JP2013214187A - Character input device, method for controlling character input device, control program and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To greatly reduce a control amount when inputting a character code by handwriting input.SOLUTION: A character input device 100A includes a hint information recognition unit 128 for recognizing hint information c4 input by handwriting, and a candidate control unit 124 that performs extraction of a candidate character from a recognition character obtained by recognizing a hand written character, on the basis of the hint information c4.

Description

  The present invention relates to a character input device, a character input device control method, a control program, and a recording medium that can perform handwritten character recognition and conversion operations by handwriting input without taking time and effort.

  Conventionally, there are several methods for inputting characters on a smartphone terminal or tablet terminal equipped with a touch panel display. For example, if a pen-type device or a finger is used as an input means and moved while touching the touch panel, the character is recognized from the trajectory, and conversion candidate characters (kanji, hiragana, kana, alphanumeric, etc.) are presented to the user. There is a method of inputting characters or sentences by selecting from the candidates by input means. There is also a method of saving a trace traced with a finger or pen as a figure (vector data or the like) or an image as it is. There is also a method of inputting characters or sentences by displaying a soft keyboard on a touch panel and selecting a key by the input means.

  Patent Document 1 describes a character recognition device that allows an operator to smoothly input characters by hand without worrying about the result of character recognition. Specifically, the character recognition device reads and displays a character image stored in the input image memory at the same location as the character determined to be erroneous by the erroneous recognition determination means.

  Further, Patent Document 2 describes a handwritten character clearing device that can print a character as an image of a handwritten character when the handwritten character recognizing unit is likely to be erroneously recognized and does not cause an error in the output result. Specifically, the handwritten character clearing device converts the handwritten character into code data for a handwritten character having a reliability of character recognition by the handwritten character recognizing means of a predetermined value or more, and character recognition by the handwritten character recognizing means. As for the handwritten character whose credit rate is lower than a predetermined value, the read data read by the reading means is assigned as data representing the handwritten character.

JP 05-108771 A (published April 30, 1993) JP 06-096268 A (published on April 8, 1994)

  However, the conventional configuration has the following problems.

  First, by recognizing characters from a trace traced with a finger or pen on the display and selecting characters to be entered from the suggested character candidates, the characters are misrecognized and cannot be entered correctly, and rewritten many times. Conversion may not be possible without re-recognition. In addition, it is necessary to input characters in units of characters, recognize characters, select an intended character from conversion candidates, and confirm it, and it is necessary to repeat this operation for the number of input characters. That is, there is a problem that an input operation in units of one character increases the operation amount. Furthermore, when there are many conversion candidates, there is a problem that it takes time to select a correct candidate.

  In addition, the method of storing a trace traced with a finger or a pen as a figure (vector data or the like) or an image as it is is intuitive but is not suitable for inputting a long sentence. In addition, there is a problem that it is difficult to retrieve and reuse a figure or image data that is not a character code.

  In addition, as for the method of displaying the soft keyboard on the touch panel, there are many cases where the soft keyboard is switched depending on the character type (kana, numbers, English, etc.) to be entered. A very complicated operation is required. In addition, unlike hard keyboards, touch typing is not possible, so inputting while looking at a soft keyboard is not intuitive, and there is also a problem that operation efficiency tends to decrease due to accidentally pressing an adjacent key. In addition, in the case of a long character string, a correct conversion candidate may be presented by changing the length of the character string to be converted, but there is a problem that the operation is very complicated and requires a lot of time.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to control a character input device and a character input device that can greatly reduce an operation amount when inputting a character code by handwriting input. To provide a method, a control program, and a recording medium.

  In order to solve the above problems, a character input device according to the present invention is a character input device that inputs characters by recognizing handwritten characters, and recognizes handwritten characters and generates a plurality of recognized characters. Recognizing means, candidate character extracting means for extracting candidate characters from the recognized characters generated by the character recognizing means, the candidate characters extracted by the candidate character extracting means are presented to the user, and the candidate characters selected by the user are A character confirmation unit that identifies a confirmed character corresponding to a handwritten character; and a hint information recognition unit that recognizes handwritten input hint information. The candidate character extraction unit includes the hint information recognized by the hint information recognition unit. Based on the above, the candidate characters are extracted from the recognized characters.

  A control method for a character input device according to the present invention is a control method for a character input device that performs character input by recognizing a handwritten character, and recognizes a handwritten character to generate a plurality of recognized characters. A candidate for extracting candidate characters from the recognized characters generated in the character recognition step based on the step, a hint information recognition step for recognizing handwritten input hint information, and the hint information recognized in the hint information recognition step A character extraction step; and a character confirmation step of presenting the candidate character extracted in the candidate character extraction step to the user and identifying the candidate character selected by the user as a confirmed character corresponding to the handwritten character. It is said.

  By the above configuration and method, hints input by handwriting are extracted when candidate characters to be presented for allowing the user to select a confirmed character corresponding to the handwritten character are extracted from a plurality of recognized characters generated by recognizing the handwritten character. Candidate characters can be extracted based on the information. Here, as the hint information, for example, when the handwritten character is a Chinese character, “reading” or “radical” of the Chinese character can be used. The hint information may be input before the input of handwritten characters or after confirming erroneous recognition of the handwritten characters.

  In this way, by using the hint information for extracting the candidate character, even if the recognition character has low recognition accuracy indicating the recognition accuracy because the handwritten character is input in a shape that makes character recognition difficult, It is possible to narrow down based on the hint information and present it to the user preferentially. Therefore, unlike the prior art, characters are misrecognized and cannot be input correctly, so there is no need to rewrite and re-recognize them.

  Further, unlike the prior art, there is no need to switch keyboard settings according to the character type (kana, numbers, alphabet, etc.) to be input, and no extra operation is required.

  Therefore, even if the handwritten character is ambiguous, it is possible to present a correct recognized character, so that the recognition rate can be improved. Moreover, since the operation for that is also hand-inputting hint information, it becomes possible to maintain the continuity of a handwritten character input and operation, and can reduce the effort concerning operation greatly.

  Furthermore, the character input device according to the present invention is characterized in that the character determination means presents a set of a plurality of candidate characters corresponding to a plurality of handwritten characters input continuously.

  According to the above configuration, in order to present a set of a plurality of candidate characters corresponding to a plurality of handwritten characters that are continuously input, it is possible to select a confirmed character corresponding to the handwritten characters. This can be done in units of character strings. As a method for collecting a plurality of candidate characters into one set, for example, morphological analysis can be used.

  Therefore, handwritten characters can be recognized with high accuracy, and even when erroneous recognition is corrected, it can be efficiently corrected to correct recognized characters.

  Furthermore, in the character input device according to the present invention, the hint information recognition means acquires hint information type information indicating the type of the hint information together with the hint information, and the hint information is obtained based on the hint information type information. It is characterized by recognition.

  With the above configuration, the handwritten input hint information can be recognized based on the type of hint information indicated by the hint information type information. For example, when a handwritten character is a Chinese character and the radical of the Chinese character is input by handwriting as the hint information, “radical” is designated as the hint information type information. Thereby, since the handwritten input input as hint information can be recognized as “radical”, the probability that the hint information can be correctly recognized is increased.

  Therefore, handwritten characters can be recognized with high accuracy, and even when erroneous recognition is corrected, it can be efficiently corrected to correct recognized characters.

  Furthermore, in the character input device according to the present invention, the hint information type information is preferably at least one of kanji reading, kanji radical, character type, and character size.

  The character input device may be realized by a computer. In this case, the control program of the input device that causes the input device to be realized by the computer by operating the computer as each of the means, and recording the program. Such computer-readable recording media also fall within the scope of the present invention.

