JPH0816727A - Document preparing device and dictionary registering method - Google Patents

Abstract

PURPOSE:To efficiently prepare a personal character recognition dictionary by easily possessing the code data of registering object characters at the document preparing device in a handwrite input system. CONSTITUTION:In document preparing processing, a character at the position of a cursor is detected and its code data are stored in a dictionary preparation object code buffer 364. At the time of dicationary preparing processing, the character at the cursor position is set as the registering object character, the code data stored in the buffer 364 are read out, and these data are registered on a personal character recognition dictionary 52 as personal dictionary data together with the feature data of a handwritten character corresponding to the registering object character.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a handwriting input type document creating apparatus for recognizing characters input by handwriting and creating a document. The present invention relates to a document creation device having the character recognition dictionary and the registration method of the personal character recognition dictionary.

[0002]

2. Description of the Related Art A device in which a display device and a coordinate input device are laminated and integrated makes it possible to input characters and figures on paper as if they were being written, and has recently come to be used in various fields. Along with this, a device that used a keyboard as an input device to create text by inputting characters, etc., also entered that information in the form of coordinate point sequence information using a stylus pen on the LCD tablet. Documents have come to be obtained by recognizing them as characters.

However, the recognition accuracy of characters is not yet practical, and research is ongoing to improve the recognition accuracy. As one of the means for improving the recognition accuracy, there is a method of using a personal character recognition dictionary in which a cursive character for writing for each individual is registered in addition to the existing standard character recognition dictionary. Data is registered in the standard character recognition dictionary so that it can be recognized by any user. Therefore, a certain recognition rate can be maintained on average. On the other hand, since it often does not support personalized peculiar characters, it is highly possible that the recognition rate will be extremely poor for some people. A personal character recognition dictionary is effective for solving such a problem.

[0004]

By the way, in order to use a personal dictionary, a tool for creating a personal character recognition dictionary is indispensable. When creating a personal character recognition dictionary, in addition to the characteristic data used for recognition processing such as coordinates, information indicating what kind of character the character is currently registered, that is, code data of the character to be registered is also required. Come on.

Conventionally, as a method of designating code data of a character to be registered, there are a method of inputting a division code defined by JIS and a method of selecting a character from a list display screen. Further, as disclosed in JP-A-1-103787, a method of inputting a character to be registered by handwriting and using code data of a character pattern obtained as a recognition result of the character as the code data of the character, to be registered By inputting characters by handwriting and selecting the desired character from each candidate character obtained as the recognition result of that character, the method of using the code data of that character as the code data of that character, the key input of the character you want to register However, there is a method in which a desired character is selected from the candidate characters obtained as the kana-kanji conversion result and the code data of the character is used as the code data of the character.

However, these methods are, for example, J.
If it is a method of inputting the kuten code defined by IS, it is necessary to look at the correspondence table each time, and if it is a method to select characters from the list display screen, Both have very troublesome problems, such as having to call the list display screen each time.

Further, in various methods as disclosed in Japanese Patent Laid-Open No. Hei 1-103787, it is necessary to enter characters to be registered by handwriting or key input regardless of document creation. There was also a very troublesome problem as well.

The present invention has been made in consideration of such circumstances, and a document creating apparatus and a document creating apparatus capable of easily obtaining code data of a character to be registered and efficiently creating a personal character recognition dictionary. The purpose is to provide a dictionary registration method.

[0009]

A document creation apparatus of the present invention recognizes a character input by handwriting using a character recognition dictionary as code data, and a document data based on the recognition result of this recognition means. A document creating means to be created, a predetermined character in the document data created by the document creating means as a registration target character, a code holding means for holding the code data, and a handwritten character corresponding to the predetermined character. Data input means for newly inputting the characteristic data, the code data held by the code holding means, and the characteristic data input by the data input means are used as personal dictionary data for the characters to be registered. The dictionary registration means for registering in the recognition dictionary is provided.

