JP2917748B2 - Character input device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character input device for processing a handwritten character every time one character is input, and more particularly to a character input device in which words to be input such as addresses and names are registered in a word dictionary. It concerns the device.

[0002]

2. Description of the Related Art FIG.
1 is a block diagram showing a conventional character input device shown in FIG.
Is a character recognition means for recognizing a handwritten input character and outputting a candidate character or a distance between the input character and the candidate character; and 3, a recognition result storage means for storing the candidate character output from the character recognition means and the distance. , 63 are character group tables storing character grouping information, and 8 is control means for controlling the entire apparatus.

Next, the operation will be described. When a writer inputs a word, he or she writes it in the input / display means 1 one by one by hand. First, when the first character of the word is entered, the handwriting data of the handwritten character is passed to the character recognition means 2 through the control means 8. The character recognition means performs character recognition from the input data, and the resulting candidate characters and distance are stored in the recognition result storage means 3.

[0004] The control means 8 refers to the character group table 63, and if any of the candidate characters stored in the recognition result storage means 3 includes a representative character of the character group table 63, the constituent characters of the corresponding character group are set as candidate characters. In addition, it is displayed on the input / display means 1, and if not, only the candidate characters are displayed. The character group table 63 is as shown in FIG. 24, where 64 is a representative character of the character group, 6
Reference numeral 5 is a constituent character of a character group. When the first-ranked candidate character is misread or rejected, the writer selects the correct character in the displayed characters. Then, the selected character becomes an input character.

The same procedure is followed for character recognition, candidate character display, and character correction for the second and subsequent characters of a word.
By inputting all the characters constituting the word in this way, one word is input.

[0006]

As described above, the conventional character input device recognizes characters one by one and inputs words. Therefore, even if a word to be input, such as a prefecture name, is predetermined, all characters are input. There was a problem that the user had to input characters and the input efficiency was poor. If a word to be entered contains a character that is not recognized, the character cannot be input by character recognition and must be used as a group constituent character in the character group table. If there is no suitable recognition target character with a similar glyph, the scribe must remember the representative character with a different glyph corresponding to the non-recognition character, which places a heavy burden on the scribe. There were challenges.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems. When a word to be input, such as a prefecture name, is determined in advance, it is possible to correctly input a word without having to write all characters. It is another object of the present invention to provide a character input device capable of efficiently inputting a word without imposing a burden on a writer even when a word to be input includes a non-recognized character.

[0008]

The character input device according to the present invention has the following elements. (A) input / display means for inputting a handwritten character, displaying a screen, and selecting on a display screen; (b) recognizing the handwritten character input by the input / display means and determining a distance or similarity between the plurality of candidate characters. (C) recognition result storage means for sequentially storing candidate characters output by the character recognition means and distance or similarity, (d) word dictionary storing words to be input, (e) inputting .Of the input target word by the display means
During the input , using the candidate characters and the distance or similarity stored in the recognition result storage means, the partial character string or the entire character string of the word in the word dictionary is compared with the input word. > Matching and extracting means for extracting candidate words including characters equal to or more than the number of characters in the middle of the input , (f) displaying the candidate words extracted by the word matching and extracting means on the input / display means and displayed A control unit that uses the candidate word selected from the candidate words by the input / display unit as an input word.

According to a second aspect of the present invention, in the character input device, the control means selects another candidate character at the time of displaying the candidate character of the input character, and the recognition result.
The contents of the result storage means are selected by the input / display means.
Means for modifying the provided character, above in accordance with the correction content
Input the above candidate word re-extracted by the written word collation extraction means
-It is displayed on the display means .

According to a third aspect of the present invention, there is provided a character input device according to the first or second aspect.
With a priority table that sets the priority for each word ,
The word collation and extraction means determines the priority in the priority table.
This is used to extract candidate words .

According to a fourth aspect of the present invention, there is provided a character input device including a frequency table for storing an input frequency for each input word;
In this method, candidate words are extracted using the frequency information in the file .

