JPH1055405A - Device and method for reading directory - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate directory data by reading a directory recorded in an original and easily and surely dividing it into each directory datum. SOLUTION: A directory reading device for reading a character concerning the directory recorded in the original and generating directory data is provided with an image reading control part 34 which reads image data by optical scanning the original where plural kinds of directory data consisting of the characters concerning the directory about plural items are arranged and recorded, a character recognition control part 36 which recognizes the character based on read image data, in item recognition control part 38 which detects the character concerning the directory corresponding to the specified item from the recognized character and a directory recognition control part 30 which detects the division of directory data every datum as against the character based on the character concerning the detected directory.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an address book reading apparatus and an address book reading method for optically reading an address book recorded on a manuscript and creating address book data divided into address data for each case. .

[0002]

2. Description of the Related Art In optical character reading (OCR), image data is detected by optically scanning characters or the like printed on paper, and characters are recognized based on the image data to form a character code. Technology. One of the applied technologies of OCR is that image data is detected from a document in which an address book is recorded and automatically converted to address book data (address data) that can be used in personal computers, word processors, portable electronic devices, and the like. Address book reading function.

In the address book, a plurality of address data composed of contents (character strings) corresponding to a plurality of items (name, address, telephone number, etc.) are arranged. Characters are read and converted into data.

When an address book recorded on a manuscript is read and converted into address book data, it is necessary to divide the data into address data for one case (for one person). Conventionally, the detection of a break position for each case has been performed as follows.

(1) When a ruled line exists in a document, a start position of an actual address data range is designated, and, for example, a horizontal ruled line is recognized and extracted from an address book below the start position. One piece of address data is determined based on positions separated by horizontal ruled lines.

(2) By reading an image of an original having no ruled lines in the original and the entire original, displaying an image of an address book, and manually specifying a delimiter position of each address data on a display screen. Determine the break position.

[0007]

However, the prior art has the following problems. (1) In the method of determining the break position of the address data based on the ruled line, the rule is relatively simply divided only by the existence of the horizontal ruled line. For example, as shown in FIG. 13, a horizontal ruled line is provided in the document. In this case, the delimiter position of the address data for each case could not be correctly detected, and the address book could not be read correctly.

That is, in the example shown in FIG. 13 (a), the range B in the figure should be handled as one piece of address data, but the range between the address and the company name or between the telephone number and the fax number should be used. Since ruled lines (solid lines) are provided, A in the figure may be treated as address data in some cases.

In the example shown in FIG. 13 (b), similarly, a horizontal ruled line (dotted line) of a line type different from the others is provided while the range D in the figure should be treated as one piece of address data. In some cases, this position is used as a delimiter for address data.

(2) In the method of inputting the designation of the break point of the address data for each address book, the work load is large,
In particular, when creating address book data for a large number of address books, an enormous amount of work is required.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described circumstances, and is intended to read an address book recorded on a manuscript and to easily and surely divide the address data into individual address data to create address book data. It is an object of the present invention to provide an address book reading device and an address book reading method capable of reading an address book.

[0012]

SUMMARY OF THE INVENTION According to the present invention, there is provided an address book reading apparatus for reading characters related to an address book recorded on a document and creating address book data. Image reading means for reading image data by optically scanning a document on which a plurality of data are arranged and recorded,
Character recognition means for recognizing characters based on the image data read by the image reading means; and item recognition means for detecting, from the characters recognized by the character recognition means, characters relating to an address book corresponding to a specific item. And address book recognizing means for detecting a break of address data for each character on the basis of characters related to the address book detected by the item recognizing means.

[0013] Thus, even if the address data is not separated by ruled lines, even from a document in which an address book consisting only of characters is recorded, the address of each case is determined based on the contents of the recorded characters themselves. The operability can be improved by dividing the data.

Further, according to the present invention, in an address book reading apparatus for reading characters related to an address book recorded on a document and creating address book data, a plurality of address data consisting of characters related to the address book for a plurality of items are arranged. Image reading means for reading image data by optically scanning the recorded document, character recognition means for recognizing characters based on the image data read by the image reading means, and image reading means Line recognition means for recognizing a line component based on the image data read by, and an effective line detection means for extracting a line component satisfying a certain condition from the characteristics of the line component recognized by the line recognition means, The sentence recognized by the character recognition unit based on the line component extracted by the effective line detection unit. Characterized by comprising the address book recognition means for detecting a break address data for each 1 against.