[0010]

According to the above construction, by designating a character in the document data with, for example, a cursor, the designated character can be set as a character to be registered. In this case, the code data is held together with the designation of the character, and by newly inputting the feature data of the handwritten character corresponding to the character, the code data and the feature data are used as personal dictionary data for the registration target character. It is registered in the recognition dictionary.

[0011]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 2 is a diagram schematically showing an external configuration of a handwriting input type document creating apparatus. This device
As shown in FIG. 2, a coordinate input device including a transparent tablet 1 and a stylus pen 2 for designating coordinates on the transparent tablet 1 and information on a two-dimensional coordinate point sequence obtained from the coordinate input device. A control device 3 mainly including a microprocessor that performs a process of cutting out one character based on the character recognition, a display device for displaying the coordinate information input from the stylus pen 2 and the character recognized, such as a liquid crystal display 4, and a control device. The external storage device 5 is accessed when the device 3 performs character recognition.

As the display device, a plasma display or the like can be used in addition to the liquid crystal display 4.
The liquid crystal display 4 is laminated and integrated with the transparent tablet 1. That is, the transparent tablet 1 laminated and integrated with the liquid crystal display 4 forms the same coordinate plane of the same size, and the information displayed on the liquid crystal display 4 can be visually recognized through the transparent tablet 1. ing. The transparent tablet 1 thus laminated and integrated
With the liquid crystal display 4, the coordinate pointing position on the transparent tablet 1 is displayed as information at the same position on the liquid crystal display 4, so that information can be input, for example, as if drawing characters / figures on paper. It has become.

FIG. 1 is a block diagram showing the internal structure of this apparatus. As shown in FIG. 1, the control device 3 includes an initial setting unit 31, an input unit 32, a character recognition unit 33, a personal dictionary creation unit 34, a display control unit 35, and a storage unit 36.

The storage unit 36 character-recognizes the data of the input data buffer 361 for temporarily storing the information of the coordinate point sequence input from the coordinate input device 1 and the data of the input data buffer 361 by using the character recognition unit 33. An input box code buffer 362 for holding the resulting character code for each input box, a character code data buffer 363 for storing the character code on the document creation screen described later together with the display position, and a personal dictionary A dictionary creation target code buffer 36 for temporarily storing a character code that is a registration target character when the creation unit 34 creates personal recognition dictionary data.
4, a display data buffer 365 for temporarily storing the data to be displayed on the liquid crystal display 4, an initial screen to be displayed when the system is initialized, and an image of the initial screen when the personal dictionary is created Screen image data 366, a screen area table 367 storing coordinate information of input frames on the screen, buttons and the like.

The initial setting section 31 is used for creating a document of this apparatus.
Initialization processing such as displaying an initial screen when registering a personal dictionary and clearing various buffers and counters is performed. The input unit 32 stores the coordinate information input from the coordinate input device and performs a process of determining an operation from the coordinate information. The character recognition unit 33 refers to the standard character recognition dictionary 51 and the personal character recognition dictionary 52 of the external storage device 5 to recognize the coordinate information stored in the input data buffer 361 as a character. . Personal dictionary creation unit 34
Is a process for creating the personal character recognition dictionary 52. The display control unit 35 controls the process of outputting / erasing the data stored in the storage unit 36 and the result recognized by the character recognition unit 33 to / from the liquid crystal display 4.

The external storage device 5 stores a standard character recognition dictionary 51 and a personal character recognition dictionary 52 that are referred to when the character recognition unit 33 recognizes characters. here,
The standard character recognition dictionary 51 is a standard dictionary created at the time of manufacturing the device. The personal character recognition dictionary 52 is a dictionary for a user who inputs his / her cursive characters one by one. As shown in FIG. 5, both dictionaries store feature data including code data of registered characters, number of strokes, and coordinate data.

