JP3333075B2 - Document processing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document processing method capable of creating a document such as a Japanese word processor or a personal computer, and more particularly to a document processing method capable of recognizing characters.

[0002]

2. Description of the Related Art Conventionally, as this type of document processing apparatus,
The one described in JP-A-63-301353 is known. In this method, when a character such as a kanji is to be input, the character to be input is divided into a basic layout and a basic character to facilitate input.

[0003] Also, the one described in Japanese Patent Application Laid-Open No. Sho 60-144786 is known. This is to synthesize a pattern when a pattern of a European character is generated. In this combination, the basic information is character layout information and basic character information. Characters are formed and displayed in accordance with the layout information and the basic characters specified by the user.

[0004] As a conventional document processing apparatus of this type, those relating to such a character pattern generation method and input method are known, and these apparatuses store character layout, the number of strokes, stroke information, and the like. Have a recognized dictionary.

As described in Japanese Patent Application Laid-Open No. 5-108874, as a workaround in the event of a failure in character recognition cutout, the user divides the cutout range or cuts out the cutout range together with the next character. There is known a device that selects a character and enables character recognition again.

[0006]

In such a conventional document processing apparatus of this type, the layout of characters, the number of strokes,
It has been considered to newly add information on synthesized characters and divided characters to a recognition dictionary having stroke information and the like.

However, if information such as character layout, the number of strokes, and stroke information is provided for all combinations of combined / divided characters, the amount of information becomes extremely large.

Also, in the apparatus described in Japanese Patent Application Laid-Open No. 5-108874, which enables character recognition again, it is necessary to always hold the read image, so that a large-capacity memory is required. Become. In addition, it is necessary to operate the character recognition again, and there is a disadvantage that the response time and the like become slow.

The present invention has been made in view of such circumstances, and has a relatively small-capacity table for synthesized characters and divided characters, and compares the recognition results of the preceding and following characters with this table. Accordingly, it is possible to provide a document processing method which can improve the recognition rate of character recognition, can be shared with a dictionary for recognition processing, and has an excellent response speed.

More specifically, in character recognition, a cut-out process for extracting data to be recognized is performed, and character recognition is performed on the cut-out data. Certain items may be divided into two characters by the clipping process. Normally, the recognition candidate is displayed only for the extracted character, and therefore, when the extraction fails, the target recognition candidate is not displayed, and the user has to input again.

Therefore, according to the present invention, a combined character recognition dictionary is provided, and it is searched whether or not the preceding and following characters exist in the combined character recognition dictionary. Display and selectable.

Conversely, in the clipping process, two characters may be clipped as one character. Also in this case, the recognition candidate is usually displayed only for the character to be cut out. Therefore, when the extraction fails, the target recognition candidate is not displayed, and it is necessary to input again. Therefore, in the present invention, the corresponding character is searched from the combined character recognition dictionary, and if there is a corresponding character as a divided character, it is displayed as one of the recognition candidates and can be selected.

[0013] Further, normally, when the extraction fails, it is necessary to delete the incorrectly recognized character and re-enter it. In the present invention, when a synthesized character candidate is selected, the corresponding character string is deleted. Characters can be replaced with one selected composite character.

In addition, when a divided character candidate is selected, the corresponding one character can be replaced with a plurality of selected divided characters, so that a recognition candidate can be selected and input even when cutting out fails. And

In the candidate display, the composite character and the divided character candidate are displayed with a mark such as a dot or underline or the color identification display so that the composite character and the divided character candidate can be distinguished at a glance on the candidate screen. Can be performed.