  As described above, the character input device according to the present invention includes a character recognition unit that recognizes handwritten characters and generates a plurality of recognized characters, and a candidate character extraction that extracts candidate characters from the recognized characters generated by the character recognition unit. A character confirming means for presenting the candidate character extracted by the means, the candidate character extracting means to the user, and identifying the candidate character selected by the user as a confirmed character corresponding to the handwritten character, and handwritten input hint information A hint information recognizing unit for recognizing, and the candidate character extracting unit extracts the candidate character from the recognized character based on the hint information recognized by the hint information recognizing unit.

  The character input device control method according to the present invention includes a character recognition step for recognizing handwritten characters to generate a plurality of recognized characters, a hint information recognition step for recognizing handwritten input hint information, and the hint information. Based on the hint information recognized in the recognition step, the candidate character extraction step for extracting the candidate character from the recognized character generated in the character recognition step and the candidate character extracted in the candidate character extraction step are presented to the user. And a character confirmation step for identifying a candidate character selected by the user as a confirmed character corresponding to the handwritten character.

  Therefore, even if the handwritten character is ambiguous, it is possible to present a correct recognized character, and the recognition rate can be improved. Moreover, since the operation for that is also handwritten input of hint information, it becomes possible to maintain the continuity of the input of the handwritten character and the operation, and it is possible to greatly reduce the labor required for the operation.

It is a functional block diagram which shows the principal part of the example of 1 structure of the character input device which concerns on embodiment of this invention. FIG. It is the front view showing the example of appearance of the character input device concerning the embodiment of the present invention. It is a flowchart which shows an example of the flow of the handwriting input process in the character input device shown in FIG. It is a transition diagram of the screen displayed in the case of the handwriting input process in the character input device shown in FIG. 1, (a) shows the example of a screen which recognized the handwritten character and displayed the 1st candidate character, (b) The example of a screen which displayed the next candidate character with respect to the misrecognized character designated from the 1st candidate character is shown, (c) shows the example of a screen which updated the next candidate character based on the input hint information, (d ) Shows an example screen when the recognized character is confirmed. It is a functional block diagram which shows the principal part of the other structural example of the character input device which concerns on embodiment of this invention. FIG. 6 is a screen example displayed during handwriting input processing in the character input device shown in FIG. 5, and shows a screen example in which the next candidate character is updated based on the input hint information and hint information type information. It is a functional block diagram which shows the principal part of the further another structural example of the character input device which concerns on embodiment of this invention. FIG. 8 is a transition diagram of a screen displayed at the time of handwriting input processing in the character input device shown in FIG. 7, (a) shows an example of a screen that recognizes a handwritten character and displays a first candidate character and a morphological analysis result; (B) shows an example of a screen displaying a part-of-speech fine classification selection area when an erroneously recognized character (string) is designated from the morphological analysis result, and (c) is an erroneously recognized character (string) designated from the morphological analysis result. ) Shows an example of a screen displaying the next candidate character string. It is explanatory drawing of the process at the time of inputting a handwritten character by data into the said character input device, (a) shows the example of a screen when the character converted into the handwritten character and the character code is included in data, (b) Shows an example of a screen in the case where print data and handwritten characters are included in the data.

  Based on FIGS. 1-9, embodiment of this invention is described in detail.

[Outline of character input device]
First, with reference to FIG. 2, the external appearance of the character input device 100 according to the present embodiment will be described. FIG. 2 is a front view illustrating an appearance example of the character input device 100.

  The character input device 100 is a device that performs character input by recognizing the handwritten character c1. Therefore, the character input device 100 can input characters based on operations on the input unit 110. As shown in FIG. 2, the character input device 100 includes an input unit 110 that can detect a user operation, a display unit 130 that can display an area and buttons input by the user, characters input by the user, recognition candidates, and the like. Is provided. For example, when the input unit 110 is realized by a touch panel and the display unit 130 is realized by a liquid crystal display, both can be integrated as shown in FIG. In the present embodiment, the input unit 110 and the display unit 130 configured as a single unit are collectively referred to as a display input unit DI.

  As shown in FIG. 2, the display input unit DI includes a handwriting input area A111 and a hint information input area A113 on the screen. The handwriting input area A111 is an area for inputting a handwritten character c1 to be recognized in order for the user to input a desired character. In the hint information input area A113, the user inputs hint information c4 for recognizing the handwritten character c1 input in the handwriting input area A111 or for selecting the next candidate character (candidate character) c3 as a recognition result. It is an area to do.

  The character input device 100 recognizes an object (figure) input from the handwriting input area A111 in units of characters and presents one first candidate character (candidate character) c2 that is a recognition result. If the presented first candidate character c2 is not intended, the user displays and selects one or more other next candidate characters c3 below the second candidate in character units, so that the characters can be selected in character units. You can enter. Alternatively, it is possible to change the characters by re-inputting them individually and making the characters recognized again.

  In the handwriting input area A111, one or more character recognition candidates (first candidate character c2, next candidate character c3) are presented in units of characters under the handwritten character c1. When there are a plurality of character recognition candidates, the user can select one of them. The character recognition candidate can present a character code string that is optimal for conversion from the recognized character string according to the priority by referring to the dictionary database and operation history (for example, the frequency of character use). Conversion candidates can be presented by so-called predictive conversion or normal conversion. The display of the next candidate character c3 below the second candidate is not essential. The character may be input again and recognized again.

  In addition, the character input device 100 recognizes the object input from the hint information input area A113 in units of characters, and recognizes the handwritten character c1 input to the handwriting input area A111 and the hint for selecting the next candidate character c3. Let it be information c4.

  For example, by handwriting input with handwriting input from the handwriting input area A111 and writing kanji characters that are not known from the hint information input area A113 as handwriting input with hiragana as hint information c4, the hiragana is converted into kanji and the whole handwritten input Can be entered in Kanji. In addition, a part of a kanji character is input as hint information c4 from the hint information input area A113 from a recognized character recognized by reading a kanji character or the like input from the handwriting input area A111, and a candidate character selected based on the kanji is presented. Can do. In addition, handwritten input of English words from the handwriting input area A111 and handwriting input of a Japanese translation as hint information c4 from the hint information input area A113 makes it possible to refer to the Japanese-English dictionary from the Japanese translation and spell the English words correctly. Can be entered.

  Below the hint information input area A113, a hint information type selection area A116 is displayed. The hint information type selection area A116 is an area in which the user specifies the type of object that is the hint information c4 input from the hint information input area A113. For example, when inputting a kanji reading as the hint information c4 in the hint information input area A113, “reading” is input from the hint information type selection area A116. In addition, when inputting a part of Chinese characters as the hint information c4 in the hint information input area A113, “radical” is input from the hint information type selection area A116. When inputting a Japanese translation as the hint information c4 in the hint information input area A113, “Japanese translation” is input from the hint information type selection area A116.

  As described above, as a result of character recognition of the handwritten character c1 input from the handwriting input area A111, at least the first candidate character c2 that is the first candidate is displayed for each character. Further, as described above, the next candidate character c3 below the second candidate may be displayed at the same time, and a user tap operation (down, up), long tap (down), etc. is displayed as a trigger. You may make it do. In addition, a button that does not convert character units (leave an object) may be displayed together with the next candidate character c3. This is effective when, for example, only some characters include external characters. Then, when the user selects the second candidate or lower by an operation such as tapping, the selected character is displayed as the first candidate.

  Furthermore, a character recognition confirmation button A115 for confirming a character code with the first candidate character c2 (including a character selected as the first candidate from the next candidate character c3) as a result of character recognition may be provided. Confirmation of conversion to character code may be performed by a specific operation such as tapping or double tapping in a specific place on the screen or where no object is displayed, or a specific gesture is input. May be executed. Moreover, you may make it determine automatically what was displayed as the 1st candidate character c2 for the predetermined time. Furthermore, when a predetermined number of characters have been written, it may be determined sequentially from those recognized in the past. In addition, when input from the handwriting input area A111 is started in a state where the first candidate character c2 is changed in accordance with the input to the hint information input area A113, the character code is determined with the first candidate character c2. May be.