A character input device according to a fifth aspect of the present invention includes a priority table in which priority is set for each word and a frequency table for storing the input frequency of the input word.
The word matching and extracting means includes the priority table and the frequency
A candidate word is extracted using a table .

According to a sixth aspect of the present invention, there is provided a character input device including a priority table and a frequency table as word information of the word dictionary.

[0014]

[Action] In the character input device according to the first aspect of the present invention, have use the candidate characters and the distance or similarity stored in the recognition result storage means for the recognition result of the input character is sequentially stored, the word A substring of a word in the dictionary or the entire string
Word matching extraction means for extracting candidate words by matching
There is a characteristic in the glow. This makes it possible to display candidate words obtained using the candidate characters of the characters within the input range and the distance or similarity even during the word input,
Since there is an input word among the candidate words, and if the candidate word is selected, the selected word is used as the input word.
You can enter words correctly without entering all words,
Furthermore, a word including a character not to be recognized can be input efficiently.

In the character input device according to the second aspect of the present invention, the control device includes means for selecting another candidate character when displaying the candidate character of the input character, and a recognition unit.
Sentence selected by input / display means
It is characterized in that means for correcting characters are provided, and candidate words are re-extracted and displayed according to the contents of correction. Thus, if the character is corrected, the probability that the word to be input is included in the candidate words increases, and the word input becomes easy.

The character input device according to the third aspect of the present invention is characterized in that a priority table in which the priority is set for each word is provided. As a result, if the priority of a word that is frequently input is increased, the probability that the word is included in the higher rank of the candidate word is increased, and the word input is facilitated.

The character input device according to the fourth aspect is characterized in that a frequency table for storing the input frequency of words is provided. As a result, as a large number of words are input, the probability that a candidate word with a high input frequency is included in a higher rank of the candidate word increases, and the word input becomes easier.

The character input device according to the present invention is characterized in that both the priority table and the frequency table are provided. Since the candidate words are extracted using both tables, word input is further facilitated.

The character input device according to the invention is characterized in that the word dictionary is provided with a priority table and a frequency table as word information. Thus, the capacity of the storage device for storing the word dictionary, the priority table, and the frequency table can be reduced.

[0020]

【Example】

Embodiment 1 FIG. An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a configuration diagram showing one embodiment of a character input device. In FIG. 1, the same components as those in the conventional example are denoted by the same reference numerals, and description thereof will be omitted. 4 is a word collating and extracting means for extracting candidate words, 5 is a word dictionary stored in a known storage device storing words to be input, and 6 and 7 are priority words for each word in a storage device (not shown). 6 is a priority table storing degrees and a frequency table storing the input frequency of each word.

Next, the operation will be described with reference to the drawings. Here, a description will be given using an example in which the word “Higashi-Kurume-shi” is input using the candidate characters and the distance as the recognition result. FIG. 2 is a processing flowchart of the character input device of the present invention. First, the scribe
"East" which is the first character of "Higashi-Kurume-shi" is input by handwriting on the input / display means 1 (step S1) .FIG. 4 shows an example of a screen of the input / display means at that time, and 9 denotes handwriting input. And a character frame used for displaying the recognition result, and 10 is an image showing the handwriting of the entered character "East." The handwriting data input here is passed to the character recognition means 2 via the control means 8. .

In the character recognition means 2, as in the conventional device,
Character recognition is performed from the handwriting data (step S2), and the distance between the plurality of candidate characters and each candidate character is output as a recognition result and stored in the recognition result storage means 3 (step S3). FIG. 5 is an example showing the contents of the recognition result storage means 3. 11 is a recognition result storage area for the first character of the word, 12 is a candidate character, and 13 is the distance between each candidate character. Input character "East"
As a recognition result for No. 10, the first candidate character "fur",
The distance is 30, the second candidate character "east", the distance 45
Is stored.