[0016]

SUMMARY OF THE INVENTION The present invention provides a character recognition process.
Storing the characters recognized by the processor in a memory;
Display the characters stored in the document in text format on the display
Along with, other recognition candidate characters of the character indicated by the cursor
Is a document processing method that displays selectable
When the cursor is used to point to the character
Characters and characters before and after the cursor position are stored in the memory
And extract one character and the one character as multiple characters
And easy-to-recognize character combinations
Search the stored character table, and select the extracted cursor
The character at the designated position and the characters before and after the cursor designated position
A character combination that includes
When the character is
The displayed one character is used as the composite character candidate in the display.
B. Another recognition candidate sentence of the character indicated by the cursor
This is a document processing method in which characters are displayed in a selectable manner at the same time. This
According to the invention, the character displayed on the display is
When instructed by the sol, it is combined with the characters that
The combined character candidate that becomes one character by
It is displayed so that it can be selected at the same time as candidate characters.
By displaying the candidate with dots, etc., the currently displayed
At a glance what candidates are
Makes it possible to significantly improve operator's correction work efficiency.
improves. In this document processing method, synthetic character candidates
Is selected, the character at the cursor position and the cursor
Select the character that exists before or after the
It is desirable to replace the
Good. Also, the part corresponding to the character indicated by the cursor
Displaying the composite character candidate so that the
desirable. The present invention also recognizes by character recognition processing.
The recognized character is stored in the memory, and the character is stored in the memory.
Characters are displayed on the display in text format, and
Selectable other recognition candidate characters of the character indicated by sol
A document processing method to be displayed, wherein the displayed characters are
When pointing with a sol, the character at the cursor
Extracted from the memory, one character and one character
Character combinations that are easy to recognize as
Search the character table stored in association with the
If the character at the cursor pointed position exists in the table,
In this case, the above-mentioned stored in association with the corresponding one character
The combination of characters is used as the divided character
B. Another recognition candidate sentence of the character indicated by the cursor
This is a document processing method in which characters are displayed so as to be selectable at the same time. This
According to the invention, the character displayed on the display is
When instructed by the sol, the designated character becomes multiple characters
If the character can be divided and the character candidate is
Characters are displayed so that they can be selected at the same time.
By displaying underline etc., it is currently displayed
Identify at a glance what a candidate is
And the operator's correction work efficiency is greatly improved.
I do. In this document processing method, divided character candidates are selected.
When selected, the character at the cursor position is selected.
Replace with a combination of multiple characters that are candidate character splits
Desirably, it is stored in a memory. Also, divided character candidates
Is displayed on the display so that it can be distinguished from other recognition candidate characters.
Is desirable. According to another aspect, the present invention provides a reading unit 101 for reading a document, a recognition unit 102 for recognizing characters written on the read document, a recognition information storage unit 103 for storing recognized characters, Storage unit 1
The display unit 104 that displays the characters stored in the storage unit 03, a shape table in which characters that are easily recognized as having the same shape are grouped for each shape, and characters that are easily recognized as a single character are associated with each other. Dictionary storage unit 105 storing a combined character recognition dictionary composed of a combined divided character table
And a specific character selected from the characters displayed on the display unit 104 and a character positioned before or after the character by referring to the synthesized character recognition dictionary stored in the dictionary storage unit 105. A combined character candidate generating unit 106 for displaying the combined character as a combined character candidate for the specific character on the display unit 104, and an adding unit for adding an identification code for identifying the combined character candidate to be a combined character. Is a document processing device comprising:

In the present invention, as the reading unit 101, a stationary scanner, a line scanner, or the like can be used. As the display unit 104, various display devices such as a CRT display device and a liquid crystal display device can be used.

As the dictionary storage unit 105, an internal memory such as a ROM, or various external storage devices such as a magnetic disk device or a floppy disk device can be used.

Recognition unit 102 and composite character candidate generation unit 10
As 6, it is convenient to use a microprocessor including a ROM, a RAM, and an I / O port. As the recognition information storage unit 103, usually, a RAM therein is used.

According to the present invention, the characters written on the read document are recognized by the recognition unit 102, stored in the recognition information storage unit 103, and displayed on the display unit 104.

Thereafter, the synthesized character candidate generating unit 106 refers to the synthesized character recognition dictionary stored in the dictionary storage unit 105, and selects a specific character selected from the characters displayed on the display unit 104 and the specific character. A composite character obtained by combining a character located before or after is displayed on the display unit 104 as a composite character candidate of the specific character, and is displayed as a composite character candidate.
Is used to identify the character
Is added.

