JP5492316B2 - Character input device and method - Google Patents

Description

  The present invention relates to a character input device and method for determining a character input by a user via some input device.

  When inputting Japanese characters with an alphabet keyboard, input a Roman character string, which is a Japanese reading, using an alphabetic character through the keyboard, and convert it into a kanji mixed text via the kana character string. Character conversion techniques such as “Romaji-kana conversion” and “Kana-kanji conversion” are known (see, for example, Patent Document 1). In this case, the character string is converted as follows. That is, the conversion of the character string is performed in the order of “input → alphabet string → (roman character conversion) → kana → (kana / kanji conversion) → sentence mixed with kanji”.

  In addition, when inputting Chinese with an alphabet keyboard, Pinyin strings that are Chinese readings are input with alphabet characters through the keyboard, and the alphabet strings are converted into Chinese Kanji strings. A character conversion technique called “pinyin conversion” is known (for example, see Patent Document 2). In this case, the character string is converted as follows. That is, the character string is converted in the order of “input → alphabet string → (pinyin conversion) → kanji character string”.

  The above-described character conversion technique may also be used when handwritten character recognition (see, for example, Patent Document 3) for recognizing characters with respect to a handwriting input via a coordinate input device is used instead of a keyboard. is there.

JP-A-7-105202 Japanese Patent Laid-Open No. 11-238051 Japanese Patent Laid-Open No. 2002-203208

  Handwritten character recognition can cause misrecognition. In a keyboard, in particular, a virtual keyboard, tactile feedback when a key is pressed cannot be obtained, so it is difficult to determine which key is pressed, and a mistake in pressing the key is likely to occur. That is, both the handwritten character recognition and the virtual keyboard may erroneously output a character string different from the character string intended by the user.

Considering the case of inputting characters using Romaji conversion or Pinyin conversion, the combinations of input alphabets can be limited. For example, in the case of Romaji conversion:
A: “a”, I: “i” ...
K: “ka”, K: “ki” ...
Since Japanese kana can be expressed by a combination of a small number of alphabets, the character strings of alphabets that can be converted into romaji are limited.

  However, the character set that can be entered with character recognition or virtual keyboard is usually fixed, so the result of character recognition or virtual keyboard judgment due to misrecognition or incorrect input becomes a combination of alphabets that cannot be converted to kana by Romaji conversion. There are many. For example, consider a case where a handwriting “ka” is input via character recognition in order to input “ka”. Since the shapes of “a” and “d” are similar, it is easy to misrecognize “a” as “d” in character recognition. If “a” written by the user is misrecognized as “d” and determined to be “kd”, there may be a combination that is not possible with Roman input.

  In character recognition, not only the recognition result of the first place but also the recognition result of the second place or less is obtained. In this case, a method can be considered in which character string candidates are created by combining a plurality of recognition results to cope with erroneous recognition. Even if “a” is misrecognized as “d”, if “a” is included in the recognition results of the second and lower ranks, the character string intended by the user “ka” may be included in the candidate. It can be done. However, if recognition processing is performed on all characters without limiting the characters, there is no guarantee that the character string resulting from character recognition will always be a correct combination as Roman characters. The same thing occurs when inputting Chinese through Pinyin conversion. As described above, in the case of a character input device using character recognition or a soft keyboard, there is a possibility of erroneous input, and thus the user may not be able to accurately input a sentence that the user wants to input.

  The present invention has been made in consideration of the above-described circumstances, and provides a character input device and method for reducing erroneous recognition and erroneous input when a user inputs a sentence via a character conversion device. For the purpose.

  In order to solve the above-described problem, the character input device according to the present invention corresponds to the input character from the reception unit that receives character input one character at a time as the input character and the first character set including one or more characters. Selecting means for selecting one character to be selected as a selected character, recording means for recording one or more characters having meaning as a character string in a certain character conversion, and recording a character that matches the selected character from the recording means Extraction means for selecting as a character and extracting all characters immediately following each recorded character from the recording means as an inputable character set, and setting the first character set as the inputable character set for new input Setting means for receiving characters.

