JP3042815B2 - Handwritten 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 an information processing apparatus such as an electronic organizer and a word processor having a handwritten character input function.

[0002]

2. Description of the Related Art Conventionally, an information processing device such as an electronic organizer or a word processor having a handwritten character input function (hereinafter referred to as a "handwritten character input device") cannot completely recognize inputted handwritten characters. For example, when a user of a handwritten character input device inputs a character to be input by handwriting, the input is performed due to a handwriting habit of the user or a character similar to the input character. In many cases, handwritten characters are not recognized as characters desired by the user.

In this case, in order to correct the first candidate character displayed as a recognition result, the user replaces the second candidate character or less in a candidate character group previously selected as a character similar to the first candidate. A desired character is selected and input from the displayed characters.

Further, the type of characters input by the user, for example, English characters and numbers, are set in the information processing device in advance, and the candidate characters for the input characters are limited to improve the recognition rate. There are methods.

[0005]

As described above, in the conventional handwritten character input device, the recognition rate of input handwritten characters is low due to the user's handwriting habits and the like. Therefore, for example, if the character cannot be accurately recognized due to the handwriting habit of the user, the user may select the character from the second candidate or lower and input the desired character, but the user may again input the same character again. Even if input, the handwritten character input device cannot recognize it accurately, and the user needs to repeat the operation of selecting from the characters below the second candidate, which is very troublesome.

When the user successively inputs the same type of character, the type of the subsequent characters is limited by the first character, but different types of characters may be used in the middle depending on the user's handwriting habit. When the character is the first candidate, the handwritten character input device selects and displays the character, so that it is necessary to perform a correction operation on the character, which is inconvenient.

For this reason, as described above, the type of character to be input is set in the handwritten character input device in advance, and by limiting the candidate characters for the input character,
Although the recognition rate can be improved, it becomes complicated for the user to select the character type each time according to the character type.

An object of the present invention to solve the above-mentioned problems is to provide a handwritten character input device capable of improving the recognition rate of input handwritten characters and easily performing an input operation.

[0009]

SUMMARY OF THE INVENTION The present invention has an input surface on which a plurality of coordinate points are set, and outputs character data which is a set of coordinate points representing handwritten characters drawn by a writing instrument on the input surface. Input means, candidate character storage means for storing candidate character data for recognizing the character data as a character, and candidate characters for retrieving candidate character data similar to the character data and arranging them in a candidate order from the first candidate. Search means for creating a data group, a recognition character storage means for sequentially storing candidate character data groups created by the search means, and a display means for displaying the candidate character data groups stored in the recognition character storage means in candidate order; A character selection unit for selecting a desired character to be adopted from among the candidate characters displayed on the display unit, and a character selected by the character selection unit as a new first candidate character Learning means for sequentially changing the arrangement of candidate character data groups stored in the recognition character storage means to create a new candidate character data group, and learning memory for sequentially storing the new candidate character data groups created by the learning means After the learning unit creates a new candidate character data group and stores it in the learning memory, the searching unit first searches the candidate memory for candidate character data in the learning memory. When there is no candidate character data input device, the candidate character data group is searched for in the candidate character storage means.

Further, the present invention includes a designation means for designating a type of a character to be input, and a character type storage means for storing type data of a candidate character selected at the time of the previous handwritten character input, wherein the display means comprises: When the type specified by the specifying unit or the type is not specified by the specifying unit, the candidate character data group stored in the recognized character storage unit based on the type data stored in the character type storage unit When there is candidate character data of the same type data, the candidate character data is displayed as a first candidate.

[0011]

According to the first aspect of the present invention, the input means comprises:
An input surface on which a plurality of coordinate points are set is provided, and character data which is a set of coordinate points represented by handwritten characters drawn with a writing instrument on the input surface is output. The candidate character storage unit stores candidate character data for recognizing the character data as a character, and the search unit searches for candidate character data similar to the character data, and arranges candidate character data arranged in the candidate order from the first candidate. Create character data group.

