radical search DB 19 (Step S54B).

  Subsequently, the radical filter unit 18 uses the radical search DB 19 to collate whether or not the kanji of the idiom candidate C character has the radical of the radical code from the handwritten character recognition control unit 12. . Then, the radical filter unit 18 selects only idioms that match the radical code from the idiom candidates. For example, as shown in FIG. 5 (B), the phrase whose second character is “word” is extracted from the idiom candidates.

  In step S53, the display control unit 14 causes the idiom candidate to be output to the prediction candidate display area 15b.

  In this manner, in the character input device 1B, the radical filter unit 18 acquires partial information of unrecognized characters based on the unrecognized characters included in the idiom predicted by the prediction conversion control unit 13. The radical filter unit 18 extracts a idiom including a part of the second character from the idiom candidates based on a part of the second character recognized as a part of the information of the unrecognized character. The partial information of the unrecognized character is, for example, radical information of the C character.

  According to the character input device 1B, since only the idiom candidate character codes corresponding to the handwritten recognition character codes are registered in the vocabulary prediction dictionary DB 17B, a system with high versatility can be used when inputting idioms quickly. can get. That is, since the radical vocabulary search DB is not provided in the vocabulary prediction dictionary DB, the versatility of the predictive conversion engine unit 40 is improved.

(Fourth embodiment)
FIG. 10 is a block diagram illustrating a configuration example of a character input device 1C according to the fourth embodiment. A character input device 1C of the fourth embodiment has substantially the same configuration as the character input device 1 of the first embodiment. About the same component as 1st Embodiment, the description is abbreviate | omitted by using the same code | symbol, and a different structure and operation | movement are mainly demonstrated.

  In the fourth embodiment, compared to the first embodiment, the radical code DB is not provided in the character code DB 16A. The character input device 1 </ b> C is provided with a relay processing unit 21 and a radical estimation database (DB) 22.

  The handwritten character recognition control unit 12 of this embodiment is an example of an entire character recognition unit. The relay processing unit 21 is an example of a partial character recognition unit. The radical estimation DB 22 is an example of an accumulation unit.

  A radical code corresponding to the handwritten recognition character code is registered in the radical guess DB 22. For example, the radical estimation DB 22 determines that the handwritten character recognition control unit 12 once recognizes “Kasa” and the relay processing unit 21 sets “Kakukanmuri” for the second character “S”. Used to infer that it was misrecognized. Similarly, for example, the radical estimation DB 22 is used to estimate a character mistakenly recognized as “P (Roman character)” as “Kozatohen”.

  The handwritten character recognition control unit 12 recognizes “word (kanji)” when, for example, “Gonben” is input by handwriting into the handwriting region 15 c. Similarly, if “Kusakamuri” is entered by handwriting, “Kaka” will be entered. recognize.

  In this way, even if any radical is input by handwriting, the handwritten character recognition control unit 12 erroneously recognizes it as another character. The radical estimation DB 22 holds a kanji radical that is misrecognized by another character and the other character in association with each other. That is, the radical estimation DB 22 statistically collects patterns that are easily misrecognized and accumulates them in advance.

  Thereby, the radical estimation DB 22 can input the handwritten recognition character code from the relay processing unit 21 and output the radical code corresponding to the handwritten recognition character code to the relay processing unit 21. Therefore, if, for example, “sentence a” is input by handwriting input, it can be estimated that the second character kana “a” has been misrecognized by the radical estimation DB 22. Accordingly, the character input device 1C can predict “culture”, “example sentence”, “style”, and the like.

  The relay processing unit 21 acquires a handwritten recognition character code from the handwritten character recognition control unit 12 and sends it to the radical estimation DB 22. Further, when receiving the radical code from the radical estimation DB 22, the relay processing unit 21 sends the handwritten recognition character code and the radical code to the prediction conversion control unit 13.

Next, an operation example of the character input device 1C will be described.
About the same step process as FIG.3, FIG.4 shown in 1st Embodiment, the description is abbreviate | omitted by attaching | subjecting the same step number. Here, only processing different from the first embodiment will be described. In the third embodiment, steps S3A and S5C are performed instead of steps S3 and S5, respectively.

