JP3534469B2 - Character processor - 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 for inputting characters, for example, a word processor, a personal computer, an electronic notebook, a personal digital assistant and the like.

[0002]

2. Description of the Related Art As a conventional technique, when a handwritten character with / without a frame is input by a pen or a mouse drag, a character can be recognized with a candidate (see FIG. 16). However, in this technique, although the recognition candidate holds the recognition candidate when the corresponding character is unconfirmed, the recognition candidate is deleted at the same time when the character is fixed.

Similarly, when a character is read by an optical character reader, a character can be recognized with a candidate (see FIG. 1).
7). However, in this technique, various editing processes could not be performed while holding the recognition candidates. For example, when a certain character string is copied, the recognition candidate cannot be selected because the recognition candidate is not held for the copied character string.

In Japanese Patent Laid-Open No. 2-156369, a kana-kana conversion is performed by instructing a phrase containing a kanji in a sentence once fixed, and a correct kanji-blurred phrase is selected and corrected from the phrase candidates generated by the kana-kanji conversion. The means for doing so is described. However, the recognition candidate of each character could not be retained. Further, Japanese Patent Laid-Open No. 6-215176 discloses that
It describes a means for improving the correction efficiency by collectively displaying correction candidates when correcting the recognized characters read by the optical character reading device. However, nothing is disclosed about holding the recognition candidates of each character.

[0005]

In the prior art, as described above, in the input by handwritten character, the recognition candidate is deleted at the same time when the character is fixed. Therefore, the recognition error is noticed after the various editing processes are performed in advance. However, there was a problem that it could not be corrected by replacing candidates.

[0006] Further, as described in Japanese Patent Laid-Open No. 2-156369, a bunsetsu containing kanji in a once-established sentence is designated to perform kana-kana conversion, and a bunsetsu candidate generated by further kana-kanji conversion is converted into a correct kanji kanji bunsetsu. It was not possible to retain recognition candidates for each character only by selecting and correcting.

Therefore, in the present invention, since recognition candidates are retained for each character, the recognition candidates are retained even after various editing processes have been performed after the character has been determined. It is possible to select and instruct from among the recognition candidates. Further, even when an editing process such as copying / moving is performed, the recognition candidate is copied / moved to the character of the copying / moving destination, so that a recognition error can be selected from the candidates and corrected. .

In the conventional technique, the character size of a character with a handwriting recognition candidate is changed (half-width conversion / full-width conversion / horizontal double-width / double-width /
When a recognition error is noticed after performing vertical double / four double / double angle cancellation), the recognition candidates could not be replaced because it has already been confirmed. Therefore, the present invention can reflect the change in the character size in response to the change in the character size.
With, it is possible to prevent the font size from being returned to the one before changing the font size when the recognition candidates are replaced .
And is intended .

In the prior art, when the character was read by the optical character reader, it was not possible to perform various editing processes while holding the recognition candidates. JP-A-6-21
The means for increasing the correction efficiency by collectively displaying the correction candidates including the correction candidates when correcting the recognized characters read by the optical character reading device described in 5176 cannot hold the recognition candidates for each character. Therefore, the present invention
When a character is read by an optical character reading device, by holding recognition candidates for each read character, it is possible to select and correct recognition errors from the candidates even after performing various editing processes. For the purpose of
There is .

Further , if the candidates are displayed forever after all the processes such as selection and replacement of the recognition candidates are completed, the display may be unsightly and the operability may be deteriorated.
It Therefore, the present invention is to enable the contents of the candidate buffer to be collectively deleted by an instruction from the user.
Is intended for .

Further, since replacement of the recognition candidates is often completed at the time of saving the document, saving the recognition candidates together saves the storage memory unnecessarily, which is less advantageous for the operator. . for that reason,
The present invention is to automatically batch deletion recognition candidate information stored in the recognition candidate buffer in response to the document storage instruction
Has an aim .

[0012]

The present invention provides a character input means such as a pen or a mouse, a character recognition means for recognizing a character handwritten by the character input means, and a character recognized by the character recognition. In a character processing device including a document buffer stored as a document and an editing unit that performs moving or copying edit processing on the document stored in the document buffer, each character of the document stored in the document buffer A candidate buffer for holding the recognition candidate, means for selecting and instructing a desired candidate from the recognition candidates for each character, and the moving or copying editing process was performed.
In this case, the character processing device is provided with means for holding and displaying recognition candidates for each character.