Next, the word collation / extraction means 4 collates the recognition result stored in the recognition result storage means 3 with the word dictionary 5, and extracts candidate words (step S4). In candidate word extraction, a word evaluation value obtained for each word including characters equal to or greater than the number of written characters is used. Words whose number of characters is less than the number of entered characters are excluded. This word evaluation value W
Is the k-th character evaluation value of the word, D _k , and the number of entered characters is n
Then, for example, it is obtained by the following equation.

[0024]

(Equation 1)

However, the character evaluation value D _k is as follows. (1) candidate character is not present in the corresponding input character (rejection) if ... C ₁ While (predetermined constant) (2) candidate character exists, if does not contain constituent characters of a word ... C ₂ ( (3) When the constituent characters of the word are included in the candidate characters… When the candidate character and the distance are used as the recognition result, the distance between the candidate character and the similarity with the candidate character are used. The similarity value of the corresponding candidate character is -1
Multiplied by

FIG. 6 is an example showing the word dictionary 5, and FIG.
The evaluation value of each word is calculated from the candidate characters and the distance shown in FIG. Here, when the C ₁ to 100, C ₂ 255 of the above calculation formula, the word evaluation value of the word "Sarashina City" and the word "Koshoku City", the distance of the candidate character "further" is 30 because 30 . In addition, for a word such as the word "Higashine" whose first character is "East", the distance between the candidate character "East" is 45 and the word evaluation value is also 45. For the other words, the word evaluation values are all 255 (C ₂ ) because the word constituent characters are not included in the candidate characters. The word matching extraction means 4 extracts words up to a predetermined number as candidate words in order from the one having the smallest word evaluation value calculated in this way, but for words having the same word evaluation value, the word dictionary 5 In the order of storage. Here, assuming that the word collating and extracting means 4 extracts a maximum of four candidate words, from the recognition result of FIG. 5 and the word dictionary 5 of FIG. 6, “Kohanshi”, “Karashishi”, “Higashineshi”, “ The four words "Higashitagawa-gun" are candidate words.

Thereafter, the control means 8 causes the input / display means 1 to display the extracted candidate words (step S5). FIG.
Is an example of the screen display of the input / display means 1 at that time, 14 is "fur" which is the first candidate character of the first character "east" of the word input in step S1, and 15 is the end of word input. An end button 16 for instructing is a candidate word display area.
Kobashi-shi, “Karakada-shi”, “Higashine”, “Higashitagawa-gun”
Two candidate words are displayed. At this point, the character input device determines whether the input of the next character has started (step S
6) It is detected whether any of the displayed candidate words has been selected with a pen or the like (step S7) and whether the end button 15 has been selected with a pen or the like (step S8). Here, when any one of the candidate words is selected, the selected candidate word is input, and when the end button 15 is selected, the word composed of the character recognition result of each input character is the input word. . However, in this embodiment, since the input word is not included in the candidate words, the next character is input (step S6).

When the second character of "Higashi-Kurume-shi" is input by hand (step S1), the character recognizing means 2 performs character recognition as in the case of the first character (step S2). The result is stored in the recognition result storage means 3 (step S3). FIG.
Is an example showing the contents of the recognition result storage means 3 after the recognition result of the second character is stored, and the recognition result of the first character is held. 17 is a recognition result storage area for the second character of the word, 18 is a candidate character, and 19 is the distance between each candidate character.

The word collating and extracting means 4 performs collation with the word dictionary 5 using the recognition result of the first character and the second character, and extracts a new candidate word (step S4). Here, the word evaluation value is calculated again by the above formula. The word "Shokuichi"
, The character evaluation value of the first character is 30 which is the distance of the candidate character “fur”, but the character evaluation value of the _second character is 255 (C ₂ ),
A total of 285 is the word evaluation value. Similarly, the word evaluation values of other words are calculated, and when four candidate words are extracted in the order of the smallest word, “Higashi-Kurume-shi” has the smallest word evaluation value.
(The word evaluation value is the sum of the distances between the candidate characters “East” and “Hi” 9
5) In the following order, “Higashiyamato” (word evaluation value is candidate character)
The sum of the distance between the east and the large 105), "Higashiosaka-shi" (the word evaluation value is the sum of the distance between the candidate characters "east" and the "large" 10
5), "Higashi-Yatsushiro-gun" (word evaluation values are candidate characters "east" and "
145).