  In addition, the character input device of the present invention is configured to receive a first character corresponding to the input character from a receiving unit that receives character input as an input character one character at a time and a first character set including one or more characters. First selection means for selecting as a selected character, second selection means for selecting one character corresponding to the input character as a second selected character from a second character set including one or more characters of one character, and the first A first storage means for storing a selected character; a second storage means for storing the second selected character; and a first recording means for recording one or more characters having meaning as a character string in a certain character conversion; First, a character that matches the second selected character is selected as a recorded character from the first recording means, and one character that immediately follows for each recorded character is extracted as an inputable character set from the first recording means. Extraction means and First setting means for setting the second character set as the inputable character set and receiving a new input character, one or more characters, and one or more converted characters obtained by converting the characters into the character The second recording means that is recorded and the second recording means search for a character that matches the second selected character as a matching character, and all converted characters associated with the matching character are extracted as extracted characters. A second extracting unit; a determining unit for determining whether one character is selected or not selected from the first selected character and the extracted character; and the first selected character and the extracted character. If it is determined that one character is selected from the first storage means, the contents of the first storage means and the second storage means are set to be empty, a new input character is accepted, and the first selected character and the extracted character One letter from home If it is determined not to-option is characterized by comprising a second setting means for accepting an input character followed by the contents while maintaining the contents of the second storage means and the first storage means.

  According to the character input device and method of the present invention, it is possible to reduce erroneous recognition and erroneous input when a user inputs a sentence via a device for converting characters.

The block diagram of the character input device of a 1st embodiment. The figure which shows an example whose input acquisition part of FIG. 1 is a virtual keyboard. The figure which shows an example of the table | surface which the inputable character recording part of FIG. 1 has recorded. The figure which shows an example of the inputable character set according to the accumulation | storage character string which the inputable character extraction part of FIG. 1 extracts. The figure which shows the database which shows the key position of the virtual keyboard of FIG. The figure which shows the graph which shows the positional relationship of the keys of the virtual keyboard of FIG. The flowchart which shows an example of operation | movement of the character input device of 1st Embodiment. The figure which shows a specific example of FIG. The figure which shows a specific example of FIG. The block diagram of the character input device of 2nd Embodiment. The figure which shows the content which the conversion character string recording part of FIG. 10 has memorize | stored. The flowchart which shows an example of operation | movement of the character input device of 2nd Embodiment. The figure which shows a specific example of FIG. The figure which shows a specific example of FIG. The block diagram of the character input device of 3rd Embodiment. The flowchart which shows an example of operation | movement of the character input device of 3rd Embodiment.

Hereinafter, a character input device and method according to an embodiment of the present invention will be described in detail with reference to the drawings. Note that, in the following embodiments, the same numbered portions are assumed to perform the same operation, and repeated description is omitted.
(First embodiment)
A character input device according to a first embodiment will be described with reference to FIG.
The character input device of this embodiment includes an input acquisition unit 101, an inputable character recording unit 102, an inputable character extraction unit 103, a character determination unit 104, a character storage unit 105, and a character string output unit 106.

The input acquisition unit 101 acquires character input information input by the user, and passes the character input information to the character determination unit 104.
For example, when the input from the user is a handwriting representing a character, the input acquisition unit 101 displays handwriting information written by the user on a screen (not shown), and acquires the handwriting information. In this case, the handwriting information is expressed as a time series of coordinates composed of the following multiple images.
(X [1] [1], Y [1] [1]), (X [1] [2], Y [1] [2]), ..., (X [1] [K [1]], Y [[1] K [1]])
(X [2] [1], Y [2] [1]), (X [2] [2], Y [2] [2]), ..., (X [2] [K [2]], Y [[2] K [2]])
...
(X [N] [1], Y [N] [1]), (X [N] [2], Y [N] [2]), ..., (X [N] [K [N]], Y [[N] K [N]])
Where N is the number of strokes of one character, K [i] is the number of points in the handwriting of the i-th drawing, and (X [i] [k], Y [i] [k]) is the k-th number of the handwriting in the i-th drawing. The coordinates of the point. This kth point is sampled by the input acquisition unit 101 at regular intervals.