[0013]

In the above structure, an optical character reader (OCR) is provided as the character input means.
Good .

In the above arrangement, a means for deleting all the recognition candidates stored in the recognition buffer is provided.
Is desirable .

In the above arrangement, it is desirable that the apparatus further comprises means for automatically deleting the recognition candidates stored in the auxiliary buffer when the document is saved.

[0017]

According to the present invention , a character input by the character input means is recognized by the character recognition means, and a recognition candidate generated based on the character recognition means is stored in a candidate buffer for each character to move or move. also it will be displayed recognition candidates against its respective character when performing editing processing <br/> replication.

[0018]

It is also possible to perform character recognition using an optical character reader as the character input means.
Is easier to recognize than handwriting.

The recognition stored in the recognition buffer
When the means for deleting all the candidates is provided , the user can arbitrarily delete the recognition candidates, so that the operation can be performed according to the user's intention and the operability is further improved.

Furthermore, when the document is saved, the auxiliary buffer is automatically
If a means for deleting the recognition candidates stored in the
Since the recognition candidate can be deleted in response to the user's document storage request, the operability is further improved.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram showing the device construction of an embodiment of the present invention. Reference numeral 1 is a central processing unit (CPU) that executes programs, and 2 is an RA that stores various data.
RO for storing control programs and fixed data
M, 4 are printer controllers, 5 are printers, 6 is a hard disk (HD) controller, 7 is a hard disk (HD), 8 is a mouse interface, 9 is a mouse,
10 is a display device, 11 is a display controller, 12 is a pen, 13 is a pen interface, 14 is a keyboard, 1
Reference numeral 5 is a key interface, 16 is a floppy disk (FD), 17 is an FD controller, 18 is an IC card, 19 is an IC card interface, 20 is a pen tablet, 21 is a pen tablet interface, and 22 is an optical character reader (OCR). The interface 23 is an optical character reader (OCR).

HD7, FD16, IC card 18
Are controlled by the above HD controller 6, FD controller 17, and IC card interface 19, respectively, and data can be loaded / saved. Further, the key interface 15 discriminates the key pressed on the keyboard 14 and converts it into the corresponding data. In addition, the mouse interface 8 and the pen interface 13
Determines the coordinates on the screen designated by the mouse 9 and the pen 12, respectively, and converts the coordinates into XY coordinates for output. In addition, the pen tablet interface 21 determines the coordinates input to the above pen tablet 20, and the XY
Convert to coordinates and output. Further, the OCR interface 22 converts the image read by the OCR 21 into bitmap data and outputs it.

2 to 4 show works used in the present invention. [Document buffer] is a storage area that is arranged in the RAM and stores the character data in the document in the format shown in the figure.
[Candidate Separator 1] is a fixed code arranged in the ROM and indicating the start of the candidate information in the format of the example in the [Document Buffer]. [Candidate separator 2] is located in ROM,
It is a fixed code indicating the end of the candidate information in the format shown in the [Document buffer]. The [horizontal double-width code] is a fixed code arranged in the ROM and indicating the horizontal double-width in the format shown in the figure in the [document buffer]. [Vertical double-width code] is a fixed code arranged in the ROM and used in the [Document Buffer] in the same format as the horizontal double-width code in the figure, and indicates a vertical double-width. [4
Double-width code] is placed in ROM and [Document buffer]
It is a fixed code that is used in the same format as the horizontal double-width code in the example shown in the figure and indicates a double-width.

[Character work] is a storage area which is arranged in the RAM and temporarily stores the character code at the cursor position. [Document Buffer Pointer] is located in RAM,
This is a storage area for temporarily storing the pointer of the [document buffer] when processing the [document buffer]. The [document buffer pointer 2] is a storage area which is arranged in the RAM and temporarily stores the pointer of the [document buffer] when processing the [document buffer]. [Candidate table] is R
The storage area is located in the AM and stores the recognition result of the handwritten character or the recognition result of the optical character in the format shown in the figure.
[Table size] is the number of candidates arranged in the ROM and generated for each recognized character.

[Candidate Table No. ] Is placed in the RAM and indicates the number of the character in the [candidate table] to be processed. Is a storage area for storing. The [candidate buffer] is a storage area which is arranged in the RAM and stores the candidate data of the character with a candidate existing in the document in the format shown in the figure. [Candidate buffer pointer] is located in RAM,
This is a storage area for temporarily storing the pointer of the [candidate buffer] when processing the [candidate buffer]. The [candidate buffer pointer 2] is a storage area that is arranged in the RAM and temporarily stores the pointer of the [candidate buffer] when processing the [candidate buffer]. [Candidate work] is RA
It is a storage area that is arranged in M and stores candidate data when examining candidate data in the [candidate buffer] one by one.