  FIG. 11 is a flowchart showing in detail the processing procedure of steps S3A to S5C in the character input processing.

  When the coordinate information of the handwritten character is received from the touch panel handwriting detection unit 11 in step S2 of FIG. 3 described above, the handwritten character recognition control unit 12 recognizes the character using the character code DB 16A and extracts the handwritten recognition character code. (Step S3A).

  Subsequently, in step S4, the display control unit 14 causes the recognized character display area 15a to output an unconfirmed recognized character.

  Subsequently, in step S5C, when the variable C is 2 or more in step S51, the relay processing unit 21 inquires the radical estimation DB 22 and extracts the radical code inferred from the character code of the C character ( Step S54C). For example, the radical code of “Gomben” is extracted.

  Subsequently, the relay processing unit 21 sends the handwritten recognition character code and the radical code to the prediction conversion control unit 13 (step S54).

  Subsequently, the predictive conversion control unit 13 searches the vocabulary prediction dictionary DB 17 from the 1st to (C-1) th character codes of the indeterminate and completed (recognized) characters and the radical code of the C character, The narrowed idiom candidates are acquired (step S55). For example, a search is performed with the completed characters “meeting” + “word”.

  In step S53, the display control unit 14 causes the idiom candidate to be output to the prediction candidate display area 15b.

  According to the character input device 1C, the handwritten recognition character code corresponding to the coordinate information of the handwritten character is only required to be registered in the character code DB 16A. Is obtained. Further, since the radical code DB is not provided at 16A in the character code DB, the versatility of the character recognition engine unit 30 is improved.

(Fifth embodiment)
In the fifth embodiment, a character input device in which the character input devices of the second embodiment and the fourth embodiment are combined is shown.

  FIG. 12 is a block diagram illustrating a configuration example of a character input device 1D according to the fifth embodiment. In the character input device 1D of the fifth embodiment, the same reference numerals are used for the same components as those of the character input devices 1, 1A, 1C of the first, second, and fourth embodiments.

  In the fifth embodiment, compared to the first embodiment, the radical code DB is not provided in the character code DB 16A. In addition, the character input device 1D is provided with a relay processing unit 21 and a radical estimation DB 22.

  In the vocabulary prediction dictionary DB 17A, the idiom candidate character code corresponding to the handwritten recognition character code and the radical information (for example, radical code) of the kanji included in the word (idiom candidate) are registered.

  Further, a radical filter unit 18 is added to the character input device 1D. The radical filter unit 18 acquires from the predictive conversion control unit 13 the character code of the idiom candidate and the radical information of the K character of the C character included in the idiom. Further, the radical filter unit 18 acquires a radical code from the relay processing unit 21. The radical filter unit 18 searches the radical information of the K character of the C character of the idiom candidate, extracts the one that matches the radical code from the predictive conversion control unit 13, and selects the narrowed idiom candidate character code This is sent to the display control unit 14.

Next, an operation example of the character input device 1D will be described.
About the same step process as FIG.3, FIG.4 shown in 1st Embodiment, the description is abbreviate | omitted by attaching | subjecting the same step number. Here, only processing different from the first embodiment will be described. In the fifth embodiment, steps S3A and S5D are performed instead of steps S3 and S5, respectively.

  FIG. 13 is a flowchart showing in detail the processing procedure of steps S3A to S5D in the character input processing procedure.

  When the handwritten character recognition control unit 12 receives the coordinate information of the handwritten character in the above-described step S2, the handwritten character recognition control unit 12 recognizes the character using the character code DB 16A and extracts the handwritten recognition character code ( Step S3A).

  Subsequently, in step S4, the display control unit 14 causes the recognized character display area 15a to output an unconfirmed recognized character.

  Subsequently, in step S5D, when the variable C is equal to or greater than 2 in step S51, the relay processing unit 21 inquires the radical estimation DB 22 and extracts the radical code inferred from the character code of the C character ( Step S54C). For example, the radical code of “Gomben” is extracted.