[0015] With this, even in a case where a line on which an address book including ruled lines or the like is recorded is mixed with a line that does not represent a delimiter position of address data, unnecessary segmentation is removed and only valid lines are removed. , The address data delimiter position is detected, so that address book data can be created reliably and easily.

Further, the effective line detecting means uses at least one of a line type and a line length as a characteristic of a line component. For example, depending on the address book, a line (for example, a solid line) indicating a division of address data
And a line (dotted line) indicating a break in the address data may be used properly, and the break position of the address data can be definitely determined by using the feature of the line component. In addition, a line representing a delimiter of address data may be longer than the length of the character string of the address data, and by using such a feature, the position of the delimiter of the address data can be similarly reliably determined.

Further, the image processing apparatus further comprises a line recognition means for recognizing a line component based on the image data read by the image reading means, wherein the address book recognition means comprises a It is characterized in that a break of address data for each case is detected based on a feature and characters related to the address book detected by the item recognition means.

[0018] With this configuration, when a manuscript in which an address book is recorded includes a line that separates address data, both the characteristics of the line and the content (character string) corresponding to the address book item are combined.
By determining the division position of the address data, the division accuracy is improved.

Further, after dividing the address data for each case based on the characteristics of the line components by the line recognition means, the address data includes characters related to an address book corresponding to a certain specific item. It is characterized in that it is determined whether or not the delimiter position is valid depending on whether or not there is a break position.

In this way, the validity of the delimiter position determined using the lines included in the address book is confirmed using characters related to the address book corresponding to the specific item. The accuracy of the data break position is improved.

Further, before detecting a break of address data by the address book recognizing means, a character representing a specific item is removed from the characters recognized by the character recognizing means in advance.

By removing character strings not included in the address data included in the address book, for example, character strings such as "address book" and "address book", which are the titles of the manuscript, in advance, Unnecessary character strings are prevented from adversely affecting subsequent processing for dividing address data, and the accuracy of dividing and recognizing address data is improved.

Further, according to the present invention, in an address book reading apparatus for reading characters from a document on which a plurality of address data including characters related to an address book for a plurality of items are arranged and recorded and creating address book data, A delimiter position detecting means for detecting a delimiter position for each piece of address data for a read character, and a delimiter position for changing a delimiter position detected by the delimiter position detector according to an instruction given arbitrarily Correction means.

[0024] With this, if the delimiter position of the address data automatically detected by the device is incorrect, the delimiter position can be arbitrarily changed by a manual operation as needed, and the delimiter position is automatically detected. The delimiter position of the address data can be reliably set by the minimum required operation according to the accuracy.

Further, the delimiter position correcting means corrects the delimiter position detected by the delimiter position detection means for one case, and thereafter determines whether or not to change the delimiter position for each case. It is characterized by the following.

This makes it easy to understand the operation of changing the delimiter position of the address data automatically detected by the apparatus. In addition, by correcting one case, the delimiter position is changed according to the same rule as described below. If the position can be determined, the change of the break position for each case can be omitted.

[0027] Further, there is provided priority changing means for determining the display priority of the item in accordance with the detection state of the character corresponding to each item by the item recognizing means, and changing the editing target for the detected character. It is characterized by.

Thus, the display priority of an item with a small content is reduced according to the detection status of the content (character string) corresponding to each item read from the address book. Operability is improved due to the selective selection.

[0029]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an address book reading device according to the present embodiment. The address book reading device reads a program recorded on a storage medium, for example, and is realized by a computer whose operation is controlled by the program.

As shown in FIG. 1, a CPU 10 and a ROM 1
2, RAM 14, floppy disk drive (FDD) 1
6, color scanner device 18, keyboard (KB) 2
0, display controller 22, VRAM (Video
RAM) 24, a color display device 26, and a color printer device 28.