The following is a description with reference to the flowcharts of FIGS. For the sake of convenience of explanation, what is supposed to be a pen or a mouse is explained only for the pen, but the same applies to the processing of the mouse.

First, the contents of the [Document Buffer] are displayed on the screen. At this time, the contents of [Candidate Separator 1] to [Candidate Separator 2] in the [Document Buffer] are excluded from the display target, so that the candidates are added. Characters can also be displayed correctly (step S101). Next, the candidate information, that is, [candidate separator 1] [candidate separator 2] is added to the character at the cursor position.
If there is candidate information (step S102), if there is no candidate information, the following candidate display processing is performed.

First, the value between [candidate separator 1] and [candidate separator 2], that is, the address on the [candidate buffer] where the candidate group of characters at the cursor position is stored is input to [candidate buffer pointer] ( Step S10
3), [Table size] in [Candidate buffer pointer]
Then, the end address of the candidate group of characters at the cursor position is input to the [candidate buffer pointer 2] (step S104). The contents on the [candidate buffer] indicated by the [candidate buffer pointer] are input to [candidate work] (step S105), and the contents of [candidate work] are displayed on the candidate window (step S106). The [candidate buffer pointer] is incremented by 2 to advance the pointer to the next candidate (step S107). If the [candidate buffer pointer] has not reached the [candidate buffer pointer 2] (step S108), steps S105 to S107.
By performing the process of, all candidates are displayed.

Subsequently, pen / key related event processing is performed. After resetting the [candidate buffer pointer] to 0 (step S109), an event is waited by a key or a pen (step S110). If any key / pen event occurs, it is checked whether it is pen down (step S111), and if it is not pen down, it is checked whether it is key input (step S112). If it is neither pen down nor key input, it is not a normal input, so the event is waited again (B). The case of key input will be described later. In the case of pen down, pen wait is continued (step S113). After pen-up, if a fixed time does not elapse (step S114), the character string is continuously input, so the event waiting is performed again (B). When a certain time has elapsed after pen-up (step S114), it is checked whether drawing has been performed (step S115).
If the candidate character in the candidate window is touched with the pen (step S124), the following candidate character replacement processing is performed (steps S125 to S127) since the drawing is not performed.

That is, first, the pointed candidate character code is input to [candidate work] (step S12).
5), BIT15 and BIT of the character code at the cursor position
By reflecting the contents of No. 7 in the [candidate work], the full-width / half-width state and the presence / absence of underlining of the [candidate work] content are matched with the character at the cursor position (step S1).
26), the character code at the cursor position on the [document buffer] is replaced with the content of [candidate work] (step S1).
After 27), the display process is performed again (A), whereby the candidate replacement process by the pen point is completed. If it is not the candidate character pen touch (step S124), the event waiting is performed again (B).

When drawing is performed (step S115)
Is a handwritten character input, so perform handwriting recognition processing,
A [candidate table] is generated (step S116). Then, [candidate table NO. ] Is input (step S117), and then [candidate table NO. ] Of the [candidate table] is copied (step S118).
As a result, the candidate group for one character is completed in the [candidate buffer]. Next, by inserting the [candidate separator 1], the value of [candidate buffer pointer], the [candidate separator 2], and the character code of the first candidate at the cursor position on the [document buffer] (step S119), the candidate is attached. The character [Document Buffer] data is completed. Further, the cursor position is updated to the position next to the insertion data (step S
After 120), [table size] is added to the content of [candidate buffer pointer] (step S121) to prepare for the next character candidate group. If it is not the last table (step S122), [candidate table NO. ] Is incremented (step S123) and step S118
~ The process of step S121 is repeated. When the processing is completed up to the last table (step S122), the display processing is performed again (A).

Next, the processing at the time of key input will be described. When the character reading key is input (step S12)
8) The optical character reading device (OCR) is operated (step S129), and after the [candidate table] is generated by the optical character reading recognition process (step S130), the same process as that after handwriting recognition is performed (step S130). C) completes the [document buffer] and the [candidate buffer].

When a character key is input (step S1)
31), the character code is inserted at the cursor position on the [document buffer] (step S132), the cursor position is updated to the position next to the inserted character (step S133), and then the display process is performed again (A).