FIG. 3 is a diagram showing the structure of the document creation screen of the embodiment. The display screen during document creation is divided into an input area 60 for inputting or instructing data and a display area 64 for displaying document data. Input area 6
At 0, an input frame 61 for inputting handwritten characters, a recognition button 62 for instructing character recognition, a write button 63 for instructing writing (display) of the recognition result, and a dictionary creation of the present invention. A dictionary creation button 66 for instructing is provided. In the display area 64, the cursor 6
5 is provided, and a character (character pattern) obtained as a recognition result is displayed at the position of the cursor 65.

FIG. 4 is a diagram showing the structure of the dictionary registration screen of the embodiment. When the dictionary creation button 66 is pressed on the document creation screen shown in FIG. 3, a dictionary registration screen as shown in FIG. 4 is displayed. The dictionary registration screen is provided with an input frame 71 for inputting handwritten characters for registration, a cancel button 72 for instructing cancellation of registration, and a registration button 73 for instructing registration.

FIG. 5 is a diagram showing the structure of the character recognition dictionary of the embodiment. Dictionary data as shown in FIG. 5 is registered in the standard character recognition dictionary 51 and the personal character recognition dictionary 52 stored in the external storage device 5 of FIG. This dictionary data consists of character code data and characteristic data. The feature data is data indicating the feature of the character, and is composed of the number of strokes and the start point and end point coordinate data for each stroke (relative coordinate data based on the start point of the first stroke).

FIG. 6 is a diagram showing the structure of the character code data buffer of the embodiment. In the character code data buffer 363 shown in FIG. 1, the character code of the character displayed on the document creation screen is stored together with the line digit data indicating the display position. For example, (1, 1) in the figure indicates the first line and the first digit, and the code data of the character "now" displayed at that position is the character code data buffer 363.
It is stored in.

Next, the operation of the embodiment will be described. In the document creation screen shown in FIG. 3, when the handwriting character is input to the input frame 61 using the stylus pen 2 as shown in FIG. 3A and the recognition button 62 is pressed, the character in the input frame 61 is recognized. To be done. Further, when the write button 63 is pressed, the result of character recognition is written at the position of the cursor 65 in the display area 64.

For example, if the character "now" is handwritten in the input frame 61 and the recognition button 62 is pressed, the handwritten character "now" is recognized, and the write button 63 is pressed to display the result "now". The character “” is displayed at the position of the cursor 65 in the display area 64.

Here, when the input person creates a personal character recognition dictionary, by operating the stylus pen 2 in the display area 64 as shown in FIG. After moving the cursor 65 to, the dictionary creation button 66 is pressed. As a result, the dictionary registration screen as shown in FIG. 4 is displayed, and the character designated by the cursor 65 is automatically set as the character to be registered. In this dictionary registration screen, when a personal character corresponding to the character to be registered is handwritten in the input frame 71, the feature data of the handwritten character and the code data of the character designated by the cursor 65 are used as personal dictionary data for personal use. It will be registered in the character recognition dictionary 52.

For example, when the document data "Today is weather" is displayed in the display area 64,
When the dictionary creation button 66 is pressed after setting the cursor 65 to the character "ha", a dictionary registration screen in which the character "ha" is registered as a character to be registered is displayed as shown in FIG. On this dictionary registration screen, by drawing a handwritten character corresponding to the registration target character “ha” in the input frame 71, the characteristic data of the handwritten character (the number of strokes and coordinates of the handwritten character “ha”) is input, and “ha Is registered in the personal character recognition dictionary 52 together with the character code data of ".

Next, a specific processing operation for realizing the above dictionary registration will be described. 7 and 8 are flowcharts showing the operation of the embodiment. At the beginning of the process, the initial setting unit 31 clears various buffers and counters and displays the image data of the screen image data 366 (step 501).

When the input person uses the stylus pen 2 to input characters, push a button, or move the cursor (step 502), the input unit 32 in the control device 3 displays the coordinate point sequence input by the stylus pen 2. By checking the values in the screen area table 367, it is checked whether or not the input person is performing the document creation processing (step 50).
3). In this case, the document creation processing is the dictionary creation button 6
This is processing performed by an operation other than pressing 6.