  Subsequently, the predictive conversion control unit 13 acquires the idiom candidate and the radical information of the kanji included in the idiom from the character code of the indeterminate and completed (authenticated) character (step S54A). When the variable C is 2 or more, the predictive conversion control unit 13 predicts using the 1st to (C-1) th characters. For example, as shown in FIG. 5A, a idiom is predicted using a “kai” as a completed character.

  Subsequently, the radical filter unit 18 searches the radical information of the K character of the C character of the idiom candidate obtained from the predictive conversion control unit 13, and matches the radical code obtained from the relay processing unit 21. Is extracted (step S55A). For example, as shown in FIG. 5 (B), a phrase whose second character is “word” is extracted from the idiom candidate group.

  In step S53, the display control unit 14 causes the idiom candidate to be output to the prediction candidate display area 15b.

  According to the character input device 1D, since the radical code DB is not provided in the character code DB 16A, the versatility of the character recognition engine unit 30 is improved. Further, since the radical vocabulary search DB is not provided in the vocabulary prediction dictionary DB 17A and radical information is registered, the configuration of the vocabulary prediction dictionary DB can be simplified. That is, even if the radical vocabulary search DB is not provided, the idiom candidates can be narrowed down by providing the radical filter unit 18 outside. Therefore, the idiom can be input promptly. In addition, a highly versatile system can be obtained.

(Sixth embodiment)
In the sixth embodiment, a character input device in which the character input devices of the third embodiment and the fourth embodiment are combined is shown.

  FIG. 14 is a block diagram illustrating a configuration example of a character input device 1E according to the sixth embodiment. In the character input device 1E of the sixth embodiment, the same reference numerals are used for the same constituent elements as those of the character input devices 1, 1B, 1C of the first, third, and fourth embodiments.

  In the sixth embodiment, compared to the first embodiment, the radical code DB is not provided in the character code DB 16A. The character input device 1E is provided with a relay processing unit 21 and a radical estimation database (DB) 22.

  In the vocabulary prediction dictionary DB 17B, idiom candidate character codes corresponding to handwritten recognition character codes are registered. The vocabulary prediction dictionary DB 17B does not hold information related to radicals.

  Further, a radical filter unit 18 and a radical search DB 19 are added to the character input device 1E. The radical filter unit 18 acquires the character code of the idiom candidate from the prediction conversion control unit 13. Further, the radical filter unit 18 acquires a radical code from the relay processing unit 21.

  The radical search DB 19 receives the character code of the C character of the idiom candidate from the radical filter unit 18 and receives the radical code from the relay processing unit 21. The radical search DB 19 searches for radical information (for example, radical code) of the K-th character Kanji and returns the matched radical information to the radical filter unit 18.

  The radical filter unit 18 acquires the character code of the idiom candidate from the prediction conversion control unit 13, acquires the radical code from the relay processing unit 21, and sends the C character code to the radical search DB 19. Then, the radical filter unit 18 extracts from the idiom candidates those that match the radical information obtained from the radical search DB 19 and sends the narrowed idiom candidate character codes to the display control unit 14.

Next, an operation example of the character input device 1E will be described.
About the same step process as FIG. 3, FIG. 4 shown in 1st Embodiment, the description is abbreviate | omitted by attaching | subjecting the same step number. Here, only processing different from the first embodiment will be described. In the sixth embodiment, steps S3A and S5E are performed instead of steps S3 and S5, respectively.

  FIG. 15 is a flowchart showing in detail the processing procedure of steps S3A to S5E in the character input processing procedure.

  When the coordinate information of the handwritten character is received from the touch panel handwriting detection unit 11 in step S2, the handwritten character recognition control unit 12 recognizes the character using the character code DB 16A and extracts the handwritten recognition character code (step). S3A).

  Subsequently, in step S4, the display control unit 14 causes the recognized character display area 15a to output an unconfirmed recognized character.