When the half-width conversion key is input (step S134), BIT1 of the character code contents at the cursor position
After turning ON 5 (step S135), the display processing is performed again (A), so that half-width conversion processing can be performed while retaining the candidate information even for the character with a candidate. When the full-width conversion key is input (step S136), BIT15 of the character code content at the cursor position is turned off (step S137), and then the display process is performed again (A), so that the candidate information is displayed even if it is a character with a candidate. Full-width conversion processing is possible while holding.

When the double-width key is input (step S
138), if the character at the cursor position is any of double-width / double-height / quadruple-size double-width characters (step S13).
9), the horizontal double-width code, the vertical double-width code, or the quadruple-width code is replaced with the horizontal double-width code (step S140).
Then, the display process is performed again (A). If the character at the cursor position is not a double-width type character (step S139), a horizontal double-width code is inserted before the character code (step S14).
After 1), by performing the display process again (A), it is possible to perform the double-width process while retaining the candidate information even for the character with a candidate.

When the double-height key is input (step S
142), if the character at the cursor position is any double-width double-width / vertical-double-width / 4-double-width double-width type character (step S14)
3), the horizontal double-width code, the vertical double-width code, or the quadruple-width code is replaced with the vertical double-width code (step S144).
Then, the display process is performed again (A). Also, if the character at the cursor position is not a double-width type character (step S143), a double-width vertical code is inserted before the character code (step S14).
After 5), by performing the display process again (A), it is possible to perform the double-height process while retaining the candidate information even for the character with a candidate.

When the quadruple-size key is input (step S
146), if the character at the cursor position is any of double-width / double-width / double-width double-width characters (step S14)
7), horizontal double-width code, vertical double-width code, or quadruple-width code is replaced with quadruple-width code (step S148)
Then, the display process is performed again (A). If the character at the cursor position is not a double-width type character (step S147), a quadruple-width code is inserted before the character code (step S14).
After 9), by performing the display process again (A), it is possible to perform the quadruple angle process while retaining the candidate information even for the character with a candidate.

When the double-width cancel key is pressed (step S150), the character at the cursor position is double-width / vertical-double width / 4.
If it is any of the double-width double-width characters (step S15)
1) Delete horizontal double-width code, vertical double-width code, or double-width code from [document buffer] (step S152)
After that, by performing the process again from the display process (A), it is possible to perform the double angle cancellation process while retaining the candidate information even for the character with a candidate. If the character at the cursor position is not a double-width type character (step S151), nothing is done and the next event is waited (B).

When the underlined shading key is pressed (step S153), the BIT of the character code contents at the cursor position
After XORing 7 (step S154), the process is performed again from the display process (A), so that even underlined characters are underlined while the candidate information is retained (underlined characters are underlined without characters). It is possible to convert and to convert underlined and non-texted characters to underlined and hatched characters.

When the candidate replacement key is input (step S155), the following candidate replacement process is performed. First, the value between [candidate separator 1] and [candidate separator 2] at the cursor position on the [document buffer], that is, the address on the [candidate buffer] where the candidate group of characters at the cursor position is stored is [candidate buffer]. Pointer] (step S
156), and the value obtained by adding [table size] to [candidate buffer pointer], that is, the end address of the candidate group of characters at the cursor position is input to [candidate buffer pointer 2] (step S157). Further, the character code at the cursor position is input to [character work], and BIT15 and BIT7 of [character work] contents are turned off (step S15).
8) By doing so, the contents of [text work] are not full-width / underlined. Then, the contents on the [candidate buffer] indicated by the [candidate buffer pointer] are input to [candidate work] (step S159), and it is checked whether the contents of [candidate work] and [character work] match (step S16).
0).

If they do not match (step S160), the [candidate buffer pointer] is incremented by +2, and the pointer is advanced to the next candidate (step S161).
Does not reach the [candidate buffer pointer 2] (step S162), steps S159 to S16.
By performing the process of 1, a search is made for the contents of [candidate work] and [character work] that match. If the [candidate buffer pointer] reaches the [candidate buffer pointer 2] before the [candidate work] and the [character work] do not match (step S162), it is not in a normal state, and therefore the display process is performed again (A). .