[0012] The recognized character storage means sequentially stores the candidate character data groups created by the search means, and the display means displays each character of the candidate character data group in the order of candidates. The character selection means selects a desired character to be adopted from the displayed candidate characters.

The learning means sequentially changes the arrangement of the group of candidate character data stored in the recognition character storage means using the selected character as a new first candidate, and sets the selected character as first candidate character data. Create a new candidate character data group,
The learning memory stores the new candidate character data group.

The search means first searches for candidate character data in the learning memory, and if no match is found, searches for candidate character data in the candidate character storage means. Therefore, when a handwritten character is input from the input means, an undesired character is displayed as a first candidate, and the desired character is selected by the character selection means, when the same handwritten character is input again from the input means, the search means The candidate character data is retrieved from a learning memory that stores the candidate character data group created by the learning means. Thus, the desired character can be displayed on the display unit as the first candidate without selecting the desired character by the character selection unit.

According to the present invention, in addition to the input means, the candidate character storage means, the search means, the recognized character storage means, the display means, the character selection means, the learning means, and the learning memory, Designation means and character type storage means are provided. The designating means designates the type of the character to be input, and the character type storing means stores the type data of the candidate character selected at the time of the previous handwritten character input. A display unit configured to display the candidate stored in the recognized character storage unit based on the type specified by the specification unit or based on the type data stored in the character type storage unit when the type is not specified by the specification unit; When the same type of candidate character data is included in the character data group, the candidate character data is displayed as a first candidate.

Therefore, when characters of the same type are successively input, the display means displays the first candidate based on the type specified by the previous specification means or the type data stored in the character type storage means. Since the character is selected and displayed, the display means displays the desired character as the first candidate.

[0017]

FIG. 1 is a block diagram showing a schematic electrical configuration of an information processing apparatus having a handwritten character input device function (hereinafter referred to as "handwritten character input device") according to an embodiment of the present invention. . A central processing unit (hereinafter abbreviated as “CPU”) 1 operates according to a program stored in a read-only memory (hereinafter abbreviated as “ROM”) 2 and controls the entire apparatus. The ROM 2 stores a program for controlling the operation of the apparatus, data necessary for handwritten character recognition, and the like. A random access memory (hereinafter abbreviated as “RAM”) 3 temporarily stores application data, programs, data necessary for handwritten character recognition, and the like, and is processed by the CPU 1.

An i / O integrated circuit (hereinafter abbreviated as “i / OLSi”) 4 detects an input state of a keyboard 5 and
PU1 is connected to the keyboard 5 via the i / OLSi4.
Can be determined. The display unit 6 uses a liquid crystal display device or the like, and includes a liquid crystal display for displaying data and the like, a liquid crystal driver for controlling the liquid crystal display, and the like. The CPU 1 displays the data on the display unit 6 by writing data to be displayed on the display memory 7.

The tablet 8 is of a transparent analog resistive film type and is mounted on the display unit 6 in a superimposed manner. When the tablet 8 is touched with a touch pen or a finger, the position information is transmitted through a control circuit 9. Is output. The tablet 8 is composed of two upper and lower panels, electrodes are formed at both ends of each panel, and a uniform resistive film is formed between the electrodes. In order to read the touched position information on the tablet 8, first, a voltage is applied to both ends of the upper flexible panel of the upper and lower panels, and resistance division of the touched position (in this case, the X coordinate) is performed. An analog voltage value corresponding to the ratio is output from the lower panel, which is converted to a digital value via the A / D conversion circuit 10 and the value of the X coordinate is converted to C value.
Read by PU1. Next, when a voltage is applied to both ends of the lower panel, an analog voltage value corresponding to the resistance division ratio at the touched position (in this case, the Y coordinate) is output from the upper panel, and the analog voltage value is output from the A / D panel. The data is converted into a digital value via the conversion circuit 10, and the value of the Y coordinate is read by the CPU 1.
As described above, the CPU 1 can know the touched position information (X coordinate, Y coordinate) on the tablet 8.