After that, the control means 8 causes the input / display means to display the extracted candidate words (step S5). FIG.
This is an example of the screen display of the input / display means 1 at that time,
0 is the first candidate character of the second input character, “Hi”, 21
Is the candidate word "Higashi-Kurume", 22 is the candidate word "Higashiyamato", 23 is the candidate word "Higashiosaka", 24 is the candidate word "
Higashi-Yatsushiro-gun ". Since the word" Higashi-Kurume-shi "21 to be input is included in the candidate words, the writer selects" Higashi-Kurume-shi "21 with a pen or the like. The means 8 detects this operation (step S7) and uses the word "Higashi-Kurume-shi" 21 as an input word (step S9).

In this embodiment, through the above steps, "
The five-letter word "Higashi-Kurume-shi" can be input by handwriting input of only the first two characters. According to the present invention, a word can be input by inputting fewer characters than the number of constituent characters of the word to be input.

Although the maximum number of candidate words to be extracted is four in this embodiment, any number of candidate words may be used. Also,
Although all the extracted candidate words are displayed on the screen at the same time, only some of the candidate words may be displayed on the screen at one time. Further, in the present embodiment, the proper noun is used as the input word, but any character string that can be stored in the word dictionary may be used. Needless to say, it can be applied to.

Embodiment 2 FIG. Here, as another embodiment of the present invention, an example will be described in which a word including a character not to be recognized in character recognition is input. Specifically, it is assumed that the word “Higashikubijo-gun” is to be input, and the second character “neck”
Is not included in the characters to be recognized in character recognition. Further, a case where a candidate character and a distance are used as a recognition result will be described. The processing flow diagram is the same as in the first embodiment.
It is.

Next, the operation will be described with reference to the drawings. First, in the same procedure as in the first embodiment, the first two characters “East” of “Higashikubiki-gun”
And "neck" have already been entered. FIG.
This is an example showing the storage content immediately after the recognition result of the second input character is stored in the recognition result storage means 3 (step S3),
Is the first candidate character, 26 is the distance to each candidate character, and 27 and 28 are the distance to the second candidate character, respectively. As shown in FIG. 10, the result of character recognition of the second character that is a non-recognition character is rejected (unrecognizable), and no candidate character is output. As for the first character, "east" is misread as "bundle" as a result of character recognition. From these recognition results, candidate words are extracted (Step S4) and displayed (Step S5). It is assumed that candidate words are extracted in the same manner as in the first embodiment. FIG.
1 is an example of the screen display of the input / display means 1 at this time, 29 is the first candidate character “bundle” of the first character, and 30 is 2
The mark "?" Indicating rejection as the recognition result of the character is displayed in the candidate word display area 16, and four candidate words obtained from the recognition results of the first and second characters are displayed.

Since the writer does not have the word "Higashikubiki-gun" to be input among the candidate words, the writer inputs the third character "castle" through step S6 (step S1). Thereafter, the character recognizing means 2 recognizes the third input character (step S).
2) The result is stored in the recognition result storage means 3 (step S)
3) is done. FIG. 12 is an example showing the contents of the recognition result storage means 3 immediately after the recognition result of the third character is stored.
Reference numeral 1 denotes an area for storing the recognition result of the third input character.
Is a candidate character, and 33 is a distance to each candidate character. 3
As a result of character recognition of the character, it can be seen that “jo” is misread as “saka”. Here, the word collation extracting unit 4 collates with the word dictionary 5 using the recognition result of the first to third characters,
A word evaluation value is obtained by the calculation described in the first embodiment, and candidate words are extracted (step S4). When the word evaluation value is obtained from the word dictionary 5 of FIG. 6 and the recognition result of FIG. 12, the word evaluation value having the smallest word evaluation value is “Higashiosaka” (the word evaluation value is the distance [50] of the candidate character “east” and C ₁ [100] and the sum [200] of the distance [50] between the candidate characters "Saka" are obtained.
(The word evaluation value is the distance [50] of the candidate character “East” and C ₁
Sum total of distance [55] between [100] and candidate character "castle" 20
5), “Higashiibaraki-gun” (the word evaluation value is the sum 205 of the distance [50] of the candidate character “east”, C ₁ [100] and the distance [55] of the candidate character “castle”), “Higashiyamato-shi "(word evaluation value, the candidate character" sum 240 of distances [90] of the sum "E" candidate character distance [50] and C ₁ [100] ") 4 single word of the candidate word.