The CPU 10 controls the entire apparatus, and controls a function of creating a document including an image by an operation according to a program stored in the ROM 12 and the RAM 14.

The ROM 12 stores programs for controlling the CPU 10 and fixed data. RAM
14 is a program for controlling the CPU 10,
It is used as a buffer memory for temporarily storing various data.

Floppy disk drive (FDD) 16
Performs reading and writing of a file such as a document and an image on a floppy disk, reading of a file of a program for controlling an operation of the apparatus, and the like.

The color scanner 18 optically reads a document on which an address book or the like is recorded and converts the document into image data. The keyboard (KB) 20 is an input device for inputting characters and performing specific instructions including operations on the color printer device 28. The keyboard 12 is assigned a "print" key, an "image input" key, and the like, in addition to a "character" key, a "numerical value" key, a "cursor" key, and the like.

The display controller 22 has a CPU
The display data developed in the VRAM 24 under the control of 10 is displayed on the color display device 26. In the VRAM 24, display data to be displayed on the color display device 26 is expanded.

The color display device 26 displays images, documents, and the like based on the display data developed in the VRAM 24 under the control of the display controller 22.

The color printer device 28 performs printing based on data such as an address book and images stored in the RAM 14. FIG. 2 is a diagram showing a functional configuration (software) of the address book reading device realized by the configuration shown in FIG.

Each section is a function realized by the CPU 10 executing a program stored in the memory. As shown in FIG. 2, the address book reading device includes:
It is composed of an address book recognition control unit 30, an address book edit control unit 32, an image reading control unit 34, a character recognition control unit 36, an item recognition control unit 38, a keyword dictionary 40, and a line recognition control unit 42.

The address book recognition control section 30 controls the entire operation of reading the address book, and controls each section in accordance with the procedures of flowcharts shown in FIGS. 5 and 6 described later. The address book recognition control unit 30 is provided with a break position correction unit 30a. The break position correction unit 30a includes a character recognition control unit 36, an item recognition control unit 38, and a line recognition control unit 42.
In the address data read from the document, which is classified based on the processing in
It is changed according to an instruction arbitrarily given from B20.

The address book editing control unit 32 edits (edits contents, newly inputs, etc.) the address book data read from the manuscript after displaying it on the display screen. The address book editing control section 32 is provided with a priority changing section 32a. The priority changing unit 32a determines the display priority of the item according to the detection state of the character corresponding to each item in the address book, and sets an editing target for the detected character. That is, the priority changing unit 32a controls so as to lower the display priority of an item whose content is less described in the document.

The image reading control unit 34 controls the color scanner device 18 to optically scan an original on which a plurality of address data including characters related to an address book for a plurality of items are recorded and arranged. Read the image data.

The character recognition control unit 36 recognizes characters based on the image data read under the control of the image reading control unit 34 and converts them into character codes. At this time, the character recognition control unit 36 also obtains a plurality of candidate characters for each character.

The item recognition control unit 38 includes the character recognition control unit 3
6, characters related to an address book corresponding to a specific item are detected by referring to dictionary data (keywords) registered in the keyword dictionary 40. The item recognition control unit 38 determines that a character string generated by combining a plurality of candidate characters corresponding to each character of the character string recognized by the character recognition control unit 36 matches a keyword registered in the keyword dictionary 40. By detecting whether or not the character is a character related to the address book corresponding to the specific item.

The keyword dictionary 40 includes the item recognition control unit 3
8 is a dictionary for recognizing a character string relating to the address book by the processing in 8, and dictionary data (keyword) is registered for each character string corresponding to each item used in the address book.

The line recognition control unit 42 extracts line components such as ruled lines on the document based on the image data read from the document by the image reading control unit 34. The line recognition control unit 42 detects, as features, the position, type (solid line, broken line, dotted line, etc.) and length of a straight line included in the image data as a line component. Further, the line recognition control unit 42 is provided with an effective line detection unit 42a. Effective line detector 42a
The line component satisfying a certain condition is obtained from the feature of the line component or the correspondence between the feature of the line component and the character position in the image read by the image reading control unit 34 (the relative position between the line and the character). And at least one of the type of the straight line and the length of the line is used as the feature of the line component.