When inputting handwritten characters, the characters are written on the tablet 8 with a touch pen or the like and then input. The character data, which is the positional information of the input handwritten characters, is read by the CPU 1 at regular intervals, and the character data is compared with candidate character data stored in the ROM 2 for recognizing handwritten characters. CPU1
Reads a character determined to be similar to the input character in this comparison as a candidate character, reads the candidate character data onto the RAM 2 in the order of the candidate in accordance with the degree of similarity,
The characters determined to be the first candidates are displayed as recognized characters. The number of candidate characters is predetermined, for example, within 10 characters.

FIG. 2 shows an example of a display screen when a handwritten character is input in the handwritten character input device shown in FIG.
This handwritten character input device has a plurality of application modes, and the example of FIG. 2 is a case where the telephone directory mode is selected. By inputting a plurality of personal names and corresponding telephone numbers using handwritten characters, the user can store those data and create a telephone directory.

The handwritten character input board 11 is a screen for inputting handwritten characters and the like.
The user writes handwritten characters in the respective frames of the input area 12 with a touch pen or the like, and inputs characters sequentially. When a character handwritten in the input area 12 starts writing a character in a frame next to the character, when no handwritten character is input for a certain period of time, or when the recognition button display 16 in the operation area 15 is touched Will be recognized. The recognized characters are sequentially displayed in the recognized character display area 13 as first candidate characters. If a character displayed in the recognized character display area 13 is different from a desired character, for example, when the character is touched with a touch pen or the like, the character is inverted (for example, when a black character is written in a white frame, (The white characters in a black frame are displayed.) In the candidate character display area 14, characters below the second candidate character similar to the character are displayed. When a desired character is touched from among the candidate characters in the display area 14, the character becomes the first candidate, and is replaced with the inverted character in the recognized character display area 13, and the desired character is displayed in the recognized character display area 13. it can.

When the adoption button display 17 is touched among the push button displays in the operation area 15, the character string in the recognition character display area 13 is adopted, and the screen moves to the application screen 18,
It is displayed in a predetermined area. At this point, the handwritten character input board 11 is in an initial state, and the user repeats the above-described operation to sequentially input handwritten characters. Further, by touching the board switching button display 20, the display screen becomes a normal keyboard input display screen. By touching the displayed keyboard characters, it is possible to input English characters / numeric characters.

FIG. 3 is a diagram showing an example of the data structure of candidate character data in the handwritten character input device shown in FIG. Normally, when a handwritten character is input, the handwritten character input device determines the character characteristics such as the stroke order, the number of strokes, the directionality of the line segment, and the positional relationship between the line segments of the input handwritten character.
A plurality of candidate character data 30 similar to the character from M2
Is retrieved and read. The ROM 2 previously stores a plurality of data representing the characteristics and corresponding candidate character data. The read plurality of candidate character data are stored as candidate character information 31 in the candidate order based on the degree of similarity to a predetermined area of the RAM 3. Each candidate character data 30 is composed of 3 bytes of data. The first 1 byte is character type data 33,
Represents character types such as half-width characters, alphabets / numbers / symbols / katakana, and uppercase / lowercase, and the next two bytes represent the character code. The candidate character number data 32 indicates the number of candidate characters, and is 1 to 10 in this embodiment.

FIG. 4 is a diagram showing an example of the data structure of the candidate character portion 35 in the handwritten character input device shown in FIG. When handwritten characters are continuously input to the handwritten character input device, first candidate characters similar to the characters are sequentially displayed in the recognized character display area 13 shown in FIG. In this case, the above-mentioned candidate character information 31 corresponding to each of the input characters is read, and constitutes the candidate character portion 35. The character number data 36 indicates the number of characters in the recognized character display area 13, that is, the number of candidate character information 31. This candidate character part 3
5 is set in a predetermined area of the RAM 3. The break of each candidate information 31 is determined by the candidate character number data 32 composed of 1 byte shown in FIG.