After that, the control means 8 causes the input / display means 1 to display the extracted candidate words (step S5). FIG.
Is an example of the screen displayed at that time. Reference numeral 34 denotes the first candidate character of the third character, "Saka", and reference numerals 35 to 38 denote the candidate words "Higashiosaka-shi", "Higashikubiki-gun", "East". It is “Ibaraki-gun” and “Higashiyamato”. Here, the writer selects the “Higashikubiki-gun” 36 with a pen or the like because the word “Higashikubiki-gun” to be input is included in the candidate word. The control means 8 detects this operation (step S7) and uses the word "Higashikubikijo-gun" as an input word (step S9).

In this embodiment, through the above steps, the word "Higashikubiki-gun" including the non-recognizable character "neck" can be correctly input by only three character inputs. Further, even though the recognition results of the input characters are all incorrect, correct word input can be performed without correcting the recognition results.

Embodiment 3 FIG. Here, another embodiment of the present invention will be described with reference to the drawings. FIG. 3 is a processing flowchart.
In FIG. 3, steps T1 to T9 correspond to step S in FIG.
Same as 1 to S9.

Next, the operation will be described with reference to the drawings. here"
An example of entering the word "Higashi-Yatsushiro-gun" will be explained.
The case where the candidate character and the distance are used as the recognition result will be described. First, it is assumed that the first two characters “Higashi” and “Hachi” of “Higashi-Yatsushiro-gun” have already been entered in the same procedure as in the first embodiment. FIG. 14 stores the recognition result of the second character (step T3).
6 is an example showing the contents of the recognition result storage means 3 immediately after the recognition. Reference numeral 39 denotes a candidate character of the first input character, and reference numeral 40 denotes the distance. 41 is a candidate character of the second input character, and 42 is its distance. From FIG. 14, it can be seen that the first character of the input is correctly recognized, but the second character is the third candidate character of the correct character. From these recognition results, candidate words are extracted from the word dictionary 5 shown in FIG. 6 in the same procedure as in the first embodiment (step T4) and displayed on the input / display means 1 (step T5). FIG. 15 shows an example of the screen display of the input / display unit 1 at this time, where 43 is the first candidate character “East” of the first input character, and 44 is the first candidate character of the second input character.
45 and 46 are candidate buttons for displaying candidate characters of input characters. In the candidate word display area 16, "Higashiyamato City", "Higashiosaka City", Four words "Higashiyamanashi-gun" and "Higashiyama-ku" are displayed, and the word "Higashi-Yatsushiro-gun" to be input is not included.

Here, the writer selects the candidate button 46 corresponding to the input second character with a pen or the like in order to correct the recognition result of the erroneously read second character. And the control means 8
The operation is detected (step T101), and candidate characters are displayed as shown in FIG. 16 (step T102). In FIG. 16, reference numeral 47 denotes a candidate character displayed on the input / display unit 1, and the candidate character 41 of the second input character stored in the recognition result storage unit 3 is sequentially displayed. When the writer selects the correct answer "8" from the displayed candidate characters 47, the control means 8 detects this operation (step T103).
Then, the recognition result is corrected (step T104). To correct the recognition result, for example, in the recognition result storage area of the corresponding character in the recognition result storage means 3, the first candidate character is set to the designated correct character, the distance is set to 0, and the second or lower candidate character is set. This is done by erasing. FIG. 17 is an example showing the contents of the recognition result storage means 3 after the correction of the recognition result. 48 is a candidate character of the second input character after correction, and 49 is its distance.
As a result of the correction of the recognition result, the character "8" in which the first candidate character is designated, its distance is 0, and the second and lower candidate characters are no longer present. Note that the distance
If similarity is used instead of separation, the first candidate sentence of the character
Character as the specified correct character, and determine the similarity
Correction by setting a sufficiently large constant
Do.