In this embodiment, characters related to the address book are read from a document (address book sheet) on which the address book is recorded as shown in FIG. 3 or FIG. Data shall be created.

The address book sheet shown in FIG. 3 is an address book including ruled lines, in which a plurality of items are arranged in a horizontal direction, and address data including contents (character strings) corresponding to each item are arranged in a vertical direction. Has been recorded. In FIG. 3, "NO."
[Name], [@], [Address] and [Remarks] are items. The ruled line is provided at a position for dividing the item line and the line of each address data, and a position for dividing the contents corresponding to each item.

The address sheet shown in FIG. 4 is an address book not including ruled lines, and the arrangement of items and address data corresponding to each item is the same as that of FIG. 3 except for the ruled lines. Next, the operation of creating the address book data in the present embodiment will be described with reference to FIG.
And a flowchart shown in FIG.

First, when an instruction to create address book data is issued, the address book recognition control unit 30 is started. The address book recognition control unit 30 first determines whether the document to be read in the address book is a document with ruled lines as shown in FIG. 3 (ruled line mode) or a document with only characters as shown in FIG. The designation of the ruled lineless mode) is input by the user through, for example, the KB 20 (step A1).

The address book recognition control unit 30 instructs the image reading control unit 34 to execute reading of image data from a document. The image reading control unit 34 controls the color scanner 18 to read one page of image data by optically scanning a document set on the color scanner 18 (step A2).

When the mode with a ruled line is specified in step A1 (step A3), the address book recognition control section 30 causes the line recognition control section 42 to detect a linear component from the image data of the document (step A3). A
4).

The line recognition control unit 42 detects a line position, a line type, a length and the like as a straight line component. In addition, as a technique for detecting such components, an existing technique can be used, and thus detailed description is omitted.

The address book recognition control unit 30 causes the character recognition control unit 36 to recognize characters from the image data of the document (step A5). The character recognition control unit 36, for example, extracts a character area from image data, cuts out a character pattern for each character, and matches the character pattern with a character pattern recognition dictionary pattern to recognize a character.

The item recognition control unit 38 includes the character recognition control unit 3.
6, a character string (keyword) corresponding to a specific item used in the address book is extracted using the keyword dictionary 40 (step A6).

For example, it is assumed that a character string "address" exists in a document, and as a result of performing character recognition on these character strings, three character candidates are obtained for each character as shown in FIG. . That is, the character candidates for the first character are “go”, “house”, and “individual”, and the character candidates for the second character are “place”, “pray”, and “fold”. Item recognition control unit 38
Checks whether each character string generated by combining the first character candidate and the second character candidate is registered in the keyword dictionary 40 as a keyword.

If "address" is registered in the keyword dictionary 40, the character string of the combination of the second candidate for the first character and the first candidate for the second character is "address". Is extracted as a keyword.

In the keyword dictionary, as shown in FIG.
Dictionary data (keywords) corresponding to each item used in the address book, such as first and last name data, address keyword data, affiliation data, title data, and specific keyword data, are registered.

The first and last name data include "Ito", "Yamada", and "Taro".
Keywords used for first and last names such as… and address keyword data include “Tokyo (Tokyo)”, “** Prefecture”, “** City”, and “**”.
Keywords used for addresses such as "town" and "** chome"
Affiliation data includes “Company”, “Laboratory”, “** Agency”, “**”
Section, "** company", "** elementary school"
For title data, keywords such as “section manager”, “department manager”, “section manager”, and for specific keyword data, “address”, “name”
It includes keywords used for the items and titles of the address book such as "telephone number", "address", "TEL", "FAX", "address book", and "address book" (note that "*"
Represents any character).

Further, the item recognition control unit 38 outputs “***”
If there is a character string such as "***-**", the postal code,
"***-***""*** (***) ***"
If there is such a character string, it is also detected that it is a telephone number (note that "*" represents an arbitrary number).

When the extraction of the keyword from the character recognition result is completed, the item recognition control unit 38 deletes an unnecessary keyword as address book data in the extracted keyword (step A7).