When [candidate work] and [character work] match (step S160), the [candidate buffer pointer] is set to +2 in order to obtain the candidate data next to the current candidate.
And move the pointer to the next candidate (step S163).
However, at this time, when the [candidate buffer pointer] reaches the [candidate buffer pointer 2] (step S164), the first candidate data is acquired, so that the [candidate separator 1] at the cursor position on the [document buffer] is selected. The value between [candidate separator 2], that is, the address on the [candidate buffer] where the candidate group of characters at the cursor position is stored is input to [candidate buffer pointer] (step S165).
Then, the contents on the [candidate buffer] indicated by the [candidate buffer pointer] are input to the [candidate work] (step S
166). In addition, B of the character code contents at the cursor position
By reflecting IT15 and BIT7 on the [candidate work] (step S167), the full width of the contents of the [candidate work] /
After matching the half-width state and the presence / absence of underlining with the character at the cursor position, the character code at the cursor position on the [Document Buffer] is replaced with the content of [Candidate work] (step S168), and then the display processing is performed again. By performing (A), the candidate replacement processing by the key is completed.

When the all candidate deletion key is input (step S169), the following all candidate deletion processing is performed. First, the [document buffer pointer] is set to the beginning of a sentence (step S170). Then, it is checked whether the content on the [document buffer] indicated by the [document buffer pointer] is [candidate separator 1] (step S171), and if not [candidate separator 1], the process proceeds to step S176.
If it is [candidate separator 1] (step S171),
The value of [document buffer pointer] is input to [document buffer pointer 2] (step S172). continue,
Until the [document buffer pointer] reaches the position of [candidate separator 2] (step S174), the [document buffer pointer] is incremented by +2 and the pointer is advanced (step S17).
3), when the position of the [candidate separator 2] is reached (step S174), the contents from [document buffer pointer 2] to [document buffer pointer] are deleted from [document buffer] (step S175) to obtain the candidate. Delete the information. Subsequently, the [document buffer pointer] is incremented by +2 and the pointer is advanced to the next character (step S17).
6) The processing of steps S171 to S176 is performed until the [document buffer pointer] reaches the end of the sentence (step S177), and then the display processing is performed again (A) to complete the candidate deletion processing. .

When the save key is input (step S1)
78), all the candidates are deleted (step S1)
After that, the normal document saving process is performed (step S18).
By doing 0), it is possible to automatically delete all candidates when saving.

When the range designation key is input (step S181), the information of [document buffer] corresponding to the designated range is stored. At this time, from [candidate separator 1] to [candidate separator 2] Is stored including the contents of (step S183), and the movement key (step S184) is used to move the storage range (step S185).
After performing the storage range copy process (step S187) with the copy key (step S186), the display process is performed again (A) to move / cancel while retaining the candidate information even if it is a character with a candidate. It is possible to perform copy processing.

Also, the half-width conversion key (step S188), full-width conversion key (step S190), horizontal double-width key (step S192), vertical double-width key (step S194), quadruple-width key (step S196), double-width key after the range is specified. When the release key (step S198) and the underlined key (step S200) are input, the above-described processing is performed for all the characters within the specified range (step S189, step S191, step S193, step S195, step S197, step S199, Step S20
After 1), by performing the display process again (A), each of the above-described processes by specifying the range can be performed while retaining the candidate information even for the character with a candidate.

After the range designation, the keys other than the above are subjected to the key processing (step S202), and then the display processing is performed again (A). Further, each key processing is performed for keys other than the above-mentioned keys before the range is specified (step S18).
After 2), the display process is performed again (A).

By the above processing, moving / copying / changing the character size / editing processing of underlining / shading while retaining the candidate information of the character with a candidate and changing the character size (half-width conversion / full-width conversion / horizontal conversion) Double angle / Vertical double angle / 4 Double angle / Double angle release)
The recognition candidate replacement processing after retaining the character size of the character with recognition candidate after performing the underlining / shading information for the character with the recognition candidate after underlining / shading processing is performed. It is possible to implement the recognition candidate replacement process, the recognition candidate all deletion process, and the process of automatically performing the recognition candidate all deletion process when the document is saved. Examples of displays in the embodiment are shown in FIGS. 12, 13, 14, and 15.

FIG. 12 shows a state in which the copying process is performed while holding the candidate information of the candidate characters, a state in which the horizontal double-width process is performed while retaining the candidate information of the candidate characters, and the candidate information of the candidate characters. 6 is a display example showing a state in which an underline shading process is performed while holding.

FIG. 13 shows a state in which the candidate is replaced by the key while keeping the double-width character state and a state in which the candidate is replaced by the pen point while keeping the double-width character state by the processing procedure of FIG. It is a display example shown.