FIG. 5 is a diagram showing an example of the data structure of the learning information section 38 in the handwritten character input device shown in FIG. When the user inputs a handwritten character into the handwritten character input device and the recognized first candidate character is different from the desired character due to the habit of the user or the like, the desired character is selected from the second and lower characters. When the user again inputs the same character by handwriting with the same habit, the previously selected character needs to be recognized as the first candidate character in order to improve the recognition rate of the character. Therefore, the learning information section 38 shown in FIG. 5 is provided, and even if the previously input handwritten character is not recognized as the first candidate character, when the character is input again under the same condition, the previously selected character is given the highest priority and the first candidate character is given the highest priority. A learning function for characters is realized.

The data structure of the learning information section 38 is shown in FIG.
This is the same as the candidate character part 35 indicated by. That is, each learning information 39 is composed of a plurality of candidate character data 30 shown in FIG. When a user inputs a handwritten character to the handwritten character input device and a candidate character similar to the character is finally adopted, candidate character information 31 having the candidate character as a first candidate is sequentially learned as learning information 39. It is stored in the information section 38. The learning character number data 40 indicates the number of learning information 39 currently stored. The learning information section 38 is stored in a predetermined area on the RAM 3.

The relationship in the processing between the candidate character section 35 and the learning information section 38 will be described with reference to FIG. For example,
Input area 1 of the handwritten character input device
2. Handwriting and input. At this time, the learning character number data 40 is set to “0”, and the learning information section 38 stores the learning information 39.
Is not stored.

Next, the handwritten character input device reads candidate character data 30 similar to the input character "RE" from the ROM 2, creates candidate character information 31, and stores it as a variable E in a predetermined area of the RAM 3. I do. However, the character “re” written by the user in the input area 12 has a habit, and the handwritten character input device determines this “re” as “wa”, and the head area 50 of the candidate character information 31 The data of "wa" is stored as one candidate character, and the data areas of "ne" and "re" are sequentially stored in the data areas of the second candidate character and the third candidate character, respectively. The data of the first candidate character "wa" is stored in a predetermined area of the RAM 3 and is stored in a variable C
Is displayed in the recognized character display area 13. The candidate character information 31 is stored as a variable D in the head area 51 of the candidate character part 35.

Next, the character "wa" is used as a first candidate character.
Although it is displayed in the recognition character display area 13, its first
When the candidate character is not the desired character, when the first candidate character “W” in the recognized character display area 13 is touched as described above, the character is inverted, and the candidate character portion 35 is displayed in the candidate character display area 14. Candidate character information 3 for stored character "wa"
Based on 1, the characters “ne” and “re” below the second candidate are displayed.

Next, when a desired character "re" is touched among the characters displayed in the candidate character display area 14, the candidate character information 31 of the candidate character portion 35 changes to candidate character information 31a, and the character “RE” becomes the new first candidate character, the previous first candidate character “Wa” shifts to the new second candidate character,
The previous second candidate character "ne" is shifted to a new third candidate character. At the same time, the variable C changes from the previous first candidate character “wa” to “re”, and the character displayed in the recognized character display area 13 is changed from “wa” to “re”.

Next, the character “RE” is displayed in the recognized character display area 13.
Is displayed, the input process is terminated. Normally, a plurality of characters are input, but for simplification of description, only one character is input and the process is terminated. By touching the adoption display button 17, the adopted character “RE” is moved from the recognized character display area 13 to the application screen 18 (see FIG. 2) and displayed, and the candidate information 31a of the candidate character portion 35 is displayed.
The learning information 39 is stored in the head area 52 of the learning information section 38 as the learning information 39a. At the same time, the contents of the variables E, C and D are all cleared.