In some cases, the address book document contains a word (“address book”) as shown in FIG. 9 as a title. This type of character string is a word that is unnecessary as address book data, and has an adverse effect on subsequent processing for separating address data on a case-by-case basis and correspondence with address book items (incorrect processing results). Is deleted from the character string of the character recognition result. Character strings such as “address book” and “list” other than “address book” are detected as unnecessary keywords.

On the other hand, the line recognition control unit 42 uses the effective line detection unit 42a to convert a straight line component (effective ruled line) for dividing address data for each address from the straight line components detected from the image data into a constant. Extraction is performed based on conditions (step A8).

Specifically, a straight line component satisfying the following condition (hereinafter referred to as an effective ruled line condition) is regarded as an effective ruled line.
First, it is substantially horizontal to the shape of the document (however, if the address data is divided in the vertical direction, it should be almost vertical), and second, the line type matches the predetermined type. Thirdly, the length of the original paper is longer than a certain ratio with respect to the length in the vertical or horizontal direction or the length of the character string.

The third condition of the effective ruled line condition is, for example, a line component for separating address data, which is usually longer than a character string and extends over most of the range in the vertical or horizontal direction of the document paper. This is based on the fact that a line component provided can be regarded as provided for classifying address data.

In the example shown in FIG. 13A, the line component extending between the address and the telephone number is the length of the character string (“Ito-9999”).
The range is shorter than “up to”), and is determined to be invalid according to the third condition. In the example shown in FIG. 13B, when the valid line type is a solid line, the dotted line is determined to be invalid according to the second condition.

The valid ruled line conditions used by the valid line detecting section 42a may be arbitrarily specified in advance in, for example, various mode settings related to the address book reading process (a document having a line direction, a line type, and a line length). The ratio may be determined by default.

By the processing of the effective line detecting section 42a,
It is possible to remove a line component that is not a delimiter of the address book data for each case, thereby improving the accuracy of the delimitation of the address data based on the line component.

Next, the address book recognition control unit 30 determines whether or not it is possible to classify address data on a case-by-case basis only by the valid line segments detected by the line recognition control unit 42 under the following conditions (hereinafter, referred to) (Referred to as a dividable condition) (step A9).

The address book recognition control unit 30 includes the following keywords (1) recognized by the item recognition control unit 38 in the character strings delimited by the respective effective ruled lines in one page.
If (2) exists, it is determined that division is possible only with the valid ruled line, and if it does not exist, it is impossible.

(1) There are a certain number of keywords determined as first and last names (for example, two main and secondary keywords). (2)
There are a certain number of keywords determined to be addresses or telephone numbers.

That is, in general, the address data does not include the data of the first and last names, and is based on the fact that either the address or the telephone number is registered in association with each other. By performing the processing in the address book recognition control unit 30, it is determined whether or not there is no problem even if the division of the address data is performed using the effective ruled line, so that the accuracy of the division position of the address data can be improved.

If the address book recognition control unit 30 determines in step A9 that division is possible, the character recognition result by the character recognition control unit 36 is provisionally divided based on the detected effective ruled line, and the address is recognized for each address. Data (Step A1)
4). That is, the address book recognition control unit 30 determines the temporary division position for the address data based on the relationship between the position of the character recognition result (character) in the image data and the position of the effective ruled line.

On the other hand, if it is determined in step A9 that the division is impossible, the address book recognition control unit 30 regards the division based on only the effective ruled lines as having low reliability, and
The image is divided using the keyword recognized by the item recognition control unit 38 (step A13).

The address book recognition control unit 30 determines, from the keywords recognized by the item recognition control unit 38, a position that satisfies the following condition (hereinafter, referred to as a temporary break condition) as a break position of one piece of address data. Then, that position is set as a temporary break position.

As specific conditions, first, keywords of a first and last name and an address or a telephone number exist in a certain range (for example, two), and second, keywords corresponding to a plurality of items are the same. Thirdly, the area in which keywords that can be regarded as one for one third are recorded is almost the same. The address book recognition control unit 30 detects a position that satisfies the above provisional division condition, and sets it as a provisional division position of one piece of address data.