Thereafter, the word matching and extracting means 4 performs matching with the word dictionary 5 using the corrected recognition result, and re-extracts new candidate words (step T4). At this time, the word evaluation value is calculated using the formula shown in the first embodiment, and candidate words are extracted in ascending order. In this example, the word having the smallest word evaluation value is “Higashi-Yatsushiro-gun” (word evaluation value Is the distance [50] of the candidate character “East” and the distance [0] of the candidate character “Eight”
(The sum of 50) and “Higashine” (word evaluation value is the sum of the distance [50] of the candidate character “east” and C ₂ [255] in the following order)
5), “Higashidagawa-gun” (word evaluation value is the sum 305 of distance [50] of candidate character “east” and C ₂ [255] 305), “Higashiyamato-shi” (word evaluation value is candidate character “east” consisting distance [50] and 4 words in total 305) of C ₂ [255].

Next, the control means 8 causes the input / display means 1 to display the extracted candidate words (step T5). FIG.
Is an example of the screen display at that time, and 50 is the modified 2
The first candidate character of the character, "8", 51 to 54 is the candidate words "Higashi-Yatsushiro-gun", "Higashine City", "Higashitagawa-gun", and "Higashiyamato City", respectively. Here, the writer selects “Higashi-Yatsushiro-gun” 51 with a pen or the like because the word “Higashi-Yatsushiro-gun” to be input is included in the candidate words. The control means 8 detects this operation (step T7) and uses the word "Higashi-Yatsushiro-gun" as an input word (step T9).

In this embodiment, through the above steps, 4
The first two characters of the character word "Higashi-Yatsushiro-gun" can be input by handwriting, and the word can be correctly input only by correcting the recognition result of one of the characters.

Embodiment 4 FIG. Here, still another embodiment of the present invention will be described with reference to the drawings. FIG. 2 is a processing flowchart.

Next, the operation will be described with reference to the drawings. An example will be described in which the word “Higashi-Kurume-shi” is input using the candidate characters and the distance as the recognition result, as in the first embodiment. As in the first embodiment, when "Higashi", which is the first character of "Higashi-Kurume City", is input by hand (step S1), the character recognizing means 2 performs character recognition (step S2), and the recognition result is obtained. Is stored in the recognition result storage means 3 (step S3). FIG. 5 shows an example in which the recognition result is stored in the recognition result storage means 3 as in the first embodiment.

The word collating and extracting means 4 collates the word dictionary 5 with the recognition result, finds a word evaluation value, and extracts candidate words (step S4). At this time, the word evaluation value is calculated by, for example, the following formula in consideration of the contents of the priority table 6.

[0047]

(Equation 2)

Where D _k is a character evaluation value.
Can be obtained by a method similar to the method shown by. P is a priority in the priority table.

For example, FIG. 19 shows the contents of the priority table, in which a preset priority is stored for each word in the word dictionary 5. Here, it is assumed that a word having a high priority value is defined to have a high priority. In FIG. 19, "Higashi Kurume City"
Has a priority of 20, and the other words have a value of 0 as the priority.

When the word collating and extracting means 4 finds word evaluation values from this priority table and extracts candidate words in ascending order, the smallest word is "Higashi Kurume City" (the word evaluation value is "East" distance [45] and priority [2
0], followed by “Kohakushi” and “Kohashiki”
(In both cases, the word evaluation value is the difference 30 between the distance [30] of the candidate character "fur" and the priority [0].) Then, "Higashine" (the word evaluation value is the distance [ 45] and the priority [0].