  Subsequently, in step S5E, the predictive conversion control unit 13 acquires a idiom candidate from the character code of an indeterminate and completed (authenticated) character (step S50). When the variable C is 2 or more, the prediction conversion control unit 13 predicts using the first to (C-1) th characters. For example, as shown in FIG. 5A, a idiom is predicted using a “kai” as a completed character.

  Subsequently, the handwritten character recognition control unit 12 determines whether or not the variable C is 2 or more (step S51). When the variable C is 1 or less, the process of step S53 described above is performed.

  On the other hand, when the variable C is 2 or more in step S51, the relay processing unit 21 inquires the radical estimation DB 22 and extracts the radical code inferred from the C character code (step S54C). For example, the radical code of “Gomben” is extracted.

  Subsequently, the radical filter unit 18 uses the radical search DB 19 to search for radical information of the K character of the C character of the idiom candidate obtained from the prediction conversion control unit 13, and the searched radical information Those matching the radical code obtained from the relay processing unit 21 are extracted (step S55A). For example, as shown in FIG. 5 (B), a phrase whose second character is “word” is extracted from the idiom candidate group.

  In step S53, the display control unit 14 causes the idiom candidate to be output to the prediction candidate display area 15b.

  According to the character input device 1E, since the radical code DB is not provided in the character code DB 16A, the versatility of the character recognition engine unit 30 is improved. Further, since the radical vocabulary search DB is not provided in the vocabulary prediction dictionary DB 17B, the versatility of the prediction conversion engine unit 40 is improved.

  Therefore, in the character input device 1E, an external device including the relay processing unit 21, the radical estimation DB 22, the radical filter unit 18, and the radical search DB 19 is added to the conventional character input device, so that idioms can be input quickly. Can be done. Therefore, an extremely versatile system can be obtained.

(Seventh embodiment)
In the seventh embodiment, the lexical prediction dictionary DB is added with a function (fuzzy search function) for additionally registering idiom candidates associated with predicted radicals in the prediction conversion dictionary.

  FIG. 16 is a block diagram illustrating a configuration example of the character input device 1F according to the seventh embodiment. The character input device 1F has substantially the same configuration as the character input device 1F of the first embodiment. About the same component as 1st Embodiment, the description is abbreviate | omitted by using the same code | symbol.

  The character input device 1F is provided with a vocabulary prediction dictionary DB 17C to which an ambiguous search function is added.

Next, an operation example of the character input device 1F will be described.
About the same step process as FIG.3, FIG.4 shown in 1st Embodiment, the description is abbreviate | omitted by attaching | subjecting the same step number. Here, only processing different from the first embodiment will be described. In the seventh embodiment, steps S3A and S5F are performed instead of steps S3 and S5, respectively.

  FIG. 17 is a flowchart showing in detail the processing procedure of steps S3A to S5F in the character input processing procedure.

  In step S4, the display control unit 14 causes the recognized character in the unconfirmed state to be output to the recognized character display area 15a.

  Subsequently, the handwritten character recognition control unit 12 sends the handwritten recognition character code to the prediction conversion control unit 13 (step S56).

  Subsequently, the predictive conversion control unit 13 acquires idiom candidates (step S57). At this time, the vocabulary prediction dictionary DB 17C has already expanded the registered vocabulary of the prediction dictionary, and the predictive conversion control unit 13 extracts the expanded idioms in anticipation of the possibility that the C-th character is a radical.

  In step S53, the display control unit 14 causes the idiom candidate to be output to the prediction candidate display area 15b.

  According to the character input device 1F, the device configuration can be simplified by adding a fuzzy search function to the vocabulary prediction dictionary DB 17C.

  The present invention is not limited to the configuration of the above-described embodiment, and any configuration can be used as long as the functions shown in the claims or the functions of the configuration of the present embodiment can be achieved. Is also applicable.

  The handwriting detection unit in each of the above embodiments detects the handwriting by using the touch or proximity of a finger or stylus to the touch panel as an input, but is not particularly limited thereto. For example, the handwriting may be detected by reading the projected light.