Therefore, with respect to the synthesized character, a synthesized character candidate is generated with reference to a synthesized character recognition dictionary including a relatively small-capacity shape table and a synthesized divided character table.
Since it is possible to display the synthesized character candidate of the specific character on the display unit 104, the recognition rate of the character recognition can be improved without increasing the dictionary processing capacity for the recognition processing so much.

In the above configuration, a selection unit for selecting an arbitrary candidate from the combination character candidates displayed by the combination character candidate generation unit 106, and a selection unit for selecting a specific character and a character located before or after the character. It is desirable to further comprise an editing unit for replacing the synthesized character candidate selected by the above.

In the case of such a configuration, the display unit 10
4, a specific character and a character located before or after the specific character can be automatically replaced with a composite character candidate.

In this case, the character at the current cursor position and the character before or after are deleted, and the selected composite character is inserted instead. At the time of this insertion, two characters are 1
Since it becomes a character, it is desirable to insert a composite character in full-width.

In the present invention, there is provided an adding section for adding an identification code for identifying a synthesized character to a synthesized character candidate . This identification code adds a mark such as a dot or underline to the synthesized character. Alternatively, when a display device capable of color display is used, a color code may be added to the composite character to display the composite character and the background in color.

When a mark such as a dot or underline is added to a composite character, if the current character corresponds to the character on the left side of the composite character candidate, a black dot is displayed at the upper left of the composite character candidate. When the character corresponds to the character on the right side of the combined character candidate, a black point is preferably displayed at the upper right of the combined character candidate.
With this configuration, it can be distinguished at a glance on the screen as a composite character candidate.

According to still another aspect, the present invention also provides a reading section for reading a document, a recognition section for recognizing characters written on the read document, and a recognition information storage section for storing the recognized characters. A display unit that displays characters stored in the recognition information storage unit, a shape table in which characters that are easily recognized as having the same shape are grouped for each shape, and a character that is easy to recognize a plurality of characters as one character. By referring to the combined character recognition dictionary stored in the dictionary storage unit that stores the combined character recognition dictionary consisting of the combined divided character table and the combined character recognition dictionary stored in the dictionary storage unit, from among the characters displayed on the display unit A divided character candidate generation unit that displays a divided character obtained by dividing the selected specific character into a plurality of characters on a display unit as a divided character candidate of the specific character; A document processing device comprising a adding unit for adding an identification code for identifying the Rukoto.

In the present invention, the reading section includes:
A stationary scanner or a line scanner can be used. Various display devices such as a CRT display device and a liquid crystal display device can be used as the display unit.

As the dictionary storage unit, an internal memory such as a ROM and various external storage devices such as a magnetic disk device and a floppy disk device can be used. ROM, RA,
It is convenient to use a microprocessor including M and I / O ports, and a RAM in the recognition information storage unit is usually used.

According to the present invention, the characters written on the read original are recognized by the recognition unit, stored in the recognition information storage unit, and displayed on the display unit. Thereafter, the divided character candidate generation unit refers to the synthesized character recognition dictionary stored in the dictionary storage unit, and generates a divided character obtained by dividing a specific character selected from among the characters displayed on the display unit into a plurality of characters. Is displayed on the display unit as a divided character candidate of the specific character, and the divided character candidate
An identification code for identifying the presence is added.

Therefore, regarding the divided characters, a divided character candidate is generated with reference to a synthesized character recognition dictionary including a relatively small-capacity shape table and a synthesized divided character table.
Since the divided character candidates of the specific character can be displayed on the display unit, the recognition rate of the character recognition can be improved without increasing the dictionary capacity for the recognition processing so much.

[0033] In the above configuration, a selection unit for selecting an arbitrary candidate from among the divided character candidates displayed by the divided character candidate generation unit, and an editing unit for replacing a specific character with the divided character candidate selected by the selection unit. Is desirably further provided.

In the case of such a configuration, an arbitrary candidate is selected from among the divided character candidates displayed on the display unit, and
A specific character can be automatically replaced with a divided character candidate.

In this case, the character at the current cursor position is deleted, and the selected divided character is inserted instead. At the time of this insertion, one character becomes two characters, but whether to insert it in full-width or half-width differs depending on the application. Therefore, it is desirable to insert according to the half-width / full-width input mode at the time of candidate selection.