According to the processing based on the keyword by the address book recognition control unit 30, 1 is correctly obtained only from the effective ruled line.
Even if the delimiter position of the address data for the case cannot be determined, the address data in the address book can be automatically delimited.

On the other hand, in step A3, if the ruled lineless mode is designated, the address book recognition control section 30 causes the character recognition control section 36 to recognize characters from the image data of the original (step A10). The keyword is recognized by the item recognition control unit 38 from the recognition result (step A11). Further, the address book recognition control unit 30
Causes the item recognition control unit 38 to delete unnecessary keywords (step A12). Step A10
The processing in A12 is performed in the same manner as the processing in steps A5 to A7 described above, and a detailed description thereof will be omitted.

In the ruled lineless mode, the address book recognition control unit 3
0 uses the keyword unconditionally recognized by the item recognition control unit 38 and determines a temporary break position of the address data based on the above-described temporary break condition (step A1).
3).

By the processing based on the keyword by the address book recognition control unit 30, even if there is no ruled line on the address book paper, the address book is based on only the characters (keywords) recorded on the paper. Address data can be automatically separated.

When the provisional separation position of the address data for each case is determined, the address book recognition control unit 30 displays a screen for specifying the recognition character string obtained by the character recognition control unit 36 and the provisional separation position. Display on the color display device 26 (step A15), and the range of the temporary division of the first one in the address book (character string area of one address data)
Is highlighted (step A16).

FIG. 10A shows a display state in which the range of the address data of the first case is inverted.
As shown in FIG. 10A, the character recognition result (character string) is displayed in the same arrangement as the arrangement of the character strings in the original from which the address book is read, and the temporary break positions are indicated by dashed lines. . In addition, since the address data for one case is represented by two character strings, the area of the character strings for two columns is highlighted.

Here, the address book recognition control section 30 controls the color display device 26 to, for example,
Address data? Is displayed and the user is asked to confirm.

If the first temporary break position is not correct, an instruction to that effect is input (step A17), and an instruction to change the reversal range is input. That is, when the user presses the “cursor” key in the vertical direction of the KB 20, the break position correction unit 30 of the address book recognition control unit 30
“a” changes the lower end position of the reversal range up and down line by line (step A18). For example, when a downward range change is instructed, the break position correcting unit 30a changes the inversion range as shown in FIG. After the reversal range is changed to the range of one desired address data, an instruction to determine the break position is input.

By this operation, even if the position automatically divided in the processing up to step A14 is incorrect, the division position can be easily changed arbitrarily by manual operation.

When the break position of one piece of address data is determined in this way, the address book recognition control unit 30 automatically determines the break position for the following address data, or the same as the first case. The user is asked to confirm whether or not to continue the operation of confirming the break position for each case.

Here, in the case where the operation of confirming the break position for each case is continued, the address book recognition control unit 30 checks the remaining data from the next line of the address data of the current case to the next line. Is displayed as one piece of tree data up to the temporary break position (temporary automatic division line) (step A22), and in the same manner as described above, while inputting an instruction to change the determination range and an instruction to determine the break position, Determine the break position.

On the other hand, when it is instructed that the future confirmation operation is unnecessary, the address book recognition control section 30 sets the temporary break positions for the second and subsequent address data every final one. (Step A20). With this operation, when the entire address book is automatically separated for each piece of address data with high accuracy, the correction confirmation operation for each piece is unnecessary, and the operability is improved.

As described above, the break position of the address data for each case is determined. Next, the address book recognition control unit 30 classifies and stores the character strings included in the address data delimited for each case according to the item recognition result by the item recognition control unit 38 for each content corresponding to each item. Then, address book data is created (step A23).

That is, the address book recognition control section 30 classifies the item recognition results into items necessary as address book data, for example, names, postal codes, addresses, telephone numbers, posts, remarks, and the like. Store in the address position corresponding to the item.

The postal code, address, and telephone number have two types, one for business use and one for home use, and it is determined which data is from the keyword. For example, an address having a keyword such as "company", "job", "** company", and "** building" is determined to be a company address, and postal and postal codes and telephone numbers before and after it are also determined to be for a company. FIG. 11 shows an example of a data structure in a state where the address book data is stored in the RAM 14.