  In addition, if the input device 100 does not want to convert the input handwritten character c1 (object) into a character code, the character input device 100 determines that the character is not converted and stores it as unconverted object information (shape and position) (see FIG. 1) or the like. Further, the converted object information may be stored in association with the converted character code string.

  In FIG. 2, the handwriting input area A111 and the hint information input area A113 are displayed on one display input unit DI, but they may be displayed separately. Further, for example, the handwritten character c1 may be displayed in a form superimposed on the display of the character code conversion result. In addition, the handwritten character c1 may be superimposed on an area for displaying the converted / unconverted result, or the area may be divided.

  For input other than handwriting input to the handwriting input area A111 and hint information input area A113, for example, input to the hint information type selection area A116, etc., numbers are written in each object, and numeric keys by hardware or software are used. You may make it selectable via.

  Hereinafter, a specific configuration example of the character input device 100 according to the present embodiment will be described.

[Configuration example 1]
A character input device 100A according to Configuration Example 1 will be described in detail with reference to FIGS. In the following description, “A” is added to a member specific to this configuration example, as in “input unit 110A”. Note that items common to members of other configuration examples may be collectively referred to as “input unit 110” with “A” omitted.

(Configuration of character input device)
First, the configuration of the character input device 100A will be described in detail with reference to FIG. FIG. 1 is a functional block diagram showing a main configuration of the character input device 100A. Note that portions not directly related to this configuration example (for example, a portion that realizes dragging of input characters, etc.) are omitted from the description of the configuration and the block diagram from the viewpoint of ensuring the simplicity of description. However, the character input device 100A may include the omitted configuration in accordance with the actual situation.

  As illustrated in FIG. 1, the character input device 100A includes an input unit 110A, a control unit 120A, a display unit 130, and a storage unit 140.

  The input unit 110A is used by the user to input for operating the character input device 100A. Specifically, the input unit 110A includes a handwritten character input unit 111, a character designation unit 112, a hint information input unit 113, a candidate selection unit 114, and an overall determination unit 115. It should be noted that display on the display unit 130 of areas and characters corresponding to each functional block of the input unit 110A is performed by the functional block or a functional block (not shown) corresponding to the functional block.

  The handwritten character input unit 111 acquires a locus handwritten by the user in the handwritten input area A111 as a handwritten character c1 that is an object (figure). The handwritten character c1 may include information such as an object shape and a writing order.

  When the first candidate character c2, which is the first candidate of the recognition result of the handwritten character c1, is erroneously recognized, the character designation unit 112 selects the user as a target for re-recognition, for example, by tapping the display of the character. Let

  The hint information input unit 113 acquires a locus handwritten by the user in the hint information input area A113 as hint information c4 that is an object (figure). The hint information c4 is used to narrow down the next candidate character c3 that is a re-recognition candidate. The hint information c4 represents, for example, kanji readings or phonetic readings, kanji radicals, character types, character sizes, and the like.

  Since the candidate selection unit 114 is a misrecognized character, the user selects the next candidate character c3 below the second candidate of the recognition result displayed below the first candidate character c2 selected by the user as a target of re-recognition. The user is made to select a properly recognized character by tapping the display of the character.

  When the user taps the character recognition confirmation button A115 displayed on the screen or the like, the overall confirmation unit 115 recognizes the character (code) recognized from the handwritten character c1 by the character (code) of the current first candidate character c2. ) Is determined to the control unit 120A.

  The controller 120A executes various controls in the character input device 100A. Therefore, the control unit 120A includes an input character recognition unit (character recognition unit) 121, a character recognition unit 122, a recognized character memory 123, a candidate control unit (candidate character extraction unit) 124, and a character input display setting unit (character confirmation unit) 125. , A display control unit 126, a character editing display setting unit 127, and a hint information recognition unit (hint information recognition means) 128A.

  The input character recognition unit 121 recognizes the handwritten character c1 and generates one or a plurality of recognized characters. Specifically, the input character recognition unit 121 acquires the handwritten character c <b> 1 input from the handwritten character input unit 111, and stores the character code resulting from the recognition by the character recognition unit 122 in the recognized character memory 123. Further, the input character recognition unit 121 displays the handwritten character c <b> 1 as an object on the display unit 130 via the display control unit 126.

  The character recognizing unit 122 acquires the handwritten character c1 as an object from the input character recognizing unit 121, and returns the result recognized in character units as a character code. Also, the hint information c4 is acquired as an object from the hint information recognition unit 128A, and the result of recognition in character units is returned.

  The recognized character memory 123 holds one or more recognized characters (character codes), which are the recognition results of each character, in units of characters cut out from the handwritten character c1.

  The candidate control unit 124 manages recognized characters stored in the recognized character memory 123. In particular, the candidate control unit 124 extracts candidate characters (first candidate character c2, next candidate character c3) from the recognized characters generated by the input character recognition unit 121 based on the hint information c4 recognized by the hint information recognition unit 128. To do. Specifically, the candidate control unit 124 holds a list in which recognized characters are arranged in order of recognition accuracy indicating the probability of recognition for each handwritten character. Then, the first candidate character c2 and the next candidate character c3 are extracted in the order of recognition accuracy, and returned in response to a request from the character input display setting unit 125 or the character editing display setting unit 127. Further, the recognition accuracy is corrected in accordance with the condition given from the character editing display setting unit 127, and the first candidate character c2 and the next candidate character c3 are extracted again. In the conventional character recognition, if the recognition accuracy (reliability) indicating the probability of recognition is lower than a predetermined value, it is excluded from the candidate characters, but the candidate control unit 124, when the hint information is input Even if the recognition accuracy is lower than a predetermined value, candidate characters are narrowed down.

  The character input display setting unit 125 sets a display form to be displayed on the display unit 130 via the display control unit 126 and instructs the display control unit 126. Further, the character input display setting unit 125 presents the candidate character extracted by the candidate control unit 124 to the user, and sets the candidate character (first candidate character c2) selected by the user as the confirmed character c5 corresponding to the handwritten character c1. Identify.

  The display control unit 126 receives an instruction for the content to be displayed on the display unit 130 and performs display on the display unit 130.

  The character editing display setting unit 127 acquires the designation of a character (a character designated by the user as a character erroneously recognized from the first candidate character c2) from the character designation unit 112, and the second and subsequent candidates corresponding to the designated character Next candidate character c3 is acquired and the display control unit 126 is instructed to display it. Here, in particular, the character editing display setting unit 127 narrows down the next candidate character c3 from a plurality of recognized characters held in the recognized character memory 123 based on the hint information c4 acquired by the hint information recognition unit 128A. To the candidate control unit 124.

  The hint information recognition unit 128A recognizes the hint information c4 input by handwriting. Specifically, the hint information recognition unit 128A acquires the hint information c4 input from the hint information input unit 113, recognizes it by the character recognition unit 122, and narrows down the recognized characters recognized from the handwritten character c1. Is sent to the character editing display setting unit 127. The hint information recognition unit 128A may determine the type of the hint information c4 when recognizing the hint information c4.

  The display unit 130 displays information handled by the character input device 100A. When the input unit 110A is realized by a touch panel and the display unit 130 is realized by a liquid crystal display, both can be configured as one body.

  The storage unit 140 stores data necessary for the processing of the control unit 120A. For example, a dictionary DB 141 to be referred to when recognizing the handwritten character c1 is stored.

(Details of character input)
The flow of handwriting input processing in the character input device 100A will be described with reference to FIGS.