Then, the control means 8 displays the extracted candidate words on the input / display means 1 (step S5). FIG.
Is an example of the screen display at that time, and 55 to 58 are the candidate words "Higashi-Kurume City", "Shohan City", "Sara Class City", ""
Here, the writer selects “Higashi-Kurume-shi” 55 with a pen or the like because the candidate wants to input the word “Higashi-Kurume-shi” 55. The control means 8 detects this operation. (Step S7), and the word “Higashi-Kurume-shi” is used as an input word (Step S9).

In the invention shown in this embodiment, a five-character word "Higashi-Kurume-shi" having a high priority can be obtained by inputting only one character by using a priority table in which the priority of each word is set in advance. , The word input operation can be performed at high speed.

Embodiment 5 FIG. Here, still another embodiment of the present invention will be described with reference to the drawings. FIG. 2 is a processing flowchart.

Next, the operation will be described with reference to the drawings. An example will be described in which the word “Higashi-Kurume-shi” is input using the candidate characters and the distance as the recognition result, as in the first embodiment. As in the first embodiment, when "Higashi", which is the first character of "Higashi-Kurume-shi", is input by hand (step S1), the character recognition means 2 performs character recognition (step S2), and the recognition result is It is stored in the recognition result storage means 3 (step S3). Here, similarly to the first embodiment, the recognition result shown in FIG.
Shall be stored in

The word matching and extracting means 4 matches the word dictionary 5 from the recognition result, finds a word evaluation value, and extracts candidate words (step S4). At this time, the word evaluation value is determined by, for example, the following formula in consideration of the value of the frequency table 7.

[0056]

(Equation 3)

Here, D _k is a character evaluation value.
Determined by the method shown in. F is the input frequency of the words in the frequency table, and _CF is a predetermined constant.

FIG. 21 shows an example of the frequency table, in which the input frequencies up to that time are stored for each word in the word dictionary 5. Words with a high input frequency are preferentially extracted. In FIG. 21, the input frequency of “Higashi-Kurume City” is 5, the input frequency of “Higashiyamato City” is 2, and the input frequency of other words is 0.
A word evaluation value is obtained using this frequency table, and candidate words are extracted in ascending order. Assuming that the constant C _{F in the} above formula is 5, the word evaluation value is the minimum at the word “Higashi-Kurume-shi” (the word evaluation value is the distance [45] of the candidate character “East” and the input frequency × C _F [5] × 5 = 25].
Next, the word evaluation value is “distance [30] of candidate character“ fur ”and input frequency × C _F [0 × 5 =
0], followed by “Higashiyamato” (the word evaluation value is the distance [45] of the candidate character “East” and the input frequency × _CF)
The difference from [2 × 5 = 10] is 35).

Next, the control means 8 causes the input / display means 1 to display the extracted candidate words (step S5). FIG.
Is an example of the screen display at that time, and 59 to 62 are candidate words, “Higashi-Kurume City”, “Shohan City”, “Sara Class City”, and “Higashiyamato City”, respectively. Here, the writer selects “Higashi-Kurume-shi” 59 with a pen or the like because the word “Higashi-Kurume-shi” to be input is included in the candidate words. The control means 8 detects this operation (step S7) and uses the word "Higashi-Kurume-shi" as an input word (step S9). At the same time, the control means 8 updates the frequency table 7 and increases the input frequency corresponding to the input word "Higashi-Kurume-shi" by one to prepare for the next word input.

In this embodiment, a five-character word "Higashi-Kurume-shi" can be input by inputting only one character by using a frequency table storing the input frequency of each word. As described above, frequently used words can be input at high speed, and the input operation can be performed efficiently.

In the first to fifth embodiments described above, an example has been described in which a candidate character and a distance are used as a recognition result. However, this may use a candidate character and similarity.

[0062]

As described above, according to the present invention, candidate words are extracted and displayed in the middle of word input using the recognition results of characters in the input range, and the candidate word selected by the scribe is selected. Is used as an input word, a word can be correctly input with a small number of character inputs, and a word including a character not to be recognized can be efficiently and easily input.

If any of the characters in the input range is corrected, a candidate word is extracted /
Since the display is re-displayed, the input efficiency can be further improved by correcting the recognition result.