  The input acquisition unit 101 is, for example, a virtual keyboard as shown in FIG. 2 and is displayed on a screen such as a display-integrated touch panel, and accepts coordinate input from the user. In the case of input, the coordinates (X, Y) on the screen tapped by the user are passed to the character determination unit 104. The input acquisition unit 101 is a keyboard, for example. In the case of input, the character of the key on the keyboard pressed by the user is passed to the character determination unit 104. The input acquisition unit 101 receives coordinate information when tapping a virtual keyboard displayed on the screen as input characters.

  The inputable character recording unit 102 records one or more characters that have meaning as characters. For example, in the case of Romaji conversion, a table as shown in FIG. 3 is held. That is, the inputable character recording unit 102 records all of one or more Roman characters having a meaning as a kana. In the case of Pinyin conversion, in addition to the table as shown in FIG. 3, conversion characters corresponding to four voices (for example, numbers “1”, “2”, “3”, “4”, respectively) are assigned.

  The inputable character extraction unit 103 searches the inputable character recording unit 102 using the stored character string of the character storage unit 105 as a search key, and passes the inputable character set to the character determination unit 104. As a search process performed by the inputable character extraction unit 103, a character string that partially matches the search key from the beginning is searched, and characters following the matched character string are extracted to form an inputable character set.

  For example, when “k” is used as a search key, the inputable character extraction unit 103 performs a search process as in (i) (ii).

(I) “ka”, “ki”, “ku”, “ke,“ ko ”,“ kya ”,“ kyu ”,“ kyo ”partially match“ k ”from the inputable character recording section from the beginning. Do
(Ii) The character {a, i, u, e, o, y} following “k” is extracted from the character string of (i) to obtain an inputable character set.
When the search key is “” (NULL) (that is, no character string is stored in the character storage unit 105), the set of the first characters of all the character strings in the inputable character recording unit 102 is an inputable character set. Become. In addition, when a part from the beginning of the search key matches all characters of the input character recording unit 102, the inputable character extraction unit 103 deletes the all-matching part and uses the remaining character string as the search key. Perform a re-search. For example, if the search key is “syak”, the “sya” part is completely matched, so “sya” is deleted, and the remaining “k” is searched again as a search key. When the inputable character recording unit 102 stores the table shown in FIG. 3, the inputable character extraction unit 103 inputs corresponding to the stored character strings stored in the character storage unit 105 as shown in FIG. Get possible character set.

The character determination unit 104 selects a character that most closely matches the input information of the input acquisition unit from the input-capable character set (character set 1) output by the input-capable character extraction unit 103, and the character storage unit 105 selects the character. To pass.
For example, when the input acquisition unit 101 acquires handwriting information, the character determination unit 104 may perform the character recognition process disclosed in the above Japanese Patent Application Laid-Open No. 2002-203208. The most similar standard pattern is selected from the standard patterns corresponding to the input character set for the input handwriting, and the character corresponding to the standard pattern is passed to the character storage unit 105.