In the present invention, there is provided an adding unit for adding an identification code for identifying a divided character to a divided character candidate . This identification code adds a mark such as a dot or underline to the divided character. Alternatively, when a display device capable of color display is used, a color code may be added to the divided characters to display the divided characters and the background in color.

When a mark such as a dot or underline is added to a divided character, if the current character is a divided character candidate, it is added to two characters that are divided character candidates so that the character can be recognized as a divided character candidate. It is preferable to connect the underlines to display. With this configuration, it can be distinguished at a glance on the screen as being a divided character candidate.

[0038]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail based on an embodiment shown in the drawings. Note that the present invention is not limited to this.

FIG. 2 is a block diagram showing the configuration of an embodiment in which the present invention is applied to a Japanese word processor. In the following, a description will be given assuming that basic functions as a Japanese word processor such as a kana-kanji conversion function are already provided.

In this figure, 1 is a CPU (central processing unit) for executing a program, 2 is a RAM for storing various data, 3 is a ROM for storing control programs and fixed data, 4 is a printer, 5 Is a printer controller, 6 is an HD (hard disk), 7 is an HD controller, 8 is a mouse, 9 is a mouse interface, and 10 is C
A display device including an RT display device or a liquid crystal display device; 11, a display controller; 12, a pen;
Is a pen interface, 14 is a keyboard, 15 is a key interface, 16 is an FD (floppy disk device), 17 is an FD controller, 18 is an IC card, 1
9 is an IC card interface, 20 is a pen tablet, 21 is a pen tablet interface, 22 is OC
R (optical character reader), 23 is an OCR interface.

HD6, FD16 and IC card 18
Are controlled by the respective HD controller 7, FD controller 17, and IC card interface 19, and can load / save data. The key interface 15 determines a key pressed on the keyboard 14 and sends data corresponding to the key to the CPU 1.

The mouse interface 9 and the pen interface 13 determine coordinates on the screen designated by the mouse 8 and the pen 12, respectively, and convert the coordinates into XY coordinates for output. The pen tablet interface 21 determines the coordinates input to the pen tablet 20, converts the coordinates into XY coordinates, and outputs the XY coordinates. OCR interface 23
The image read at 22 is converted into bitmap data and output.

The OCR 22 can read an original on which various characters are described. The CPU 1 recognizes characters written on the read document. A sentence buffer for storing the recognized result is secured in the RAM 2.
U1 stores the recognized character in the sentence buffer, and displays the stored contents of the sentence buffer on the display device 10.

FIGS. 3 to 7 are explanatory diagrams showing an example of display in this embodiment. These show correction screens of recognition candidates after the document is read by the OCR 22 in FIG. 2, and the text displayed on each screen indicates the content of the read document.

In these figures, A indicates a display screen of the display device 10. B indicates the position of the cursor, and a recognition candidate D of the character with the cursor B is displayed on the screen A as a window screen C.

If the recognition candidate D is a candidate for a composite character,
A black dot is displayed at the upper left of a character, and an underline is displayed below a character if the character is a candidate for a divided character. If the read character is incorrect, the pen E selects the correct character from the recognition candidates D. F is a reverse display indicating the currently selected character. G is guidance, in which text information such as paper size and character size is displayed.

3 shows a state in which the cursor B is located at the character "(" and candidates for the character "(" are displayed. As candidates for the character "(", "(, ku, <, i, @" Although "I, い" is displayed, the character "(" and the next character ")" at the cursor B may be combined to form "I, ぃ". A dot is displayed at the upper left of “I, ぃ”.

FIG. 4 shows a state in which the cursor B is located at the character ")" and a candidate for the character ")" is displayed. Although "),>, i, ぃ" are displayed as candidates for the character ")",
Since a combination with the previous character “(” may result in “I, ぃ”, a dot is displayed at the upper right of “I, ぃ” as a composite character candidate.

FIG. 5 shows the candidate "I" in the state of FIGS. 3 and 4.
FIG. 9 is a diagram showing a state in which is selected. "("")"
Are combined to form “i”, and new candidates for “i” are “i, ぃ, (), (>, ku), ku
>, <), And <>] are displayed.