  In each of the above embodiments, the case where one radical, such as “Gomben”, is used as a part of the character has been described. However, the idiom is expressed using a combination of two or more radicals. You may make it extract. Moreover, you may use a part of radical or a part of characteristic Kanji other than a radical.

  Moreover, the character input device of each said embodiment is applicable to various apparatuses other than portable information terminals, such as a smart phone and a tablet terminal.

  The present invention also includes a program that supplies a character input program that realizes the functions of the above-described embodiments to a character input device via a network or various storage media, and that is read and executed by a computer in the character input device. It is.

  INDUSTRIAL APPLICABILITY The present invention is useful for a character input device, a character input method, a character input program, and the like that can shorten the time required to input a idiom.

DESCRIPTION OF SYMBOLS 1, 1A, 1B, 1C, 1D, 1E, 1F Character input device 11 Touch panel handwriting detection part 11a Touch panel 12 Handwritten character recognition control part 13 Predictive conversion control part 14 Display control part 15 Display part 15a Recognition character display area 15b Prediction candidate display Area 15c Handwritten area 15d Operation key screen area 15e Recognized character candidate display area 16, 16A Character code DB
16a radical code DB
17, 17A, 17B, 17C Vocabulary prediction dictionary DB
17a radical vocabulary search DB
18 radical filter part 19 radical search DB
21 relay processing unit 22 radical estimation DB
25 Operation key screen area 25a Confirmation key 25b, 25c Arrow key 25d Full-width / half-width key 25e Character candidate key 25f Delete key 30 Character recognition engine section 40 Predictive conversion engine section

Claims (11)

A character input device for inputting characters,
A handwriting detector for detecting handwriting;
A character recognition unit that recognizes the handwriting detected by the handwriting detection unit as a character;
A idiom prediction unit that predicts an idiom including a first character recognized by the character recognition unit and a part of a second character subsequent to the first character;
A character input device comprising: The character input device according to claim 1, further comprising:
A character confirmation unit for confirming the character recognized by the character recognition unit;
The idiom prediction unit is a character input device that predicts a idiom that is recognized by the character recognition unit and includes a part of the first character and the second character that has not been determined by the character determination unit. The character input device according to claim 1 or 2,
The character recognition unit is a character input device that recognizes the first character and a part of the second character. The character input device according to claim 1 or 2,
The character recognition unit recognizes the first character, recognizes a part of the second character as another character, and recognizes the character recognized as the other character as the second character. A character input device comprising: a partial character recognition unit that recognizes a part of a character. The character input device according to claim 4, further comprising:
A storage unit that stores other characters recognized by the whole character recognition unit and a part of the second character in association with each other;
The partial character recognition unit is a character input device for recognizing a character recognized as the other character as a part of the second character based on information stored in the storage unit. The character input device according to any one of claims 3 to 5,
The idiom prediction unit includes a first idiom prediction unit that predicts a idiom including the first character, and a idiom including a part of the second character among the idioms predicted by the first idiom prediction unit. A character input device comprising: a second idiom prediction unit to extract. The character input device according to claim 6,
The first idiom prediction unit predicts a idiom that includes the first character, acquires information on a part of unrecognized characters included in the predicted idiom,
The second idiom prediction unit is based on a part of the unrecognized character information acquired by the first idiom prediction unit and a part of the second character recognized by the character recognition unit, A character input device for extracting a idiom including a part of the second character from idioms predicted by the first idiom prediction unit. The character input device according to claim 6,
Acquiring a part of information of the unrecognized character based on the unrecognized character included in the idiom predicted by the first idiom prediction unit; and a part of the information of the acquired unrecognized character; and the character recognition A character input device that extracts a idiom including a part of the second character from idioms predicted by the first idiom prediction unit based on the part of the second character recognized by the part. A character input device according to any one of claims 1 to 8,
A character input device in which a part of the character is a radical of the character. A character input method for inputting characters,
Detecting handwriting; and
Recognizing the detected handwriting as a character;
Predicting an idiom that includes the recognized first character and a portion of a second character that follows the first character;
A character input method.   A character input program for causing a computer to execute each step of the character input method according to claim 10.

Images