  In addition, for example, when the input acquisition unit 101 is a virtual keyboard, the character determination unit 104 performs determination based on the coordinate information (X, Y) from the input acquisition unit 101. The character determination unit 104 selects a key with reference to the coordinates corresponding to the coordinate information acquired by the input acquisition unit 101 and the position coordinates of each key of the virtual keyboard. If the virtual keyboard has the keyboard layout as shown in FIG. 2, it has the position information of each key as shown in FIG. A close key is selected, and a character corresponding to the key is passed to the character storage unit 105. FIG. 5 is a key position database describing the position of each key of the virtual keyboard used in the input acquisition unit. The character determination unit 104 determines a key corresponding to the coordinates input by the user based on the position information of this DB. The coordinates in FIG. 5 are determined so that the upper left of FIG. 2 is the origin, the x coordinate increases as going to the right, and the y coordinate increases as going down. The node represents each key on the keyboard, and the arc represents the positional relationship between the keys. A distance between keys may be described by assigning a distance to a node.

  Further, for example, when the input acquisition unit 101 is a keyboard, the character determination unit 104 refers to the key that the input acquisition unit 101 is pressed and the arrangement information of a plurality of keys arranged on the keyboard, and determines the character. select. More specifically, for example, when the input acquisition unit 101 is a keyboard and the output key from the input acquisition unit 101 is included in the inputable character set, the character determination unit 104 outputs the character corresponding to the output key as it is. To do. On the other hand, when the output key from the input acquisition unit 101 is not included in the inputable character set, the character determination unit 104 determines the character corresponding to the output key having the closest positional relationship among the keys included in the inputable character set. Is output. Here, the proximity of the keys may be obtained from a graph structure describing the positional relationship between the keys as shown in FIG. 6, for example. Each node represents a key, each arc represents a positional relationship between the keys, and a distance is assigned to each arc. In order to determine that there is a high possibility that the closest key is most selected, for example, a larger weight coefficient is assigned to a node from a certain key as the key is closer.

  Furthermore, the character determination unit 104 may prepare a mode for setting all characters recorded in the inputable character recording unit 102 and may be set so as to be switched to this mode.

  The character storage unit 105 stores the characters output by the character determination unit 104 as a character string arranged in time order.

  The character string output unit 106 outputs the character string stored in the character storage unit. When the user deletes the character string output by the character string output unit 106, the character determination unit 104 prepares a mode for setting all characters recorded in the inputable character recording unit 102. It may be set to be switched to.

  Next, an example of the operation of the character input device according to the first embodiment will be described with reference to FIGS. 7, 8, and 9. In the example shown here, the flow of processing when the character determination unit 104 performs character recognition processing using handwriting information and the inputable character recording unit 102 records a Roman alphabet conversion table as shown in FIG. 3 will be described. To do. As a specific example, a case where “ka” is written to input the character “ka” will be described.

(I) First, as shown in FIG. 8, the inputable character extraction unit 103 calculates an inputable character set that can be input to the first character before the first character is input (step S703). In this case, since no character is stored in the character storage unit 105, the inputable character extraction unit 103, as shown in FIG. 8, starts the head of all the character strings included in the inputable character recording unit 102 (FIG. 3). A set of characters,
{A, i, u, e, o, k, s, t, n, h, m, y, r, w, g, z, d, b, p, c} (*)
Is calculated as an inputable character set (step S703).

(Ii) Next, as shown in FIG. 8, when the input acquisition unit 101 inputs “k” as the first character, the character determination unit 104 sets the input character set (*) as the character set 1 and sets the character set. 1 is determined to be “k” similar to the input handwriting (step S701). The character storage unit 105 stores “k” (step S702). The character string output unit 106 outputs “k” (step S704). The inputable character extraction unit 103 searches the inputable character recording unit 102 (FIG. 3) using “k” as a search key, and as shown in FIG.
{A, i, u, e, o, y} (**)
Is calculated as the next inputable character set (step S703).

  (Iii) Subsequently, when “a” is input as the second character, as shown in FIG. 9, the character determination unit 104 sets the input character set (**) as the character set 1 and “A” similar to the input handwriting is selected from (step S701). The character storage unit 105 stores “ka” (step S702). The character string output unit 106 outputs “ka” (step S704). As shown in FIG. 9, the inputable character extraction unit 103 performs a search process on the inputable character recording unit 102 (FIG. 3) using “ka” as a search key. Here, the inputable character extraction unit 103 can input (*) as in the case where the character string is not stored in the character storage unit 105 because “ka” completely matches the character string in FIG. Calculate as character set.