Next, when the user inputs the same character "RE" by handwriting with the same habit as the previous time, the handwritten character input device outputs candidate character data similar to the "RE" in the learning information section 38. It is checked whether or not the information 39 is stored as the head area 53, that is, the first candidate character data. When there are a plurality of pieces of learning information, the learning information 39a, 39b, 39
.. are sequentially searched. In this case, since the candidate character data of the character “RE” is stored in the head area 53 of the learning information 39a, the learning information 39a becomes the candidate character information 31 and is newly stored as a variable E. Therefore, the first candidate character “RE” of the candidate character information 31 is displayed in the recognized character display area 13. Here, the character "re"
If there is no candidate character data similar to in the leading area 53 of the learning information 39, the candidate character data 30 is read from the ROM 2 as described above.

Thus, the handwritten character input device recognizes the character "wa" due to the user's handwriting habit when the user inputs the character "re" by handwriting. Depending on the function, the character "RE" can be determined.

As described above, the case where one handwritten character is input has been described. However, when a plurality of handwritten characters are input, candidate character information 31 similar to the handwritten character is added to the candidate area 35 in the candidate character group 35. The first candidate characters of the plurality of candidate character information 31 are sequentially stored in the variable C and displayed in the recognized character display area 13.

When the plurality of first candidate character data stored in the variable C is adopted by touching the adoption button display 17, the learning information is stored in the plurality of candidate character information 31 stored in the candidate character portion 35. Feedback to the unit 38 is provided.

At this time, the handwritten character input device determines whether or not the candidate character information 31 of the adopted character candidate character portion 35 is in the learning information 39 of the learning information portion 38. In this case, regardless of the order of the arrangement of the candidate character data in the adopted candidate character information 31, it is determined whether or not the candidate character data in the candidate character information 31 and the candidate character data in the learning information 39 all match. to decide. As a result of the determination, if the matching learning information 39 is present in the learning information section 38, the handwritten character input device deletes the learning information 39 and places the adopted candidate character information 31 at the head position of the learning information section 38. Register as new learning information 39. At this time, if the matching candidate character information 31 is not in the learning information section 38, the leading area 52 of the learning information section 38 is used as new learning information 39.
And shifts the position of the subsequent learning information 39 one by one. Also, in this case, when there is no free space in the memory capacity of the learning information section 38, the handwritten character input device deletes the learning information 39 in the tail area of the learning information section 38 and replaces the new learning information 39 with the head position. To be stored. By repeating such processing, feedback to the learning information unit 38 is performed for all the adopted character strings.

Although the contents described above are the basic contents of the learning for the handwritten characters input to the handwritten character input device, in order to further improve the recognition rate for the handwritten characters, as previously described, it is necessary to input the previously input characters. It is necessary to determine the first candidate of the next input handwritten character based on the type of the handwritten character. For example, a handwritten character is input to the handwritten character input device, and a recognized character display area 1 is input.
3 and if the desired character is different, the selected character is selected from the candidate character display area 14 and the desired character is displayed in the recognized character display area 13. In this case, the handwritten character input device
The displayed character is stored as a first candidate and the character type of the character is stored. If there is the same character type among candidate characters similar to the next input handwritten character, it is set as the first candidate. At this time, when there are two or more candidate characters, the candidate character lower in the candidate order is the first candidate.

In the case where handwritten characters of the same character type are continuously input, a function for setting the character type is provided in the handwritten character input device.
The handwritten character input device determines the first candidate character based on only this character type, and prohibits the learning function or the processing based on the character type of the previously input handwritten character.