In FIG. 14, according to the processing procedure of FIG. 3, the candidate is replaced by the key while keeping the underlined character state, and the candidate is replaced by the pen point while keeping the underlined character state. It is a display example which shows a state.

FIG. 15 is a display example showing a state in which all recognition candidate deletion processing has been performed.

[0054]

According to the present invention, even if a recognition error is noticed after a moving or copying edit process is performed on a character with a handwriting candidate, the correction can be made by replacing the candidate and the operability is improved.

[0055]

As a character input means, an optical character reading is used.
When using the sampling device, even if a recognition error is noticed after moving or copying edit processing for input by the optical character reading device, correction by candidate replacement is possible and operability is improved. To do.

Furthermore, the recognition stored in the recognition buffer is performed.
When the means for deleting all the knowledge elements is provided , the recognition candidates can be arbitrarily deleted, so that the operation according to the operator's intention can be performed and the operability is further improved.

Then, when the document is saved, the auxiliary buffer is automatically
If a means for deleting the recognition candidates stored in the
It is because it is avoided saving unnecessary data when the document is saved, further improving the operability.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an information processing apparatus of the present invention.

FIG. 2 is a diagram showing an example of a document buffer according to an embodiment of the present invention.

FIG. 3 is a diagram showing an example of a candidate table and a candidate buffer in one embodiment of the present invention.

FIG. 4 is a diagram showing an example of a structure of a candidate buffer according to an embodiment of the present invention.

FIG. 5 is a diagram showing a flowchart showing a process in one embodiment of the present invention.

FIG. 6 is a diagram showing a flowchart showing a process in one embodiment of the present invention.

FIG. 7 is a diagram showing a flowchart showing a process in one embodiment of the present invention.

FIG. 8 is a diagram showing a flowchart showing a process in one embodiment of the present invention.

FIG. 9 is a diagram showing a flowchart showing a process in one embodiment of the present invention.

FIG. 10 is a diagram showing a flowchart showing a process in one embodiment of the present invention.

FIG. 11 is a diagram showing a flowchart showing a process in one embodiment of the present invention.

FIG. 12 is a view showing a state in which a copying process is performed while retaining candidate information of a candidate character and a state in which horizontal double-width processing is performed while retaining candidate information of a candidate character in the present invention; It is a figure which shows the state which underlined the hatching process, holding information.

FIG. 13 is a diagram showing a state in which a candidate is replaced by a key while holding a double-width character state and a state in which a candidate is replaced by a pen point while holding a double-width character state in the present invention.

FIG. 14 is a diagram showing a state in which a candidate is replaced by a key while holding a state of underlined shaded characters, and a state in which a candidate is replaced by a pen point while holding a state of underlined shaded characters in the present invention. .

FIG. 15 is a diagram showing a state in which all recognition candidate deletion processing has been performed in the present invention.

FIG. 16 is a diagram showing a state where handwriting input with / without a frame is performed in the conventional information processing apparatus.

FIG. 17 is a diagram showing a state in which character reading is performed by an optical character reading device in a conventional information processing device.

[Explanation of symbols]

1 CPU 2 RAM 3 ROM 4 Printer controller 5 printers 6 HD controller 7 HD 8 mouse interface 9 mice 10 Display 11 Display controller 12 pens 13 Pen controller 14 keyboard 15 key interface 16 FD 17 FD controller 18 IC card 19 IC card interface 20 pen tablet 21 Pen tablet interface 22 OCR interface 23 OCR

Claims (4)

(57) [Claims] 1. A character input unit such as a pen or a mouse, a character recognition unit for recognizing a character handwritten by the character input unit, and a document buffer for storing the character recognized by the character recognition as a document. A character processing device having an editing unit for performing moving or copying edit processing on a document stored in the document buffer, in which recognition recognition is held for each character of the document stored in the document buffer. And a candidate buffer for selecting and instructing a desired candidate from the recognition candidates for each character, and means for holding and displaying the recognition candidates for each character when the moving or copying editing process is performed. Character processing device. As claimed in claim 2, wherein said character input means, claims, characterized in that it comprises an optical character reader (OCR)
1. The character processing device according to 1 . 3. A contracting means comprising means for deleting all recognition candidates stored in the recognition buffer.
The character processing device according to claim 1 or 2 . 4. A document automatically said Hou auxiliary character processing apparatus of claim 1, wherein further comprising means for deleting the stored recognition candidate in the buffer during storage.