  For the third and subsequent characters, the same process may be repeated each time a character is input.

  In the example shown here, if character determination is performed without limiting the character set for the second character, if an ambiguous handwriting is input whether the second character is “a” or “d”, “ Although the possibility of misrecognizing as “d” increases, the character determination unit 104 does not misrecognize “a” as “d” by limiting the character set.

  According to the first embodiment described above, the character determination of the next input character is performed by limiting to the character set that may be input next from the character combination table and the character string input so far. As a result, the possibility of erroneous recognition and erroneous input can be reduced.

(Second Embodiment)
A character input device according to a second embodiment will be described with reference to FIG.
In addition to the character input device of the first embodiment, the character input device of this embodiment includes a conversion character string recording unit 1001, a character string conversion unit 1002, a first character determination unit 1003, a first character storage unit 1004, and candidates. A character string output unit 1005 and an accumulated character string control unit 1006 are included. The second character determining unit 104 and the second character accumulating unit 105 are the same as the character determining unit 104 and the character accumulating unit 105 of the first embodiment, although the names are different.

  The conversion character string recording unit 1001 records a search character string for search and a conversion character string corresponding to the character string. The conversion character string recording unit 1001 records a Roman character kana conversion table as shown in FIG. 11, for example, in the case of Roman character kana conversion. FIG. 11 is an example of a database recorded in the converted character string recording unit, and shows the correspondence between Roman characters and converted kana.

  The character string conversion unit 1002 searches the converted character string recording unit 1001 using the character string output from the character string output unit 106 as a search key. The character string conversion unit 1002 refers to the conversion character string recording unit 1001 in order to search for a conversion character string. The Roman characters in FIG. 11 are searched, and a kana (converted character string) corresponding to the searched Roman characters is output. More specifically, the character string conversion unit 1002 first searches for a search character string that matches the search key, and extracts and outputs a conversion character string corresponding to the matched search character string. In addition, when there are a plurality of matching search character strings, the character string conversion unit 1002 outputs all corresponding conversion character strings.

  The first character determination unit 1003 is the same as the character determination unit 104 except that the character set is different. The character set 1 to be determined by the first character determination unit is a fixed character set. The character set 1 is, for example, all characters. Note that the character set 2 of the character determination unit 104 is the same as in the first embodiment and is an inputable character set.

  The first character storage unit 1004 stores the characters output by the first character determination unit 1003 as a character string arranged in time order.

  The candidate character string output unit 1005 outputs one or more candidate character strings using the character string from the first character storage unit 1004 and the converted character string from the character string conversion unit 1002 as candidate character strings. If the character string from the first character storage unit 1004 and the converted character string from the character string conversion unit 1002 are common character strings, the candidate character string output unit 1005 combines these character strings into one. .

  When the user selects one of the candidate character strings output from the candidate character string output unit 1005, the accumulated character string control unit 1006 performs the character strings of the first character storage unit 1004 and the second character storage unit 105. The contents of the buffer are erased, and the contents of the first character storage unit 1004 and the second character storage unit 105 are emptied. That is, if the accumulated character string control unit 1006 determines that one character is selected from the candidate character strings output from the candidate character string output unit 1005, the first character accumulation unit 1004, the second character accumulation unit 105, Is set to empty, a new input character is accepted, and if it is determined that no character is selected from the candidate character strings output from the candidate character string output unit 1005, the first storage means and the first (2) It is set so as to accept the input character following the content while maintaining the content with the storage means.