FIG. 7 is a flowchart showing the processing contents in the handwritten character input device shown in FIG. In step a1, the process is started. In this step, if the character type is specified, the data is stored in variable A
Is stored in a predetermined area of the RAM 3. The variable A has the same data structure as the character type data 33 shown in FIG. 3, and the maximum bit (MSB) 34 specifies the character type, and when specified, its value is “1”. The remaining 7 bits other than the maximum bit 34 indicate the character type, and the character type is determined based on the data. Therefore, if no character type is specified, the value of the variable A is 00h (hexadecimal).

In step a2, it is determined whether a handwritten character has been input. If handwritten characters are entered,
The process proceeds to step a3 where candidate character data similar to the character is read. At this time, as described above, the first candidate character data in the head area 52 of each learning information 39 of the learning information section 38 is searched, and if there is similar candidate character data, the learning information 39 including the candidate character data is stored in a variable. If there is no similar candidate character data, the candidate character data is retrieved and read from the ROM 2 and the candidate character information 31 arranged in the candidate order is stored as a variable E. In step a4, the value of the variable A is determined. When the value of the variable A is not 0, that is, when there is a designation of a character type, in step a5,
If there is a candidate character of the same character type as the designated character type in the candidate character information 31 that is the variable E, the candidate character is determined as the first candidate character, and the process proceeds to step a9.

When the value of the variable A is 0 in step a4, that is, when there is no designation of the character type, the process proceeds to step a6. In step a6, the value of the variable B is determined. Variable B is the character type data of the previous handwritten character.
The initial value stored in a predetermined area of the RAM 3 is 00
h. In step a6, the value of the variable B is 0
At 0h, that is, when the character type of the previous handwritten character is not stored, the first candidate character data in the head area of the candidate character information 31 that is the variable E is directly transferred to step a8 as the first candidate character data.

In step a6, when the value of the variable B is not 00h, that is, when the character type data of the previous handwritten character is stored in the variable B, candidate character data having the same character type data as the character type data is stored. The search is performed from the candidate character information 31 that is the variable E, and if the search is successful, the candidate character data is set as the first candidate character. If the candidate character data cannot be searched, the first candidate character data in the head area of the candidate character information 31 as the variable E is used as the first candidate character data as it is.

In step a8, the character type data of the determined first candidate character data is stored as a variable B. In step a9, the determined first candidate character data is added to and stored in the variable C storing the character string in the recognized character display area 13, and the candidate character information 3 including the first candidate character data is stored.
1 is added to the variable D shown in FIG. 4 and stored. Variable C
Alternatively, when adding data to the variable D, if the storage capacity of the variable C or the variable D exceeds a predetermined storage capacity, the data is not added.

Next, the process proceeds from step a9 to step a10. If there is no touch of the push button display in the operation area 15 in step a10, the process proceeds to step a11.

Next, in step a11, when the character displayed in the recognized character display area 13 is touched with a touch pen or the like, the process proceeds to step a12, the character is displayed in reverse video, and the process returns to step a2. In this case, the second and lower characters of the touched characters are displayed in the candidate character display area 14 based on the candidate character information 31 stored in the variable D.

Further, in the above-described state of step a12, when a desired candidate character is touched from among the characters displayed in the candidate character display area 14 in step a13, a step is performed to make the touched character a first candidate character. a14
In, the touched character and the highlighted character are switched, and in step a15, the first candidate character is determined. In step a16, the character type data of the first candidate character data is stored in the variable B. In step a17, the first candidate character data is stored in the variable C. Based on the replacement of the first candidate characters in step a14, the order of the arrangement of the candidate character data of the candidate character information 31 of the variable D is changed. Return to a2.

When there is no touch of the display character in step all, the process returns to step a2. Step a
In step 13, when there is no touch on the displayed candidate character, the process proceeds to step a20.

At step a10, each operation button display 1
When 5 is touched, processing is performed in accordance with the touched operation button display. For example, when the adoption button display 17 is touched, the data of the variable C is passed to the application program, and the process proceeds to step a18. That is, the characters displayed in the recognized character display area 13 are displayed on the application screen 18. Next, in step a19, feedback to the above-described learning information section 38 is performed, that is, learning is performed on all the characters displayed and adopted in the recognition display area 13. In step a20, if there is another input other than the inputs in step a2, step a10, and step a11, the input processing is performed. If there is no other input processing in step a20, the process returns to step a2.