Here, if the input person is performing the document creation processing (Yes in step 503), the input unit 32 is used.
Checks specifically what kind of processing is being performed.
For example, when the stylus pen 2 is located at the position of the cursor 65 in the display area 64, the input unit 32 determines that the cursor has moved, and moves the processing to the display area 64 (Yes in step 504).
s). As a result, the display control unit 35 moves the position of the cursor 65 according to the operation of the stylus pen 2 (step 505a), and changes the code data of the character displayed at the new cursor position to the character code data buffer 36.
3 is obtained and stored in the dictionary creation target code buffer 364 (step 505). Here, in the character code buffer 363, as shown in FIG. 6, the row and column numbers of the display position and the code data of the characters existing therein are stored.

If the stylus pen 2 is in the input frame 61 of the input area 60, the input section 32 determines that the character is input (Yes in step 506). At the time of character input, the display control unit 35 displays the handwriting of the stylus pen 2 on the display device 4 via the display data buffer 365 (step 5).
07), and further, the input unit 32 uses the input data buffer 361.
The coordinate data is stored in (step 508). here,
The input data buffer 361 stores X and Y coordinate values of input data for each input frame 61.

If the stylus pen 2 is present on the recognition button 62 in the input area 60, the input section 32 determines that it is recognition processing, and moves the processing to the character recognition section 33 (Yes in step 509). As a result, the character recognition unit 33 causes the input frames 6
Characters are extracted one by one based on the input data buffer 361 for each one, and the following character recognition processing is performed (step 51).
0).

That is, in the character recognition processing, the registered patterns of the standard character recognition dictionary 51 and the personal character recognition dictionary 52 having the same number of strokes as the input data stored in the external storage device 5 are used. First, the two-dimensional coordinate data stored in the input data buffer 361 is converted into the character recognition dictionary 51,
It is converted into the form of relative coordinates with the starting point of the first image stored in 52 as a reference. Next, the relative coordinate information and the distance at each coordinate point of the start point and the end point of one stroke are calculated. Next, the total value of the distances of the respective coordinate points is calculated, and the value obtained by dividing the total value by the number of strokes is calculated as the evaluation value. Then, the character having the smallest evaluation value is set as the recognition candidate.

Such recognition processing is performed for each input frame 61, and when the recognition candidates are calculated for all the input frames 61, the processing moves to the display control unit 35. The display control unit 35 erases all handwritings, that is, handwritten characters displayed in the input frame 61 (step 511), and displays the recognition result of the handwritten characters there (step 512). Furthermore, the display control unit 35
Stores the character recognition result for each input frame 61 in the input frame code buffer 362 (step 513). here,
The input frame code buffer 362 stores the input frame number and the character code data of the recognition result displayed there.

If the stylus pen 2 is located on the write button 63 in the input area 60, the input section 32 determines that the writing is to be performed, and moves the processing to the display control section 35 (step 514).
Yes). As a result, the display control unit 35 writes a character (character) corresponding to the character code data stored in the input frame code data buffer 362 at the position of the cursor 65 in the display area 64, and the code data of the same character is converted into a character. Store in code data buffer 363 (step 5)
15). Next, the display control unit 35 erases the characters in the input frame 61 (step 516), and the input frame code buffer 36
2 is initialized (step 517).

At this time, the display control unit 35 moves the cursor 65 in response to the writing of the character and stores the code data of the character on the cursor 65 in the dictionary creation target code buffer 364 (step 518).

Then, in the above step 503,
When the stylus pen 2 is on the dictionary creation button 66, the input unit 32 determines that the dictionary is created, and moves the process to the personal dictionary creation unit 34 (No in step 503).

When creating a dictionary, the personal dictionary creating section 3
4, first, loads the screen image data 366 of the storage unit 36 via the initial setting unit 31 to switch the display screen from the document creation screen to the dictionary registration screen shown in FIG. 4 (step 519). Next, the personal dictionary creation unit 34
It is checked whether or not the character code data is stored in the dictionary creation target code buffer 364 (step 520).
Normally, the cursor 6 on the document creation screen is displayed immediately before switching screens.
The code data of the character that is above 5 is stored in the dictionary creation target code buffer 364.