  FIG. 3 is a flowchart showing an example of the flow of handwriting input processing in the character input device 100A. FIG. 4 is a transition diagram of a screen displayed during the handwriting input process in the character input device 100A. FIG. 4A shows a screen example in which the handwritten character c1 is recognized and the first candidate character c2 is displayed. ) Shows a screen example in which the next candidate character c3 is displayed for the erroneously recognized character designated from the first candidate character c2, and (c) shows a screen in which the next candidate character c3 is updated based on the input hint information c4. An example is shown, and (d) shows an example of a screen when a recognized character is confirmed.

  First, the handwritten character input unit 111 acquires the handwritten character c1 from the handwritten input area A111 (S11).

  Next, the handwritten character c1 input by the input character recognition unit 121 is acquired, and character recognition is performed by the character recognition unit 122 (S12; character recognition step). The input character recognition unit 121 stores the recognized character (character code) obtained as a result of this recognition in the recognized character memory 123 in association with the recognition accuracy for each character of the handwritten character c1.

  Next, the character input display setting unit 125 acquires the first candidate character c2 that is the first candidate for each character of the handwritten character c1 from the recognized character memory 123 via the candidate control unit 124, and the display control unit 126. Send to. And the display control part 126 displays the 1st candidate character c2 on the display part 130 with the handwritten character c1 sent from the input character recognition part 121 (S13, Fig.4 (a)).

  In the example of FIG. 4A, handwritten input “go to Kyoto” and the characters “K”, “part”, “ni”, “line”, “ku”, which are the first candidate characters c2 of the recognition result, The handwritten characters c1 are displayed in association with each other.

  Next, the character edit display setting unit 127 determines whether or not there is a character selected by the user via the character designating unit 112 as an erroneously recognized character from the displayed first candidate character c2 (S14). If an erroneously recognized character is designated (YES in S14), the character editing / display setting unit 127 selects the next candidate character c3 corresponding to the designated erroneously recognized character (first candidate character c2) as a candidate control unit 124. Via the recognized character memory 123. At this time, the candidate control unit 124 associates the designated first candidate character c2 with the recognized object (for one character included in the handwritten character c1) in response to the request from the character editing display setting unit 127. A predetermined number is extracted from the recognized characters in the order of recognition accuracy and returned as the next candidate character c3. Then, the character editing display setting unit 127 causes the display unit 130 to display the next candidate character c3 in association with the designated erroneously recognized character (first candidate character c2) via the display control unit 126 (S15, FIG. 4 (b)).

  In the example of FIG. 4A, the “part” of the first candidate character c2 recognized from the “city” of the handwritten character c1 is a misrecognized character. In the example of FIG. 4B, when the “part” is designated by a tap or the like, the “part” of the designated first candidate character c2 is recognized from the “city” of the handwritten character c1 recognized. The “evil” and “rejected” are displayed as the next candidate character c3 under “part” which is the first candidate character c2. In FIG. 4B, two next candidate characters c3 are displayed, but the number to be displayed is not limited.

  Next, the character editing / display setting unit 127 determines whether there is a candidate character designated by the user via the candidate selection unit 114 as a correct recognized character from the displayed next candidate character c3 (S16). If no candidate character is designated (NO in S16), the hint information input unit 113 acquires the hint information c4 from the hint information input area A113. Then, the hint information recognition unit 128A sends the result obtained by character recognition of the hint information c4 by the character recognition unit 22 to the character edit display setting unit 127 (S17; hint information recognition step).

  Next, the display edit display setting unit 127 sets conditions for extracting candidate characters from the recognized characters based on the result obtained by character recognition, and instructs the candidate control unit 124 to extract candidate characters. In response to this instruction, the candidate control unit 124 is associated with the object (one character included in the handwritten character c1) in which the first candidate character c2 designated as the erroneously recognized character is recognized according to the set condition. A predetermined number of candidate characters are extracted from the recognized characters in the order of recognition accuracy. The candidate character with the highest recognition accuracy is set as the first candidate character c2, and then a predetermined number of candidate characters with the highest recognition accuracy are returned as the next candidate character c3. Thereafter, the character editing / display setting unit 127 corresponds to the first candidate character c2 and the first candidate character c2 designated as the erroneous recognition characters by using the first candidate character c2 and the next candidate character c3 acquired from the candidate control unit 124. Then, the display screen is updated via the display control unit 126 so as to display instead of the next candidate character c3 that has been displayed (S18; candidate character extraction step).

  In the example of FIG. 4C, as the hint information c4, the reading kana of the “city” of the handwritten character c1 is input as “t” by phonetic reading. This indicates “G” of “Kyoto”, but “Miyako” may be entered as a knot. In addition, whether reading aloud or reading a knot may be input by distinguishing katakana and hiragana, or may not be distinguished.

  Then, from the recognized characters recognized from the “city” of the handwritten character c1, a recognized character whose sound reading is “g” is extracted, the “city” having the highest recognition accuracy is set as the first candidate character c2, and the recognition accuracy is next. High “betting” and “堵” are displayed as the next candidate character c3. Thus, when the candidate character is updated based on the hint information c4, both the first candidate character c2 and the next candidate character c3 are updated.

  Next, the character editing display setting unit 127 determines whether there is a candidate character designated by the user via the candidate selection unit 114 as a correct recognized character from the updated next candidate character c3 (S19). In step S19, when a candidate character is designated by the user from the displayed next candidate character c3 (YES in S19), the character editing / display setting unit 127 sets the designated next candidate character c3 as a new first candidate character c2. Then, the display screen is updated via the display control unit 126 so as to display instead of the old first candidate character c2 (S20). Similarly, when a candidate character is designated by the user from the displayed next candidate character c3 in step S16 (YES in S16), the designated next candidate character c3 is set as a new first candidate character c2, and the display screen is displayed. Update (S20).

  If the display screen is updated in step S20 and if no candidate character is specified by the user from the displayed next candidate character c3 in step S19 (NO in S19), an error is detected by the user from the first candidate character c2. The process returns to the process for acquiring the designation of the recognized character (S14). As described above, when the candidate character is updated based on the hint information c4, the first candidate character c2 is updated. Therefore, when the first candidate character c2 is a correct recognition result, the user selects the candidate character. No operation is required.

  Next, returning to step S14, the character editing display setting unit 127 determines whether there is a character selected by the user as a misrecognized character from the first candidate character c2, and when the misrecognized character is designated ( In S14, YES), the above-described processing (S15 to S20) is repeated. That is, the user can change another first candidate character c2 (for example, “K” and “Line” in FIG. 4A) by repeating the above-described processing (S15 to S20).

  If no erroneously recognized character is designated in step S14 (NO in S14), that is, when the overall confirmation unit 115 detects an operation of the character recognition confirmation button A115 by the user, the character input display setting unit 125 The recognition result of the handwritten character c1 is determined by all the first candidate characters c2 prepared so far (S21; character determination step). Then, the character input / display setting unit 125 displays the determined fixed character c5 in a predetermined character font instead of the handwritten character c1, the first candidate character c2, and the next candidate character c3 (FIG. 4D).

  In addition, after displaying the 1st candidate character c2 in step S13, the said 1st candidate character c2 is not selected as a misrecognized character, but when the predetermined time passes or the 1st candidate character c2 of a predetermined number of characters is added and displayed. At this point, the character editing display setting unit 127 determines that no erroneously recognized character has been designated (NO in S14), and the character input display setting unit 125 may determine the recognition result using the first candidate character c2. Good. Similarly, after the next candidate character c3 is displayed in steps S15 and S18, the character editing display setting is set when the predetermined time has elapsed or the handwritten character c1 is input without the next candidate character c3 being selected. The unit 127 may determine that the next candidate character c3 has not been selected (NO in S16 or S19), and the character input display setting unit 125 may determine the recognition result based on the first candidate character c2 at that time. Note that the predetermined time for waiting and the predetermined number of characters displayed for the first candidate character c2 may be determined from the operation history of the user.