  Next, an example of the operation of the character input device according to the second embodiment will be described with reference to FIG. 12, FIG. 13, and FIG. In the example shown here, the first character determination unit 1003 and the second character determination unit 104 perform character recognition processing using the handwriting information, and the inputable character recording unit 102 records the Roman character conversion table as shown in FIG. A flow of processing in the case of being performed will be described. As a specific example, a case where “ka” is input to input the character “ka” will be described.

(I) First, as shown in FIG. 13, the inputable character extraction unit 103 calculates an inputable character set that can be input to the first character before the first character is input (step S703). Since no character string is stored in the second character storage unit 105, the inputable character extraction unit 103, as shown in FIG. 13, starts the head of all the character strings included in the inputable character recording unit 102 (FIG. 3). Set of characters {a, i, u, e, o, k, s, t, n, h, m, y, r, w, g, z, d, b, p, c} (*)
Is calculated as an inputable character set (step S703).

(Ii) Next, as shown in FIG. 13, when the input acquisition unit 101 inputs “k” as the first character, the first character determination unit 1003 reads the input handwriting from the character set 1 (= all characters). "K" is determined (step S1201), and the first character storage unit 1004 stores "k" (step S1202). On the other hand, as shown in FIG. 13, the second character determination unit 104 sets the character set (*) that can be input as the character set 2 and determines “k” that is similar to the input handwriting from the character set 2 (step In step S701, the second character storage unit 105 stores “k” (step S702). As shown in FIG. 13, the inputable character extraction unit 103 performs a search process on the inputable character recording unit 102 (FIG. 3) using “k” stored in the second character storage unit 105 as a search key.
{A, i, u, e, o, y} (**)
Is calculated as an inputable character set (step S703). The character string conversion unit 1002 searches the conversion character string recording unit 1001 (FIG. 11) using “k” output from the character string output unit 106 as a search key, and outputs “k” as the conversion character string ( Step S1203). Next, the candidate character string output unit 1005 combines the character string stored in the first character storage unit 1004 and the output from the character string conversion unit 1002, and outputs “k” as the candidate character string (step S1204). ). Finally, when the user selects one, the stored character string control unit 1006 erases the contents of the character string buffers of the first character storage unit 1004 and the second character storage unit 105, and the first character storage unit 1004. The contents of the second character storage unit 105 are emptied (step S1205), but the user does not select at this point, so nothing is done.

(Iii) Next, as shown in FIG. 14, it is assumed that “a” is written as the second character in a shape similar to “d”. At this time, as shown in FIG. 14, the first character determination unit 1003 determines “d” similar to the input handwriting from the character set 1 (= all characters) (step S1201). The first character storage unit 1004 stores “kd” (step S1202). On the other hand, as shown in FIG. 14, the second character determining unit 104 sets the character set (**) as the character set 2 and determines “a” similar to the input handwriting from the character set 2 (step S701). The second character accumulating unit 105 accumulates “ka” (step S702). As shown in FIG. 14, the inputable character extraction unit 103 performs a search process for the inputable character recording unit 102 (FIG. 3) using “ka” as a search key (step S703). Here, “ka” completely matches the character string in FIG. 3, so (*) is calculated as the input character set as in the case where the character string is not accumulated in the second character accumulation unit. The character string conversion unit 1002 performs a search process on the conversion character string recording unit 1001 (FIG. 11) using “ka” output from the character string output unit 106 as a search key, and outputs “ka” as a conversion character string. (Step S1203). Further, as shown in FIG. 14, the character string conversion unit 1002 may also pass the search character string “ka” as it is to the candidate character string output unit 1005. Next, the candidate character string output unit 1005 combines the character string stored in the first character storage unit 1004 and the output from the character string conversion unit 1002, and outputs “kd” and “ka” as candidate character strings. (Step S1204). Finally, as shown in FIG. 14, when the user selects one of the candidate character strings “kd” and “ka”, the accumulated character string control unit 1006 displays the first character accumulation unit 1004 and the second character. The contents of the character string buffer in the storage unit 105 are deleted, and the contents of the first character storage unit 1004 and the second character storage unit 105 are emptied (step S1205).