As shown in FIG. 11, the data buffer can record two kinds of information, that is, company information and home information. On the other hand, there are various types of information, most of which are company information.

Therefore, the priority changing unit 32a of the address book editing control unit 32 determines whether the company information and the home information are included in one address book composed of a plurality of address data determined by the address book recognition control unit 30. It is determined which information is larger, and a mode is set to give priority to the display of the larger information.

For example, when there is more home information in the address book data, display of the home information is prioritized by the priority changing unit 32a, and the address book editing control unit 32 sets the home information as the edit target in FIG. ) (Steps A24 and A26). In a state where the home information is displayed, the address book editing control unit 32 edits the address book data such as confirmation, correction, and additional registration of the address book data according to the input for each item from the KB 20 (step A2).
7).

Similarly, if there is more company information, the address book editing control unit 32 sets the editing target as company information in FIG.
2 (b) (Step A24,
A25).

Thus, the one with more information described in the address book can be set as the initial display mode.
The operation amount of the display switching can be reduced, and the operability is improved.

Of course, when the address book editing control unit 32 is in the editing state of the address book data and inputs a priority mode switching instruction from the KB 20 by operating the “switch” key, the home information editing screen is displayed. (FIG. 12A) and the editing screen for company information (FIG. 12B) can be switched between each other.

The edited and corrected address book data is stored as a file on a floppy disk by the FDD 16, for example. The present invention can be applied not only to an address book reader as a peripheral device of a word processor or a personal computer but also to a built-in function of a word processor or application software of a personal computer.

In addition, not only an address book but also an automatic division of data per case when reading and recognizing a manuscript of a list in which a plurality of items of data are arranged, such as a list of purchased items, is performed. Can also be used.

Further, the method described in the above-described embodiment can be implemented as a program that can be executed by a computer, such as a magnetic disk (floppy disk, hard disk, etc.), an optical disk (CD-ROM, DVD, etc.), a semiconductor memory, etc. It can be provided stored in a medium. The computer that realizes the address book reading device according to the present embodiment reads the program recorded on the recording medium, and executes the above-described processing by controlling the operation of the program.

[0100]

As described above in detail, according to the present invention, it is possible to read an address book recorded on a manuscript, and to easily and surely divide the address book into individual address data to create address book data. It becomes.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an address book reading device according to an embodiment of the present invention.

FIG. 2 is a diagram showing a functional configuration (software) of the address book reading device realized by the configuration shown in FIG. 1;

FIG. 3 is a view showing an example of an address book including ruled lines handled in the embodiment.

FIG. 4 is a diagram showing an example of an address book that does not include a ruled line handled in the embodiment.

FIG. 5 is a flowchart for explaining a part of the operation of creating address book data in the embodiment.

FIG. 6 is a flowchart for explaining a part of an operation of creating address book data in the embodiment.

FIG. 7 is an exemplary view for explaining a keyword recognition process performed by an item recognition control unit according to the embodiment.

FIG. 8 is a view for explaining an example of data registered in a keyword dictionary according to the embodiment.

FIG. 9 is a diagram illustrating an example of a title included in an address book document.

FIG. 10 is an exemplary view showing a display example of a state in which the range of the address data of the first case is reversed in the embodiment;

FIG. 11 is a view showing an example of a data structure of address book data stored in a RAM according to the embodiment.

FIG. 12 is a view showing an example of a display mode of each information according to a priority mode in the embodiment.

FIG. 13 is a view for explaining a conventional address data dividing method.