(Summary)
As described above, according to the present configuration example, when editing a recognition result that has been misrecognized, the recognition characters are narrowed down only by tapping or handwriting input. Thereby, it is possible to increase the possibility that a correct recognized character can be presented by a laborious input method. Further, by using hint information for extraction of candidate characters, even if the recognition character has low recognition accuracy indicating the accuracy of recognition because the handwritten character is input in a shape that makes character recognition difficult, the hint information It is possible to narrow down based on and present to the user preferentially.

[Configuration example 2]
Next, the character input device 100B according to Configuration Example 2 will be described in detail with reference to FIGS. Note that, in the following description, “B” is added to members unique to this configuration example as “input unit 110B”. In addition, about the matter which is common with the member of another structural example, "B" may be abbreviate | omitted and may be named generically like the "input part 110." For convenience of explanation, members having the same functions as those shown in the configuration example 1 are given the same reference numerals, and explanation thereof is omitted.

(Configuration of character input device)
First, the configuration of the character input device 100B will be described in detail with reference to FIG. FIG. 5 is a functional block diagram showing a main configuration of the character input device 100B. Note that portions not directly related to this configuration example (for example, a portion that realizes dragging of input characters, etc.) are omitted from the description of the configuration and the block diagram from the viewpoint of ensuring the simplicity of description. However, the character input device 100B may include the omitted configuration in accordance with the actual situation.

  As illustrated in FIG. 5, the character input device 100B includes an input unit 110B, a control unit 120B, a display unit 130, and a storage unit 140. Note that the display unit 130 and the storage unit 140 have the same functions as the display unit 130 and the storage unit 140 of Configuration Example 1, respectively, and thus description thereof is omitted.

  The input unit 110B performs input for the user to operate the character input device 100B. Specifically, the input unit 110B includes a hint information type selection unit 116 in addition to the handwritten character input unit 111, the character designation unit 112, the hint information input unit 113, the candidate selection unit 114, and the overall determination unit 115. It should be noted that the display of the area and characters corresponding to each functional block of the input unit 110B on the display unit 130 is performed by the functional block or a functional block (not shown) corresponding to the functional block.

  The hint information type selection unit 116 displays hint information type information c6 for designating the type of hint information c4 input from the hint information type selection area A116 (FIG. 6) provided in the vicinity of the hint information input area A113 on the screen. get.

  In addition to the hint information type selection unit 116, the handwritten character input unit 111, the character designation unit 112, the hint information input unit 113, the candidate selection unit 114, and the overall determination unit 115 each have the same functions as those in the configuration example 1. Therefore, the description is omitted.

  The controller 120B executes various controls in the character input device 100B. Therefore, the control unit 120B includes an input character recognition unit 121, a character recognition unit 122, a recognized character memory 123, a candidate control unit 124, a character input display setting unit 125, a display control unit 126, a character edit display setting unit 127, and hint information. A recognition unit (hint information recognition means) 128B is provided. That is, the control unit 120B is configured to include a hint information recognition unit 128B instead of the hint information recognition unit 128A, as compared to the control unit 120A (FIG. 1) of the configuration example 1.

  The hint information recognition unit 128B recognizes the hint information c4 input by handwriting. In particular, the hint information recognition unit 128B acquires the hint information type information c6 indicating the type of the hint information c4 together with the hint information c4, and recognizes the hint information c4 based on the hint information type information c6. Specifically, the hint information recognition unit 128B acquires the hint information c4 input from the hint information input unit 113 and the hint information type information c6 input from the hint information type selection unit 116, and the character recognition unit 122 Recognition is performed, information for narrowing down the recognized characters recognized from the handwritten character c1 is generated, and the information is sent to the character editing display setting unit 127. The hint information recognition unit 128B uses the hint information type information c6 when recognizing the hint information c4. In the configuration example 1, the type of the hint information c4 is determined by the hint information recognition unit 128A.

  In addition to the hint information recognition unit 128B, the input character recognition unit 121, the character recognition unit 122, the recognized character memory 123, the candidate control unit 124, the character input display setting unit 125, the display control unit 126, and the character edit display setting unit 127 are Since each has the same function as that of the configuration example 1, the description thereof is omitted.

(Details of character input)
The flow of handwriting input processing in the character input device 100B will be described with reference to FIGS. FIG. 6 is a screen example displayed during handwriting input processing in the character input device 100B, and shows a screen example in which the next candidate character c3 is updated based on the input hint information c4 and hint information type information c6.

  Since only the process in Step S17 of FIG. 3 is different from the process of Configuration Example 1, the description of the other steps is omitted.

  As shown in FIG. 3, when a candidate character is not designated as a correct recognized character from the displayed next candidate character c3 in step S16 (NO in S16), the hint information input unit 113 starts from the hint information input area A113. The hint information c4 is acquired. At this time, in the character input device 100B, the hint information type selection unit 116 acquires the hint information type information c6 from the hint information type selection area A116 displayed on the screen together with the hint information input area A113 (FIG. 6). Then, the hint information recognition unit 128B considers the hint information c4 input from the hint information input unit 113 by the character recognition unit 122 and the type indicated by the hint information type information c6 input from the hint information type selection unit 116. Are recognized and sent to the character editing display setting unit 127 (S17; hint information recognition step). Then, the display editing display setting unit 127 sets conditions for extracting candidate characters from the recognized characters based on the result obtained by character recognition using the hint information type information c6, and the candidate control unit 124 is set as a candidate. Instructs extraction of characters. The subsequent processing is the same as that of the configuration example 1.

  In the example of FIG. 6, after selecting “radical” as the hint information type information c6, by inputting only the “Ozama” portion as the hint information c4, the candidate characters are narrowed down and “Miyako” is extracted. The probability that

  Here, in FIG. 6, as the hint information type information c6, “Kanji reading (kana)”, “Kanji radical”, “character type”, and “character size” are listed. However, the hint information type information c6 is not limited to these, and any information can be used as long as it can be specified as the type of the hint information c4.

  Examples of the “character type” include “kana-kanji”, “alphabet”, “number”, “emoticon”, and “pictogram”. However, other characters that can be specified as character types, such as “Hiragana”, “Katakana”, “Kanji”, “Numbers”, “Symbols”, and characters used in other countries such as “Latin characters” and “Russian characters” Any can be used.

  In the case of “character type”, “character type” is initially displayed in the hint information type selection area A116, and when “character type” is selected, “kana-kanji”, “alphabet”, “number”, “emoticon”, “pictogram” Or the like may be displayed to allow the user to select a character type. Alternatively, a character type may be selected in advance from “kana / kanji”, “alphabet”, “number”, “emoticon”, “pictogram”, and the like, and the selected character type may be displayed in the hint information type selection area A116 from the beginning. The “emoticon” is a pseudo image expressed by a combination of characters such as “(^^)”, and the pictogram is an image (including an animated one) associated with a character code.

  For example, when “Kana-Kanji” is selected as the “character type” before handwriting input, when “shi” is input by handwriting, the alphabet “l” (L small letter) and the number “1” are also candidates. You can prevent it from going up. Conversely, when “character type” is selected after handwritten input, characters of all character types may be displayed as candidate characters before selection, and candidate characters may be narrowed down by the selected character type after selection.

  In the case of “Large / Small”, for the characters that make up a pair of large and small such as “Ya” and “N”, “Y” and “y”, a refinement that specifies whether it is large or small is performed before handwritten input or handwritten Can be executed after input.

  In the case of “large” and “small”, as in the case of “character type”, “large” is initially displayed in the hint information type selection area A116, and when “large” is selected, “large” “small” ”May be displayed to allow the user to select between large and small. Alternatively, one of “large” and “small” may be selected in advance, and the selected large / small distinction may be displayed in the hint information type selection area A116 from the beginning.