The characters "i" are "(" and ")",
"(" And ">", "ku" and ")", "ku" and ">",
In some cases, the character can be divided into “<” and “)” and “<” and “>”, so “(” and “)”, “(”
And “>”, “ku” and “)”, “ku” and “>”, “<” and “)”, and “<” and “>” are underlined.

FIG. 6 shows a state in which "Buddha, Den, I, Mu" is displayed as a candidate for the character "Buddha" of the read document.
Since “Buddha” may be recognized as “Buddha” by combining “I” and “M”, “I” and “M” are underlined and displayed as division candidates.

FIG. 7 is a diagram showing a state in which a pair of candidates "i" and "mu" are selected in the state of FIG. One character of the character "French" indicated by the cursor B in the text disappears, the selected "I" and "M" are displayed, and "I, I, Buddha" is newly displayed as a candidate for "I". I have.

The dots and underlines displayed in these figures are identification marks for indicating the combined character candidates and the divided character candidates, respectively. It is not necessary to use a dot and an underline as the identification mark. Therefore, for example, when the display device 10 is a display device capable of displaying colors, the dots and underlines are displayed in a background such as red and blue instead of the dots and underlines to display synthesized character candidates and divided lines. Character candidates may be distinguished.

FIG. 8A is an example in which the synthesized character candidates of the window screen C of FIG. 3 are displayed in color with a red background H, and FIG. 8B is a divided character candidate of the window screen C of FIG. Are displayed in color on a blue background I. In this way, instead of dots and underlines, colors such as red and blue can be displayed on the background to distinguish between synthesized character candidates and divided character candidates.

FIGS. 3 to 7 are explanatory diagrams showing an example of display in this embodiment. These can also display the colors of the figure on the background to distinguish the synthesized character candidate from the divided character candidate.

FIG. 9 shows an example of storing character data in the text buffer. This example shows a case where the text buffer contains a character string "Oh ..." and only a character "I" has a recognition candidate.
As shown in this figure, candidate characters are stored between the character separator 1 and the character separator 2 only when there is a recognition candidate.

FIGS. 10, 11, 12 and 13 show the contents of the text buffer around the cursor position in the states of FIGS. 3, 5, 6 and 7, respectively. As shown in these figures, the sentence buffer does not store the combined character candidates or the divided character candidates with dots or underscores.
As will be described later, the synthesized character candidate and the divided character candidate are generated and displayed in relation to the characters before and after.

FIG. 12 is an explanatory diagram showing a shape code table in which a plurality of characters which are easily recognized as having the same shape in the recognition process are grouped. This shape code table is RO
It is stored in M3.

In the shape code table, the end of each shape code is represented by null. For example,
“(”, “Ku”, and “<” represent a unit and are represented by the code 8001H.

FIG. 15 is a composite divided character table showing a mutual relationship between two characters and a character recognized as a group when the two characters are likely to be recognized as a group. For example, the character "Buddha" is 80
This indicates that characters in a group represented by codes 04H and 8005H are often recognized as a group and appear.

FIG. 16 is an explanatory diagram showing the contents of the internal management information table. This internal management information table shows information internally managed for candidate display when a candidate is displayed. The contents of this information are
It consists of the display half-width position, character code, and type. The types of information include normal candidates, candidates with a cursor on the left side for composite characters (left target), candidates with a cursor on the right side for composite characters (right target), candidates on the left side of split characters, and split characters. Is a candidate to the right of
For example, "0001", "0002", "0"
003 ”,“ 0004 ”, and“ 0005 ”.

FIG. 17A shows the contents stored in the internal management information table in the state shown in FIG. The detailed display state of the recognition candidate D in this case is shown in FIG.
Is shown in FIG. 18A shows the storage contents of the internal management information table in the state shown in FIG.
The detailed display state of the recognition candidate D in this case is shown in FIG.

The operation of this embodiment will be described below with reference to the flowcharts shown in FIGS. FIG. 19 is a flowchart showing the processing content when selecting a candidate after recognition is completed. In this process, first, a normal candidate is displayed for the character at the cursor position (step S1).
Next, the display of the synthesized character candidate and the display of the divided character candidate are performed (step S2, step S3), and the process waits for the key of the user (step S4). Here, when the cursor movement key is pressed, the candidate display is temporarily erased (step S
5) The designated cursor is moved (step S6).