If the dictionary creation target code buffer 364 is empty, the dictionary cannot be registered, so the personal dictionary creation unit 34 displays an error message (step 52).
1) Immediately switch to the document creation screen (step 52)
7). If the dictionary creation target code buffer 364 is not empty, the character designated by the cursor 65 is set as the registration target character.

Here, the input person draws a handwritten character corresponding to the character in the input frame 71 provided on the dictionary registration screen with the stylus pen 2 to input the characteristic data (number of strokes and coordinates) of the individual handwritten character. Is performed (step 522). At this time, the registration button 73 is pressed to register the data in the dictionary, and the cancel button 72 is pressed to cancel the registration.

When the cancel button 72 is pressed (No in step 523), the personal dictionary creating section 34 stops the registration process and switches to the document creating screen (step 527).
If the registration button 73 is pressed (step 523)
Yes), the personal dictionary creating unit 34 confirms whether or not the feature data of the handwritten character is input (step 52).
4) If no input has been made yet, an error message is displayed (step 525) and the process returns to step 522.

If the handwritten character feature data is input, the personal dictionary creating unit 34 displays the feature data and the code data currently stored in the dictionary creation target code buffer 364 in the data structure of the recognition dictionary shown in FIG. The data is shaped according to the above, and is registered as personal dictionary data in the personal character recognition dictionary 52 of the external storage device 5 (step 526).

After that, the process proceeds to the initial setting unit 31, which returns the screen to the first document creation screen (step 527) and then returns to step 502. In this way, since the character in the document data displayed on the document creation screen can be designated as the registration target character, the code data of the character can be easily acquired. As a result, personal dictionary data can be efficiently registered.

Next, another embodiment of the present invention will be described. In the above-described embodiment, the cursor 65 is moved on the document creation screen to specify the character at the position of the cursor 65 as the registration target character. However, here, the input recognized last during the document creation is input. Characterized by automatically setting a character as a character to be registered.

That is, when creating the document data such as "Today is the weather" in the example of FIG. 3, for example, when the character "ha" of "Today" is handwritten, the "ha" is recognized. The result is the character to be registered. At this time, if the dictionary creation button 66 is pressed, the dictionary creation process can be performed with the character "ha" as the registration target character.

9 and 10 are flow charts showing the operation of another embodiment. The basic processing flow is shown in FIG.
Since it is the same as FIG. 8 and FIG. 8, only different points will be described here. 7 and FIG. 8 are different from FIG. 7 and FIG. 8 in that a process (step 613a in FIG. 10) of storing the result of the last character recognition in the handwritten character recognition process in the dictionary creation target code buffer 364 is added, and the process of moving the cursor. With the exception of the process of storing the code data of the character on the cursor 65 in the dictionary creation target code buffer 364 (step 505 in FIG. 8), the code data of the character on the cursor 65 in the writing process is targeted for dictionary creation. This is a point except the process of storing in the code buffer 364 (step 518 in FIG. 8).

That is, in the recognition processing of the handwritten character input in the input frame 61 (Yes in step 609),
The display control unit 35 stores the code data obtained as the recognition result in the input frame code buffer 363 (step 613), and further stores the code data of the last character obtained at that time in the dictionary creation target code buffer 364. (Step 613a). In the cursor movement process (Yes in step 604), the display control unit 35 only moves the cursor 65 according to the operation of the stylus pen 2 (step 605), and in the writing process (Yes in step 614). The display controller 35 displays the display area 64.
By only writing the character (character pattern) obtained as the recognition result at the position of the cursor 65 of (step 615 to 617), the input character recognized last during the document creation is always registered as the character to be registered. To be ringed.

As a result, in the dictionary creating process (No in step 603), the personal dictionary creating unit 34 sets the last character as the character to be registered and performs the same process as above (step 620 to 620). 626).