  Furthermore, “character type” and “large” and “small” do not conflict with other types, and can be selected together with “reading” and “radical”. Whether it is “reading” or “radical” may be selected by the user, or the character input device 100B may determine which is being input.

  Note that “character type” and “large / small” may be set before handwritten input from the handwritten character input unit 111 and used for narrowing down in initial character recognition.

(Summary)
As described above, according to this configuration example, the handwritten input hint information can be recognized based on the type of hint information indicated by the hint information type information. This increases the probability that the hint information can be recognized correctly.

[Configuration example 3]
Next, the character input device 100C according to Configuration Example 3 will be described in detail with reference to FIGS. Note that, in the following description, “B” is added to members unique to this configuration example as “input unit 110 </ b> C”. Note that items common to members of other configuration examples may be collectively referred to as “input unit 110” with “C” omitted. For convenience of explanation, members having the same functions as those shown in the structural examples 1 and 2 are denoted by the same reference numerals, and description thereof is omitted.

(Configuration of character input device)
First, the configuration of the character input device 100C will be described in detail with reference to FIG. FIG. 7 is a functional block diagram showing a main configuration of the character input device 100C. Note that portions not directly related to this configuration example (for example, a portion that realizes dragging of input characters, etc.) are omitted from the description of the configuration and the block diagram from the viewpoint of ensuring the simplicity of description. However, the character input device 100C may include the omitted configuration in accordance with the actual situation.

  As shown in FIG. 7, the character input device 100C includes an input unit 110C, a control unit 120C, a display unit 130, and a storage unit 140. Note that the display unit 130 and the storage unit 140 have the same functions as the display unit 130 and the storage unit 140 of Configuration Examples 1 and 2, respectively, and thus description thereof is omitted.

  The input unit 110C is used by the user to input for operating the character input device 100C. Specifically, the input unit 110C includes a handwritten character input unit 111, a character designation unit 112C, a hint information input unit 113, a hint information type selection unit 116, a candidate selection unit 114, and an overall determination unit 115. That is, as compared with the input unit 110 </ b> B (FIG. 5) of the configuration example 2, the character specifying unit 112 </ b> C is provided instead of the character specifying unit 112. It should be noted that the display of the area and characters corresponding to each functional block of the input unit 110C on the display unit 130 is performed by the functional block or a functional block (not shown) corresponding to the functional block.

  The character designating unit 112 </ b> C misrecognizes the first candidate character c <b> 2 that is the first candidate for the recognition result of the handwritten character c <b> 1 or the morphological analysis result (candidate character) c <b> 7 obtained by performing the morphological analysis on the first candidate character c < In such a case, when the user taps the display of the character or the character string, the user is made to select the target for re-recognition. In addition, when the morphological analysis result c7 is misrecognized, the character designating unit 112C displays the part-of-speech sub-category selection area A117 having contents corresponding to the part-of-speech of the character string of the selected morpheme analysis result c7. Get selection.

  In addition to the character designation unit 112C, the handwritten character input unit 111, the hint information input unit 113, the hint information type selection unit 116, the candidate selection unit 114, and the overall determination unit 115 have the same functions as those of the configuration examples 1 and 2, respectively. The description thereof is omitted.

  The controller 120C executes various controls in the character input device 100C. Therefore, the control unit 120C includes an input character recognition unit (character recognition unit) 121C, a character recognition unit 122, a recognized character memory 123C, a candidate control unit 124, a character input display setting unit 125, a display control unit 126, a character editing display setting unit. 127C, a hint information recognition unit 128B, and a part of speech recognition unit 129. That is, as compared with the control unit 120B (FIG. 5) of the configuration example 2, the control unit 120C replaces the input character recognition unit 121, the recognized character memory 123, and the character editing display setting unit 127 with the input character recognition unit 121C, The configuration includes a recognized character memory 123 </ b> C and a character editing / display setting unit 127 </ b> C, and further includes a part-of-speech recognition unit 129.

  Similarly to the input character recognition unit 121 of the configuration example 1, the input character recognition unit 121C acquires the handwritten character c1 input from the handwritten character input unit 111 and recognizes the character code as a result of recognition by the character recognition unit 122. Store in the character memory 123C. Further, the input character recognition unit 121 displays the handwritten character c <b> 1 as an object on the display unit 130 via the display control unit 126.

  The part-of-speech recognition unit 129 returns a character code of a morphological analysis result c7 (character string) obtained by performing morphological analysis on the first candidate character c2 which is a first candidate of recognition result of the handwritten character c1. Note that data such as a dictionary to be referred to when performing morphological analysis may be stored in the dictionary DB 141 or the like in the storage unit 140.

  The input character recognizing unit 121C particularly recognizes the character code of the character string obtained as a result of the morphological analysis of the first candidate character c2 which is the first candidate of the recognition result of the handwritten character c1 by the part-of-speech recognition unit 129. Store in 123C.

  The recognized character memory 123C holds one or more recognized characters (character codes) that are the recognition results of each character or character string in character units cut out from the handwritten character c1 and character string units obtained by morphological analysis. .

  Here, the character string obtained by handwriting recognition can be decomposed into character strings such as nouns, verbs, adjectives, particles and auxiliary verbs by performing morphological analysis on the character string of the recognition result. For example, in the example shown in FIG. 8, since “ni” is determined to be a particle, when “ni” is tapped in the morphological analysis result c7, other particles such as “to”, “ha”, and “to” are displayed. Is displayed as a candidate character string. At this time, for example, kanji nouns with the same reading may be displayed as candidate character strings with lower priority, or character strings other than particles may not be displayed. That is, the candidate character string is displayed with priority given to the part of speech determined by the morphological analysis.

  However, morphological analysis is especially wrong when you enter it from the middle of a sentence or when you write a part of a noun that can be written in kanji in hiragana (such as “Kyoto” is “Kyoto”). There is. Therefore, it is possible to specify such as changing the part of speech from “others” (FIG. 8B) in the subcategory sub-category selection area A117 or correcting the mistake in the break of the phrase by attaching it to the preceding or following character. May be. In addition, if you intend to write “knowledge” but “knowledge” is misrecognized as “weave”, both of them are correct to the same extent if there is only one character. ”, It can be seen that the accuracy of“ knowledge ”is high. Thereby, the priority of a candidate character string can be changed appropriately.

  Similarly to the character editing display setting unit 127, the character editing display setting unit 127C acquires the designation of a character (a character that the user has designated as a character erroneously recognized from the first candidate character c2) from the character designation unit 112C. The next candidate character c3 after the second candidate corresponding to the designated character is acquired, and the display control unit 126 is instructed to display it. In addition, the character editing display setting unit 127C, based on the hint information c4 acquired by the hint information recognition unit 128B, narrows down to the next candidate character c3 from a plurality of recognized characters held in the recognized character memory 123C. The control unit 124 is instructed. Then, the candidate control unit 124 extracts candidate characters (first candidate character c2 and next candidate character c3) from the recognized characters generated by the input character recognition unit 121 based on the hint information c4 recognized by the hint information recognition unit 128. To do.

  Further, when the character editing display setting unit 127C obtains the specification of the character string (the character string specified by the user as the character string erroneously recognized from the morphological analysis result c7) from the character specifying unit 112C, the character editing display setting unit 127C The display control unit 126 is instructed to display the classification selection area A117, and the user is allowed to select a fine classification according to the part of speech of the designated character string. In addition, the character editing display setting unit 127C generates candidate character strings (first candidate, next candidate characters subsequent to the second candidate) corresponding to the designated character string generated so as to match the selected subcategory. Column (candidate character) c8) is acquired, and its display is instructed to the display control unit 126. In addition, the character editing display setting unit 127C converts a plurality of recognized character strings generated so as to match the selected subcategory based on the hint information c4 acquired by the hint information recognition unit 128B to candidate character strings. The candidate control unit 124 is instructed to narrow down. The candidate control unit 124 then selects a candidate character string (morpheme analysis result c7, next candidate character string c8) from the recognized character string generated by the input character recognition unit 121 based on the hint information c4 recognized by the hint information recognition unit 128. To extract.