When a candidate is selected in the key waiting state in step S4, if it is a normal candidate, a normal candidate selection process (step S7) is performed. If a candidate is a composite character candidate, a composite character candidate selection process (step S8) Is performed, and if it is a divided character candidate, a divided character candidate selection process (step S9) is performed. If it is not a cursor movement key or a candidate selection, other processing is performed.

Normal candidate display in step S1, step S1
5, candidate display erasure, cursor movement processing in step S6,
Since the normal candidate selection processing in step S7 is the same processing as the conventional processing, detailed description is omitted.

FIG. 20 is a flowchart showing the contents of the combined character candidate display process. In this process, CODE, CODE are stored in the RAM 2 as areas for storing character codes.
The areas B and CODEN are secured (hereinafter, these areas are simply referred to as CODE, CODEB and CODEN).

Then, first, the character code at the cursor position is extracted and stored in CODE (step S11).
Similarly, the character code of the character before the cursor position is extracted and stored in CODEB (step S12), and the character code of the character next to the cursor position is extracted and CODE is extracted.
N (step S13). However, if there is a recognition candidate, the character code of the first candidate is set, and if there is no character, null is set.

Next, CODE, CODEB, CODEN
14 is searched in the shape code table of FIG. 14, and CODE, CODEB, CODEN
Check whether each character code of is in the target character. If there is a corresponding character in the target character of the shape code table, the shape code at the head of the table is CODE,
CODEB and CODEB (Step S1)
4).

As the composite characters, CODEB and COD
Since E is generated by the combination of CODE and CODEN, first, it is checked whether or not there is a combination character candidate in the combination of CODEB and CODE.

For this purpose, the composite divided character table shown in FIG. 15 is searched (step S15), and the left code / right code and CODEB / C of each table of the composite divided character table are searched.
Check whether the ODE matches (step S1)
6). If there is no match, step S2
Proceed to 1 to check the next combination of CODE and CODEN.

If there is a match in step S16, it is checked whether or not the combined character in the matched table is a shape code (step S17). If it is not a shape code, the code is a composite character code, and the code is displayed as a composite character candidate (step S1).
8).

In step S17, if it is a shape code, the synthesized character code in the shape code table is checked (step S19), and all target characters following the matching shape code are displayed as synthesized character candidates (step S20). .

When displaying as a synthesized character candidate, a dot is displayed at the upper right of the character. Alternatively, when displaying in color, it is displayed with a red background. And the next CODE and C
The processing from step S15 to step S20 is repeated for the combination of ODEN.

FIG. 21 is a flowchart showing the contents of the divided character candidate display processing. In this process, first, the character code at the cursor position is extracted and stored in CODE (step S31). Next, the CODE character code is searched to see if it is in the target character in the shape code table, and a matching shape code is stored in the CODE (step S32).

Next, the composite divided character table is searched (step S33), and it is checked whether or not the CODE is a composite character of the composite divided character table (step S33).
If not, the process ends. If there is, the process proceeds to steps S35 to S38.

In the processing of steps S35 to S38, first, the right and left codes corresponding to the matched combined character code are stored in CODEB and CODEN (step S35).
35).

Next, for CODEB, the shape code in the shape code table is searched. If there is no match, CODEB is set as the left target character. If there is a match, all target characters in the corresponding table are set as left target characters. Characters are set (step S36). Similarly, for CODEN, the shape code in the shape code table is searched, and the obtained target character is set as the right target character (step S37).

Then, all combinations of the left target character obtained in step S36 and the right target character obtained in step S37 are displayed as divided character candidates (step S38). When displayed as a divided character candidate, an underline is displayed over the left target character and the right target character.
Alternatively, when displaying in color, it is displayed with a blue background.

FIG. 22 is a flowchart showing the contents of the composite character candidate selection process. In this process, since the selected character is specified in the internal management information table of FIG. 16, first, it is determined whether or not the selected character is the left target of the composite character (step S41). If the target is not the left target but the right target, the cursor is returned one character to the left in order to obtain the left target character (step S42).