[0050]

As described above, according to the present invention, when a handwritten character is inputted from the input means and an undesired character is displayed as the first candidate character on the display means, the selected second and lower characters are displayed. Is created and stored as a first candidate character data group, so that when the same handwritten character is input again from the input means, the created new candidate character data group is searched for, so that the desired Characters can be displayed on the display means as first candidates.

According to the present invention, when continuously inputting the same type of handwritten characters, the display means displays the character designated by the designating means or the character type stored at the time of the previous handwritten character input. Since the first candidate character is selected and displayed based on the type data stored in the type storage means, a desired character can be displayed.

Therefore, it is possible to obtain a handwritten character input device which can improve the recognition rate of input handwritten characters and can easily perform an input operation.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic electrical configuration of a handwritten character input device according to one embodiment of the present invention.

FIG. 2 is a diagram illustrating an example of a display screen when a handwritten character is input in the handwritten character input device illustrated in FIG. 1;

FIG. 3 is a diagram showing an example of a data structure of candidate character information 31 in the handwritten character input device shown in FIG.

FIG. 4 is a diagram showing an example of a data structure of a candidate character portion 35 in the handwritten character input device shown in FIG.

FIG. 5 is a diagram illustrating an example of a data structure of a learning information unit in the handwritten character input device illustrated in FIG. 1;

FIG. 6 is a diagram for explaining processing contents of a learning information unit 38;

FIG. 7 is a flowchart showing processing contents of the handwritten character input device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 CPU 2 ROM 3 RAM 4 i / OLSi 5 Keyboard 6 Display part 7 Display memory 8 Tablet 9 Control circuit 10 A / D converter 11 Handwritten character input board 12 Input area 13 Recognized character display area 14 Candidate character display area 15 Operation Area 16 Recognition button display 17 Adoption button display 18 Application screen 20 Board switching button display 30 Candidate character data 31 Candidate character information 33 Character type data 35 Candidate character part 38 Learning information part 39 Learning information

Continuation of the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G06K 9/03 G06K 9/68

Claims (2)

(57) [Claims] 1. An input means having an input surface on which a plurality of coordinate points are set, and outputting character data as a set of coordinate points representing handwritten characters drawn by a writing instrument on the input surface; Candidate character storage means for storing candidate character data for recognizing a character as a character, and search means for searching for candidate character data approximate to the character data and creating a candidate character data group arranged in the order of candidates from the first candidate. Recognition character storage means for sequentially storing candidate character data groups created by the search means; display means for displaying candidate character data groups stored in the recognition character storage means in candidate order; and candidate characters displayed on the display means. Character selecting means for selecting a desired character to be adopted from among the characters, and a character selected by the character selecting means as a new first character.
Learning means for sequentially changing the arrangement of candidate character data groups stored in the recognition character storage means as candidate characters and creating new candidate character data groups, and sequentially adding new candidate character data groups created by the learning means A learning memory for storing, after the learning means creates a new candidate character data group and stores it in the learning memory, first searches for a candidate character data group in the learning memory, A handwritten character input device for retrieving a candidate character data group in the candidate character storage means when there is no corresponding character. 2. A display device comprising: a designation unit for designating a type of a character to be input; and a character type storage unit for storing type data of a candidate character selected at the time of a previous handwritten character input. If the type specified by the character type is not specified by the specifying unit, the candidate character data group stored in the recognized character storage unit is the same as the candidate character data group stored in the recognized character storage unit based on the type data stored in the character type storage unit. 2. The handwritten character input device according to claim 1, wherein when there is candidate character data of the type data, the candidate character data is displayed as a first candidate.