For the third and subsequent characters, the same process may be repeated each time a character is input. In the example described here, the character string conversion unit 1002 outputs only the result of performing the Roman character kana conversion based on FIG. 11 and the character string before conversion (in the above example, “ka”) as converted characters. However, the following character string may be added to the converted character string.
Character string converted to kana / kanji (for example, “OK” or “Section” for “ka”)
Predicted character string that matched forward (for example, “pretty” for “ka”, “signboard”, etc.)
According to the second embodiment described above, the character input device of the first embodiment is based on the premise that a character is always input by performing Romaji conversion, whereas the second embodiment. In this character input device, it is also taken into consideration that characters are directly input without performing Romaji conversion. In the above example, even if the user intends to write “kd” without using Romaji conversion, it is not necessary to input the character again because “kd” remains in the candidate character string. In the character input device according to the second embodiment, it is possible to input an alphabetic character string directly or to input another character string using character string conversion such as Romaji conversion without switching the mode in either case. Can be entered.

(Third embodiment)
A character input device according to a third embodiment will be described with reference to FIGS. 15 and 16.
The character input device of this embodiment is almost the same as the character input device of the second embodiment, but only the first character determination unit 1501 is different.
The first character determination unit 1501 receives the inputable character set from the inputable character extraction unit 103 as in the second character determination unit 104, determines the character set 1 according to the inputable character set, and determines the character set 1 The character that best matches the input information of the input acquisition unit 101 is selected from among the characters, and the character is passed to the first character storage unit 1004.

  In the second embodiment, the first character determination unit 1003 performs character determination using a fixed character set 1 (normally all characters), whereas in the third embodiment, the first character determination unit 1501. Determines the character set 1 to be subjected to character determination based on the inputable character set obtained by the inputable character extraction unit 103. Usually, as the character set 1, a character set obtained by excluding an inputable character set from all characters is used.

  An example of the operation of the character input device of the third embodiment is as shown in FIG. 16, and only the step S1201 of FIG. 12 is changed to step S1601 of FIG. 16 from the character input device of the second embodiment. is there. The only content that has been changed is that the character set 1 of the first character determination unit 1501 has been determined according to the character set that can be input.

  According to the third embodiment described above, it is possible to reduce overlapping characters between the contents of the character set 1 of the first character determination unit and the contents of the character set 2 of the second character determination unit, and the candidate character string output unit Can reduce the amount of processing for combining the characters with duplicates into one. Moreover, character determination can be performed more efficiently than character determination based on all characters.

  According to the embodiment described above, an alphabetic character combination table that can be converted by Romaji conversion, Pinyin conversion, or the like is prepared in advance, and the character string input so far is collated with the character combination table. Next, a character set that can be input next is calculated, and character determination is performed by limiting the character set for the next input character. In this way, by performing character determination for the next input character only from the character combination table and the character string input so far, to the character set that may be input next, erroneous recognition and error are detected. The possibility of input can be reduced.

DESCRIPTION OF SYMBOLS 101 ... Input acquisition part, 102 ... Inputable character recording part, 103 ... Inputable character extraction part, 104 ... Character determination part, 2nd character determination part, 105 ... Character storage part, 106 ... Character string output unit, 1001 ... Conversion character string recording unit, 1002 ... Character string conversion unit, 1003, 1501 ... First character determination unit, 1004 ... First character storage unit, 1005 ... Candidate character string output unit, 1006 ... Accumulated character string control unit.