  In this way, the character editing display setting unit 127C can present a set of a plurality of candidate characters (next candidate character string c8) corresponding to a plurality of handwritten characters c1 that are continuously input. Similarly, the character editing display setting unit 125 can present a set of a plurality of candidate characters (morpheme analysis results c7) corresponding to a plurality of handwritten characters c1 that are continuously input.

  In addition, the character recognition unit 122, candidate control unit 124, character input display setting unit 125, display control unit 126, hint other than the input character recognition unit 121C, the recognized character memory 123C, the character editing display setting unit 127C, and the part of speech recognition unit 129. Since the information recognition unit 128B has the same function as that of the configuration examples 1 and 2, the description thereof is omitted.

(Details of character input)
The flow of handwriting input processing in the character input device 100C will be described with reference to FIGS. FIG. 8 is a transition diagram of a screen displayed during the handwriting input process in the character input device 100C. FIG. 8A is a screen example in which the handwritten character c1 is recognized and the first candidate character c2 and the morphological analysis result c7 are displayed. (B) shows an example of a screen displaying the part-of-speech sub-category selection area A117 when a misrecognized character (column) is designated from the morphological analysis result c7. (C) is designated from the morphological analysis result c7. A screen example in which the next candidate character string c8 is displayed with respect to the erroneously recognized character (string) is shown.

  Note that, among the steps in FIG. 3, the description of the processing steps having the same contents as those in Configuration Example 1 is omitted.

  In the character input device 100C, when the handwritten character c1 is acquired, the input character recognizing unit 121C acquires the input handwritten character c1, and the character recognizing unit 122 performs character recognition (character recognizing step) and the part of speech recognizing unit 129. The morpheme analysis is performed by (morpheme analysis step) (S12). The input character recognition unit 121C stores the recognized character (character code) obtained as a result of character recognition in the recognized character memory 123 in association with the recognition accuracy for each character of the handwritten character c1, and obtains the result of morphological analysis. The recognized character string (character code) (morpheme analysis result c7) is stored in the recognized character memory 123 in association with the character of the handwritten character c1.

  Next, the character input display setting unit 125 acquires the first candidate character c2 that is the first candidate for each character of the handwritten character c1 from the recognized character memory 123 via the candidate control unit 124, and the display control unit 126. Send to. At the same time, the character input display setting unit 125 acquires the morphological analysis result c7 corresponding to the character of the handwritten character c1 from the recognized character memory 123 via the candidate control unit 124, and sends it to the display control unit 126. The display control unit 126 displays the first candidate character c2 and the morpheme analysis result c7 on the display unit 130 together with the handwritten character c1 sent from the input character recognition unit 121 (S13, FIG. 8A).

  In the example of FIG. 8A, “go to Kyoto” is input by handwriting, and the characters “K”, “part”, “ni”, “line”, and “ku”, which are the first candidate characters c2 of the recognition result, The handwritten characters c1 are displayed in association with each other. At the same time, the character strings “Kyobe”, “Ni”, and “Go”, which are morphological analysis results c7, are displayed in association with the handwritten character c1.

  As shown in FIG. 8A, the morphological analysis result c7 is displayed as a character string grouped by nouns, verbs, adjectives, particles, auxiliary verbs, adverbs, etc., as a result of the morphological analysis performed by the part of speech recognition unit 129. Is done. The morphological analysis result c7 may be presented so that the user can determine the part of speech by changing the color according to the part of speech or displaying the part of speech name horizontally.

  Next, the character editing / display setting unit 127 acquires a selection of a misrecognized character from the displayed first candidate character c2 via the character specifying unit 112 (S14). At the same time, the character editing / display setting unit 127 acquires the selection of the erroneously recognized character string from the displayed morpheme analysis result c7 via the character designating unit 112 (S14). If the selection of the misrecognized character string from the morpheme analysis result c7 is acquired (YES in S14), it is classified by part of speech for causing the user to select a fine classification according to the part of speech of the character string of the selected morpheme analysis result c7. The fine classification selection area A117 is displayed (FIG. 8B). In FIG. 8B, the part-of-speech classification selection area A117 is displayed below the hint information input area A113 (position of the hint information type selection area A116). For example, the selected morphological analysis result c7 It may be displayed close to the character string.

  In the example of FIG. 8B, since “Kyobe”, which is the morphological analysis result c7, is tapped by the user, “Kyobe” is determined to be a noun. (First name and last name), “Person name (last name only)”, “Person name (first name only)”, “Emoji” and “Others” are displayed. In addition, as long as the items displayed as the fine classification can be specified as the fine classification of each part of speech, the contents and the number can be arbitrarily selected.

  Next, when the character editing / display setting unit 127 acquires the selection of the fine classification from the fine classification selection area A117 for each part of speech via the character designation unit 112 by the user, the designated erroneously recognized character string (morpheme analysis result c7) Is obtained from the recognized character memory 123 via the candidate control unit 124. At this time, in response to a request from the character edit display setting unit 127, the candidate control unit 124 applies an object (one character included in the handwritten character c1) in which the recognized character included in the designated morphological analysis result c7 is recognized. A predetermined number is extracted and returned in the order of recognition accuracy from a character string obtained by combining the associated recognized characters so as to match the selected subcategory. Then, the character editing display setting unit 127 replaces the extracted first candidate character string with the next candidate character instead of the misrecognized character string (morpheme analysis result c7) and the extracted second and subsequent character strings. The column c8 is associated with the erroneously recognized character string and is displayed on the display unit 130 via the display control unit 126 (S15, FIG. 8C).

  At this time, in the character input device 100C, the hint information input area A113 and the hint information type selection area A116 are displayed on the screen, and the hint information c4 and the hint information type information c6 for narrowing down candidate character strings can be input. .

  In the example of FIG. 8C, the user designates “Kyobe” as the morphological analysis result c7 as the misrecognized character string, and further selects “place name” as the subcategory of parts of speech. “Kyoto” is displayed, and “Toto”, “Toho”, and “Tokyo” are displayed as the next candidate character string c8.

  Here, as a candidate character string, not only a combination of recognized characters as a result of character recognition (in the example of FIG. 8C, “Kyoto” “Toto” “Toho”) but also “Tokyo” A character string obtained by converting the recognized character (string) based on a predetermined dictionary (predictive dictionary) may be presented together. As such conversion, conversion from Japanese to English, such as Kana-Kanji conversion or Japanese-English translation, can be used. Further, when the recognized character string is confirmed, information (subclass name etc.) for specifying the subcategory of the part of speech selected by the user in the determination process is added, and the recognized character string is added to the user registration dictionary (dictionary DB 141). You may register. For example, “person name (last name)” is added as additional information to the recognition character string selected after “person name (last name)” is selected as a subcategory of part of speech, and is registered in the dictionary. Further, when registering the recognized character string in the dictionary, it may be learned to increase the priority so that it can be preferentially presented in future recognition. For example, the recognition accuracy may be corrected.

  Next, as shown in FIG. 3, when a candidate character string is not designated as a correct recognized character from the displayed next candidate character string c8 in step S16 (NO in S16), the hint information input unit 113 Hint information type information c6 is obtained from the hint information type selection area A116, and hint information c4 is obtained from the hint information input area A113. Then, the display edit display setting unit 127 sets a condition for extracting a candidate character string based on the result obtained by character recognition using the hint information type information c6, and the candidate control unit 124 is set as a candidate character string. Is extracted (S17; hint information recognition step). The subsequent processing is the same as that of the configuration example 1.