As described above, even when the last input character recognized during document creation is used as the registration target character, the acquisition of character code data can be simplified and the efficiency of creating a personal character recognition dictionary can be improved. Is something that can be done.

The present invention is not limited to the above embodiment. For example, in the above embodiment, the handwritten character recognition processing is performed by matching the distance between the registered character in the dictionary and the input coordinates, but for example, the character is abstracted into a certain basic shape and matching with that shape is performed. Other character recognition methods may be used as long as the evaluation value can be calculated. Further, the data structure of the recognition dictionary may be changed accordingly. In short, the present invention can be implemented with various modifications without departing from the scope of the invention.

[0048]

As described above, according to the present invention, the acquisition of the character code data to be registered in the personal character recognition dictionary is performed by using the buffered code, which is necessary for the conventional registration work. It is possible to minimize the troublesome code input process and to easily register personal cursive characters to improve the character recognition rate. To be

[Brief description of drawings]

FIG. 1 is a block diagram showing an internal configuration of a document creating apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram schematically showing an external structure of the document creation device.

FIG. 3 is a diagram showing a configuration of a document creation screen according to the embodiment.

FIG. 4 is a diagram showing a dictionary registration screen of the same embodiment.

FIG. 5 is a diagram showing a data structure of a character recognition dictionary of the embodiment.

FIG. 6 is a diagram showing a data structure of a character code data buffer of the embodiment.

FIG. 7 is a flowchart for explaining the operation of the embodiment.

FIG. 8 is a flowchart for explaining the operation of the embodiment.

FIG. 9 is a flowchart for explaining the operation as another embodiment of the present invention.

FIG. 10 is a flowchart for explaining the operation as another embodiment of the present invention.

[Explanation of symbols]

1 ... Transparent tablet, 2 ... Stylus pen, 3 ... Control device, 4 ... Liquid crystal display, 5 ... External storage device, 31 ...
Initial setting section, 32 ... Input section, 33 ... Character recognition section, 34 ...
Personal dictionary creation unit, 35 ... Display control unit, 36 ... Storage unit,
361 ... Input data buffer, 362 ... Input frame code buffer, 363 ... Character code data buffer, 364 ...
Dictionary creation target code buffer, 365 ... Display data buffer, 366 ... Screen image data, 367 ... Screen area table, 51 ... Standard character recognition dictionary, 52 ... Personal character recognition dictionary.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kazuo Koyama 1381-1 Shinmachi, Ome-shi, Tokyo Within Toshiba Computer Engineering Co., Ltd. (72) Kenji Suzuki 70 Yanagicho, Saiwai-ku, Kawasaki-shi, Kanagawa Toshiba Corporation Yanagimachi factory

Claims (4)

[Claims] 1. A recognition means for recognizing a character input by handwriting as code data using a character recognition dictionary, a document creation means for creating document data based on a recognition result of the recognition means, and a document creation means by the document creation means. Code holding means for holding a predetermined character in the created document data as a registration target character and the code data, and data input means for newly inputting feature data of handwritten characters corresponding to the predetermined character A dictionary registration unit for registering the code data held by the code holding unit and the characteristic data input by the data input unit in the character recognition dictionary as personal dictionary data for the character to be registered. A document creation device characterized by the above. 2. The code holding means sets a character at a cursor position in the document data as a registration target character,
2. The code data is retained.
Document creation device described. 3. The document creating apparatus according to claim 1, wherein the code holding unit holds an input character recognized last during creation of the document as a character to be registered and holds the code data thereof. 4. A dictionary registration method of a document creation device for recognizing characters input by handwriting as code data using a character recognition dictionary and creating document data based on the recognition result, wherein a predetermined value in the document data is specified. Is the character to be registered,
The code data is held, the feature data of handwritten characters corresponding to the predetermined character is newly input, and the code data and the feature data are used as personal dictionary data for the character to be registered as the character recognition dictionary. A dictionary registration method characterized by being registered in.