If the target is a left target in step S41, two characters are temporarily deleted from the current cursor position in order to change to a composite character candidate (step S43). next,
The selected synthesis candidate character is stored in the CODE (step S44), and the target character in the shape code table is searched for the CODE (step S45).

If there is no target character (step S
46) and inserts CODE at the cursor position (step S)
47) If there is a target character, the character separator 1, all target characters in the corresponding table, and the character separator 2 are inserted at the cursor position. In this case, character separator 1,
Reference numeral 2 denotes a full-width separator. Thus, the character designated by the cursor is replaced with the selected synthesized character candidate.

FIG. 23 is a flowchart showing the contents of the divided character candidate selection process. In this process, first, it is determined whether or not the selected character is the left character of the divided character (step S51). If it is not the left character, the target of the internal management information table is set to the table immediately before the current divided character left (step S52).

Next, one character is deleted from the cursor position (step S53), the character code of the left of the divided character is stored in the CODE from the internal management information table (step S54), and the CODE is stored in the shape code table. Is searched for (step S55).

Here, if there is no target character (step S
56) and inserts CODE at the cursor position (step S)
57) If there is a target character, the character separator 1, all target characters in the corresponding table, and the character separator 2 are inserted at the cursor position (step S58).

Thus, the left character of the divided character can be processed. Next, the character code on the right of the divided character is taken out from the next internal management information table and stored in CODE (step S59), and the same processing as in steps S55 to S58 is performed on CODE (step S6).
0). As a result, the character specified by the cursor is replaced with the selected divided character candidate.

Thus, the synthesized character candidates are displayed,
By replacing the character specified by the cursor with the selected synthesized character candidate, one character is originally cut out to 2 characters.
Even if the character is divided, the synthesized character that is not divided can be displayed as a recognition candidate, and it is not necessary to perform recognition again. Further, when a composite character is selected, two characters are automatically replaced with one character, so that there is no need to manually insert or delete characters.

Further, by displaying the divided character candidates and replacing the character designated by the cursor with the selected divided character candidate, even if two characters are originally cut out to become one character, the divided character is regarded as a recognition candidate. Can be displayed, and it is not necessary to perform recognition again. Further, when a divided character is selected, one character is automatically replaced with two characters, so that there is no need to manually insert or delete characters.

Further, synthesized character candidates and divided character candidates are displayed on the candidate screen together with normal candidates. For this reason, by displaying the synthesized character candidate and the divided character candidate with dots and underlines, it is possible to identify at a glance what the currently displayed candidate is. In this case, in a device capable of color display, it is also possible to identify the device by color coding.

[0089]

According to the present invention, with respect to a combined character or a divided character in character recognition, a combined character candidate or a divided character candidate is generated by referring to a relatively small-capacity combined character recognition dictionary to generate a specific character. Can be displayed as synthesized character candidates or divided character candidates, so that the recognition rate of character recognition can be improved without significantly increasing the dictionary capacity for recognition processing. Also, the characters displayed on the display
Is indicated by the cursor,
Character candidates are displayed so that they can be selected at the same time as other recognition candidate characters.
Are displayed, dots or underscores
To display the currently displayed candidate.
Can be identified at a glance
You. In this case, if the device is capable of displaying color,
Can be identified by different colors
The efficiency is greatly improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of the present invention.

FIG. 2 is a block diagram showing a configuration of an embodiment in which the present invention is applied to a Japanese word processor.

FIG. 3 is an explanatory diagram showing an example of display in the embodiment.

FIG. 4 is an explanatory diagram showing an example of display in the embodiment.

FIG. 5 is an explanatory diagram showing an example of display in the embodiment.

FIG. 6 is an explanatory diagram showing an example of display in the embodiment.

FIG. 7 is an explanatory diagram showing an example of display in the embodiment.

FIG. 8 is an explanatory diagram showing an example in which a window screen is displayed in color in the embodiment.

FIG. 9 is an explanatory diagram showing an example in which character data is stored in a text buffer in the embodiment.