[Explanation of symbols]

 Reference Signs List 10 CPU 12 ROM 14 RAM 16 Floppy disk device (FDD) 18 Color scanner device 20 Keyboard (KB) 22 Display controller 24 VRAM (Video RAM) 26 Color display device 28 Color printer device 30 ... address book recognition control section 30a ... break position correction section 32 ... address book edit control section 32a ... priority change section 34 ... image reading control section 36 ... character recognition control section 38 ... item recognition control section 40 ... keyword dictionary 42 ... line Recognition control unit 42a: valid line detection unit

Claims (12)

[Claims] 1. An address book reading device for reading characters related to an address book recorded on a manuscript and creating address book data, wherein a plurality of address data including characters related to an address book for a plurality of items are arranged and recorded. Image reading means for reading image data by optically scanning the read original, character recognition means for recognizing characters based on the image data read by the image reading means, and character recognition means for recognizing the characters. Item recognition means for detecting a character associated with an address book corresponding to a specific item from the extracted characters; and, based on the character associated with the address book detected by the item recognition means, separating address data for each character. Address book reading device, comprising: an address book recognizing means for detecting an address book. 2. An address book reading apparatus for reading characters related to an address book recorded on a manuscript and creating address book data, wherein a plurality of address data comprising characters related to an address book for a plurality of items are arranged and recorded. Image reading means for reading image data by optically scanning the read original, character recognition means for recognizing characters based on the image data read by the image reading means, and character reading means for reading characters by the image reading means. Line recognition means for recognizing a line component based on the image data obtained, from the characteristics of the line component recognized by the line recognition means,
Effective line detecting means for extracting a line component satisfying a certain condition; and dividing address data for each character for the character recognized by the character recognizing means based on the line component extracted by the effective line detecting means. Address book reading device, comprising: an address book recognizing means for detecting an address book. 3. The effective line detection means uses at least one of a line type and a line length as a characteristic of a line component.
Address book reader as described. 4. The image processing apparatus according to claim 1, further comprising: a line recognition unit configured to recognize a line component based on the image data read by the image reading unit; 2. The document creation apparatus according to claim 1, wherein a break of address data for each case is detected based on a feature and characters related to the address book detected by the item recognition unit. 5. After segmenting address data for each case on the basis of the characteristics of line components by the line recognition means, the address data includes characters relating to an address book corresponding to a certain specific item. 5. The document creation apparatus according to claim 4, wherein whether or not the separation position is valid is determined based on whether or not the document is separated. 6. The method according to claim 1, wherein a character representing a specific item is removed from characters recognized by said character recognizing means in advance before detecting a break of address data by said address book recognizing means. The document creation device according to claim 2. 7. An address book reading device for reading characters from a document on which a plurality of address data including characters related to an address book for a plurality of items are arranged and recorded, and creating address book data, wherein: Delimiter position detecting means for detecting a delimiter position for each piece of address data for each character, and delimiter position correcting means for changing a delimiter position detected by the delimiter position detector in accordance with an arbitrary instruction A document creation device comprising: 8. The delimiter position correcting means corrects the delimiter position detected by the delimiter position detection means for one case, and thereafter determines whether or not to change the delimiter position for each case. The document creation device according to claim 7, wherein 9. A system according to claim 1, further comprising a priority changing unit for determining a display priority of the item according to a detection state of a character corresponding to each item by the item recognizing unit, and changing an editing target for the detected character. 2. The document creation device according to claim 1, wherein: 10. An address book reading method of reading characters related to an address book recorded on a manuscript and creating address book data, wherein a plurality of address data consisting of characters related to an address book for a plurality of items are arranged and recorded. Image data is read by optically scanning the scanned document, characters are recognized based on the read image data, and characters related to an address book corresponding to a specific item are detected from the recognized characters. A method of reading an address book, comprising detecting, based on the detected characters related to the address book, a break of address data for each character with respect to each character. 11. An address book reading method for reading characters related to an address book recorded on a manuscript and creating address book data, wherein a plurality of address data including characters related to an address book for a plurality of items are arranged and recorded. The scanned original is scanned optically to read image data, recognize characters based on the read image data, and recognize line components. A method of reading an address book, comprising extracting a line component satisfying a condition, and detecting a segment of address data for each recognized character based on the extracted line component. 12. A program for reading characters related to an address book recorded on a manuscript and creating address book data, wherein a plurality of address data including characters related to an address book for a plurality of items are arranged. Image data is read by optically scanning the recorded document, characters are recognized based on the read image data, and characters related to an address book corresponding to a specific item are recognized from the recognized characters. A computer-readable recording medium storing a program for detecting and detecting a break of address data for each character based on the detected character of the address book.