Claims (10)

Accepting means for accepting input of characters one by one as input characters;
First selection means for selecting, as a first selection character, one character corresponding to the input character from a first character set including one or more characters of one character;
Second selection means for selecting, as a second selection character, one character corresponding to the input character from a second character set including one or more characters of one character;
First storage means for storing the first selected character;
Second storage means for storing the second selected character;
First recording means for recording one or more characters having meaning as a character string in a certain character conversion;
A first extraction that selects a character that matches the second selected character from the first recording means as a recording character, and extracts all the characters that immediately follow each recording character from the first recording means as an inputable character set. Means,
First setting means for setting the second character set as the inputable character set and receiving a new input character;
A second recording means for recording one or more characters in association with one or more converted characters obtained by converting the characters into the character;
A second extraction unit that searches the second recording unit for a character that matches the second selected character as a matching character, and extracts all converted characters associated with the matching character as extracted characters;
Determining means for determining whether one character is selected or not selected from the first selected character and the extracted character;
When it is determined that one character is selected from the first selected character and the extracted character, the contents of the first storage means and the second storage means are set to be empty and a new input character is accepted. When it is determined that no character is selected from the first selected character and the extracted character, the contents of the first storage means and the second storage means are maintained and the input character following the content is maintained. And a second setting means for receiving the character input device. If there is no second selected character, the first extracting means sets all one or more characters used as the first character among the characters recorded in the first recording means as an initializable character set. Extract and
The character input device according to claim 1, wherein the second selection unit selects one character corresponding to the input character as the second selected character from the initial possible character set.   The character input device according to claim 1, wherein the first character set includes all characters recorded in the first recording means.   The character input device according to claim 1, wherein the first character set includes characters obtained by removing the inputable character set from all characters recorded in the first recording unit. The accepting means accepts handwriting information of the input characters;
The said 1st selection means and the said 2nd selection means perform a character recognition process using the said handwriting information, respectively, The said 1st selection character and the said 2nd selection character are selected, respectively. Item 5. The character input device according to any one of Items 4. The accepting means accepts the coordinate information when tapped on the virtual keyboard displayed on the screen as the input character,
The first selection means and the second selection means select the first selection character and the second selection character, respectively, with reference to the coordinates corresponding to the coordinate information and the position coordinates of each key of the virtual keyboard. The character input device according to claim 1, wherein the character input device is a character input device. The accepting means accepts a character corresponding to a pressed key of the keyboard as the input character;
The first selection unit and the second selection unit select the first selection character and the second selection character, respectively, with reference to the key and arrangement information of a plurality of keys arranged on the keyboard. The character input device according to claim 1, wherein the character input device is a character input device.   The said 1st recording means has recorded the alphabet input by the Romaji input as a 1 or more character string which has a meaning as the said character string, The any one of Claim 1 to 7 characterized by the above-mentioned. The character input device described in 1.   8. The method according to claim 1, wherein the first recording unit records an alphabet input by Chinese pinyin input as a character string of one or more characters having meaning as the character string. The character input device according to claim 1. Accepting means accepts character input one character at a time as input characters,
The first selection means selects, as a first selection character, one character corresponding to the input character from a first character set including one or more one character.
A second selection means selects, as a second selection character, one character corresponding to the input character from a second character set including one or more characters of one character;
First storage means, said first selected character to accumulate,
Second storage means, said second selected character to accumulate,
The first recording means records one or more characters having meaning as a character string in a certain character conversion ;
The first extracting means selects a character that matches the second selected character from the first recording means as a recording character, and the next character immediately after each recording character is set as an inputable character set from the first recording means. Extract all,
The first setting means accepts a new input character by setting the second character set as the inputable character set.
A second recording means records one or more characters in association with one or more converted characters obtained by converting the characters into the character ;
The second extraction means searches the second recording means for a character that matches the second selected character as a matching character, and extracts all converted characters associated with the matching character as extracted characters,
The determination means determines whether to select one character or not one character from the first selected character and the extracted character,
If it is determined that the second setting means selects one character from the first selected character and the extracted character, the contents of the first storage means and the second storage means are set to be empty. When a new input character is received and it is determined that no character is selected from the first selected character and the extracted character, the contents of the first storage means and the second storage means are maintained. A character input method characterized by receiving an input character following the content.

Images