FIG. 10 is an explanatory diagram showing the contents of a text buffer around the cursor position in the state of FIG. 3 in the embodiment.

11 is an explanatory diagram showing contents of a text buffer around a cursor position in the state of FIG. 5 in the embodiment.

FIG. 12 is an explanatory diagram showing the contents of a sentence buffer around the cursor position in the state of FIG. 6 in the embodiment.

FIG. 13 is an explanatory diagram showing the contents of a text buffer around the cursor position in the state of FIG. 7 in the embodiment.

FIG. 14 is an explanatory diagram illustrating storage contents of a shape code table according to the embodiment;

FIG. 15 is an explanatory diagram illustrating storage contents of a combined divided character table according to the embodiment;

FIG. 16 is an explanatory diagram illustrating an internal management information table according to the embodiment;

FIG. 17 is an explanatory diagram showing storage contents of an internal management information table in the state shown in FIG. 3 in the embodiment.

FIG. 18 is an explanatory diagram showing storage contents of an internal management information table in the state shown in FIG. 6 in the embodiment.

FIG. 19 is a flowchart showing processing contents when selecting a candidate after recognition is completed in the embodiment.

FIG. 20 is a flowchart illustrating the contents of a combined character candidate display process according to the embodiment.

FIG. 21 is a flowchart illustrating details of a divided character candidate display process according to the embodiment.

FIG. 22 is a flowchart illustrating the contents of a combined character candidate selection process in the embodiment.

FIG. 23 is a flowchart illustrating details of a divided character candidate selection process in the embodiment.

[Explanation of symbols]

 Reference Signs List 1 CPU 2 RAM 3 ROM 4 Printer 5 Printer controller 6 HD (hard disk) 7 HD controller 8 Mouse 9 Mouse interface 10 Display device 11 Display controller 12 Pen 13 Pen interface 14 Keyboard 15 Key interface 16 FD (Floppy disk device) 17 FD controller 18 IC card 19 IC card interface 20 Pen tablet 21 Pen tablet interface 23 OCR (optical character reader) 22 OCR interface

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G06K 9/03 G06K 9/62 G06K 9/68

Claims (6)

(57) [Claims] 1. A character recognized by a character recognition process is stored in a memory, the character stored in the memory is displayed on a display in a sentence format, and another recognition candidate character designated by a cursor is selected. A document processing method for displaying a character at a position indicated by a cursor and characters before and after the position indicated by the cursor from the memory. A character table in which one character is stored in association with a combination of characters that can be easily recognized as a plurality of characters is searched, and one of the character at the extracted cursor designated position and the character before and after the cursor designated position is included. When a combination of characters exists in the table, one character stored in association with the corresponding combination of characters is regarded as the composite character candidate. A document processing method for simultaneously displaying the recognition candidates on a spray. 2. When the synthesized character candidate is selected,
The characters present before or after the characters and the cursor indication position of the cursor indication position, the document processing method according to claim 1, wherein storing in the memory by replacing the selected synthetic character candidates. 3. A part corresponding to a character designated by a cursor.
Displaying the synthesized character candidate so that the character can be identified.
3. The document processing method according to 1 or 2 . 4. A character recognized by the character recognition processing is stored in a memory, the character stored in the memory is displayed on a display in a sentence format, and another recognition candidate character indicated by a cursor is selected. A document processing method for displaying a character as possible, wherein when a displayed character is designated by a cursor, a character at the cursor designated position is extracted from the memory, and one character and the one character are recognized as a plurality of characters. A character table is stored in which a combination of easy-to-use characters is stored in association with each other, and when the character at the extracted cursor-indicated position exists in the table, the character table stored in relation to the corresponding one character is stored. A document processing method for displaying a combination of characters as divided character candidates on the display simultaneously with the recognition candidates . When the method according to claim 5, wherein the divided character candidates are selected,
5. The document processing method according to claim 4 , wherein the character at the cursor designated position is replaced with a combination of a plurality of characters that are selected divided character candidates and stored in the memory. 6. The method of claim 4 displaying the divided character candidates to another recognition candidate characters as identifiably the display or
5. The document processing method according to item 5 .