  The hint information c4 for narrowing down the candidate character string of the morphological analysis result c7 is basically the same as the hint information c4 for narrowing down one candidate character. However, in the case of a candidate character string, since there are a plurality of characters, the following may be performed. (1) The designation of “radical” may be entered in order from the first character of the candidate character string. In this case, the display form is changed, for example, by changing the display color by associating the character in the morphological analysis result c7 with the input to the hint information input area A113 so that it can be understood which character radical is being input. Is preferred. Of course, such a display may be performed in the same way even in the case of one character. (2) For characters that are recognized correctly in the candidate character string, a button or the like that can skip the input of the hint information c4 is displayed in the corresponding character column of the hint information input area A113, or a skip operation is assigned. You may prepare gestures.

(Summary)
As described above, according to the present configuration example, when an input handwritten character is recognized, character recognition and morphological analysis are performed in units of characters. As a result, the context and the meaning of the word can be reflected and recognized.

[Modification]
As described above, the character input device 100 is a device that inputs characters by recognizing the handwritten character c1 drawn on the touch panel. However, the input method of the handwritten character c1 is not limited to the touch panel. For example, the handwritten character input unit 111 may read image data of the handwritten character c1 stored in advance from the storage medium.

  The image data of the handwritten character c1 may be, for example, scanned and read paper on which the character is handwritten. In the case of image data, data such as the stroke order and input speed cannot be acquired. Therefore, since the amount of information that can be acquired in real time from the touch panel is large, there is a high possibility that a highly accurate character recognition result can be presented.

  In addition, when the handwritten character c1 is input from the data file, some characters may have been converted into character codes (FIG. 9A). In this case, the character input device 100 can convert only the portion that has not been converted into the character code as the handwritten character c1 and convert it into the character code (first candidate character c2) by the above-described processing. Also, with respect to the part c11 that has been converted into the character code, if the image data of the handwritten character can be used, the part that has been converted into the character code is returned to the image data, and the entire part including the part that has not been converted into the character code Can be converted into a character code by the above-described processing. Thereby, for example, when it is originally two characters but is mistakenly converted into a character code as one character, or vice versa, conversion into a character code can be performed again. In that case, it is preferable that the range of one character can be designated in a handwritten state.

  In addition, when the image data includes type characters and handwritten characters, for example, in the case of scanned image data in which characters are handwritten on a paper on which type characters are written, the character input device 100 uses the type characters c12. Can be converted into a character code c2 ′ using an OCR (optical character reader) technique, and the handwritten character c1 can be converted into a character code (first candidate character c2) by the above-described processing (FIG. 9B). ).

  Furthermore, not only re-recognition of the read image data and handwritten characters, but also handwritten handwritten characters may be recognized, and the recognition result may be replaced with the read image data and handwritten characters.

[Combination of configurations (technical means) included in the embodiment]
Note that the configurations included in the above-described embodiments can be combined as appropriate. That is, the present invention is not limited to the above-described embodiment, and various modifications can be made within the scope shown in the claims. That is, embodiments obtained by combining technical means appropriately modified within the scope of the claims are also included in the technical scope of the present invention.

[Example of software implementation]
Finally, each block of the character input device 100, particularly each block of the character input unit 110 and the control unit 120, may be configured by hardware logic, or may be realized by software using a CPU as follows. Good.

  In the latter case, the character input device 100 stores a CPU that executes instructions of a program that realizes each function, a ROM (Read Only Memory) that stores the program, a RAM that expands the program, the program, and various data. A storage device (recording medium) such as a memory is provided. An object of the present invention is a recording medium on which a program code (execution format program, intermediate code program, source program) of a control program of the character input device 100, which is software that realizes the functions described above, is recorded so as to be readable by a computer. This can also be achieved by supplying to the character input device 100 and reading and executing the program code recorded on the recording medium by the computer (or CPU or MPU).

  Examples of the recording medium include tapes such as magnetic tapes and cassette tapes, magnetic disks such as floppy (registered trademark) disks / hard disks, and disks including optical disks such as CD-ROM / MO / MD / DVD / CD-R. IC cards (including memory cards) / optical cards, semiconductor memories such as mask ROM / EPROM / EEPROM / flash ROM, PLD (Programmable logic device), FPGA (Field Programmable Gate Array), etc. Logic circuits can be used.

  Further, the character input device 100 may be configured to be connectable to a communication network, and the program code may be supplied via the communication network. The communication network is not particularly limited as long as it can transmit the program code. For example, the Internet, intranet, extranet, LAN, ISDN, VAN, CATV communication network, virtual private network, telephone line network, mobile communication network, satellite communication network, and the like can be used. The transmission medium constituting the communication network may be any medium that can transmit the program code, and is not limited to a specific configuration or type. For example, even with wired lines such as IEEE 1394, USB, power line carrier, cable TV line, telephone line, and ADSL (Asymmetric Digital Subscriber Line) line, infrared rays such as IrDA and remote control, Bluetooth (registered trademark), IEEE 802.11 wireless, HDR ( It can also be used by radio such as High Data Rate (NFC), Near Field Communication (NFC), Digital Living Network Alliance (DLNA), mobile phone network, satellite line, and digital terrestrial network. The present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission.

  As described above, in this specification, the means does not necessarily mean a physical means, and includes the case where the function of each means is realized by software. Further, the function of one means may be realized by two or more physical means, and the function of two or more means may be realized by one physical means.

  The present invention is widely applicable to devices that perform character input by recognizing handwritten characters, for example, smartphone terminals and tablet terminals equipped with a touch panel display.

100 (100A, 100B, 100C) Character input device 121 (121C) Input character recognition unit (character recognition means)
124 Candidate control unit (candidate character extraction means)
125 Character input display setting part (character confirmation means)
128 (128A, 128B) Hint information recognition unit (hint information recognition means)
c1 Handwritten character c2 First candidate character (candidate character)
c3 Next candidate character (candidate character)
c4 Hint information c5 Fixed character c6 Hint information type information c7 Morphological analysis result (candidate character)
c8 Next candidate character string (candidate character)
S12 Character recognition step S17 Hint information recognition step S18 Candidate character extraction step S21 Character confirmation step

Claims (7)

A character input device for inputting characters by recognizing handwritten characters,
Character recognition means for recognizing handwritten characters and generating a plurality of recognized characters;
Candidate character extraction means for extracting candidate characters from the recognized characters generated by the character recognition means;
Character confirmation means for presenting candidate characters extracted by the candidate character extraction means to the user, and specifying the candidate characters selected by the user as confirmed characters corresponding to the handwritten characters;
A hint information recognition means for recognizing handwritten hint information,
The character input device, wherein the candidate character extraction unit extracts the candidate character from the recognized character based on the hint information recognized by the hint information recognition unit.   The character input device according to claim 1, wherein the character determination unit presents a set of a plurality of candidate characters corresponding to a plurality of handwritten characters input continuously.   The hint information recognizing means acquires hint information type information indicating a type of the hint information together with the hint information, and recognizes the hint information based on the hint information type information. The character input device according to 2.   4. The character input device according to claim 3, wherein the hint information type information is at least one of kanji reading, kanji radical, character type, and character size. A method for controlling a character input device that performs character input by recognizing handwritten characters,
A character recognition step for recognizing handwritten characters and generating a plurality of recognized characters;
A hint information recognition step for recognizing handwritten hint information;
A candidate character extraction step for extracting candidate characters from the recognized characters generated in the character recognition step based on the hint information recognized in the hint information recognition step;
A character input device including: a candidate character extracted in the candidate character extraction step; and a character determination step for specifying the candidate character selected by the user as a fixed character corresponding to the handwritten character. Control method.   A control program for operating the character input device according to any one of claims 1 to 4, wherein the control program causes a computer to function as each means.   A computer-readable recording medium on which the control program according to claim 6 is recorded.

Images