Since the candidate words are extracted by using the priorities set in advance and the stored input frequencies, the words which are frequently input are easily included in the candidate words, and the efficiency of the input operation is further reduced. Can be raised.

Further, by incorporating the priority and the input frequency as components of the word dictionary, the capacity of the storage device can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a character input device of the present invention.

FIG. 2 is a flowchart of the character input device of the present invention in the first, second and fifth embodiments.

FIG. 3 is a flowchart of a character input device according to the present invention in a third embodiment.

FIG. 4 is a display example of an input / display unit according to the first embodiment.

FIG. 5 is an example of contents stored in a recognition result storage unit in the first embodiment.

FIG. 6 is an example showing the contents of a word dictionary.

FIG. 7 is a display example of input / display means in the first embodiment.

FIG. 8 is an example of contents stored in a recognition result storage unit in the first embodiment.

FIG. 9 is a display example of input / display means in the first embodiment.

FIG. 10 is an example of storage contents of a recognition result storage unit in the second embodiment.

FIG. 11 is a display example of input / display means in the second embodiment.

FIG. 12 is an example of storage contents of a recognition result storage unit according to the second embodiment.

FIG. 13 is a display example of input / display means in the second embodiment.

FIG. 14 is an example of storage contents of a recognition result storage unit in the third embodiment.

FIG. 15 is a display example of input / display means in the third embodiment.

FIG. 16 is a display example of input / display means in the third embodiment.

FIG. 17 is an example of storage contents of a recognition result storage unit in the third embodiment.

FIG. 18 is a display example of input / display means in the third embodiment.

FIG. 19 is an example of a priority table according to the fourth embodiment.

FIG. 20 is a display example of input / display means in the fourth embodiment.

FIG. 21 is an example of a frequency table according to the fifth embodiment.

FIG. 22 is a display example of input / display means in the fifth embodiment.

FIG. 23 is a block diagram showing a conventional character input device.

FIG. 24 is an example of a conventional character group table.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Input / display means 2 Character recognition means 3 Recognition result storage means 4 Word collation extraction means 5 Word dictionary 6 Priority table 7 Frequency table

Claims (6)

(57) [Claims] 1. A character input device has the following elements:, (a) input and display means for input and screen display and selection on the display screen of the handwritten characters, handwritten characters input by (b) the input and display means (C) recognition result storage means for sequentially storing the candidate characters output by the character recognition means and the distance or similarity, and (d) inputting. (E) a word dictionary storing target words, (e) a word dictionary of the input target words by the input / display means.
During the input , using the candidate characters and the distance or similarity stored in the recognition result storage means, the partial character string or the entire character string of the word in the word dictionary is compared with the input word. > Matching and extracting means for extracting candidate words including characters equal to or more than the number of characters in the middle of the input , (f) displaying the candidate words extracted by the word matching and extracting means on the input / display means and displayed A control unit that uses the candidate word selected from the candidate words by the input / display unit as an input word. Wherein said control means, when displayed candidate character in the input character, means for selecting another candidate characters, upper
The contents of the recognition result storage means are input and displayed by the above input / display means.
Means for correcting the selected character is provided, and the word matching / extracting means extracts candidate words re-extracted according to the contents of the correction.
2. The character input device according to claim 1, wherein the character is displayed on the input / display unit . 3. A includes a priority table which sets the priorities for each word, claims the word collating extracting means, characterized in that the extraction of the candidate words using the priority in the priority table The character input device according to claim 1 or 2. 4. A frequency table for storing an input frequency for each input word , wherein said word collating and extracting means extracts a candidate word using frequency information in said frequency table. The character input device according to claim 1 or 2, wherein: Comprise 5. word frequency table for storing input frequency of the input word with the priority table which sets the priorities for each, the said word collating extracting means using the priority table and frequency table candidate 3. The character input device according to claim 1, wherein a word is extracted. 6. The character input device according to claim 1, wherein the word dictionary includes a priority table and a frequency table as word information.