JPH07230530A - Character recognition device and method for estimating reading - Google Patents

Abstract

PURPOSE:To provide a character recognition device and a reading estimating method capable of estimating the reading of a name entered on paper. CONSTITUTION:When a CPU 21 recognizes a name entered on a business card through a character recognizing part 25 on the side of the character recognition device 20, a system ROM 23 is accessed, a name dictionary stored in the ROM 23 is referred to and three leading characters of the name are checked by the name dictionary. When the name concerned exists in the dictionary, the CPU 21 extracts a 1st reading candidate from the name dictionary and outputs the extracted result to a host device 10 as an estimated result.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition device used in a business card management system for managing a plurality of business cards, and more particularly to a character recognition device and a pronunciation estimation method capable of estimating the reading of a name written on a business card. .

[0002]

2. Description of the Related Art Conventionally, a business card management system comprises a host device and a character recognition device, and is configured so that the host device manages data of a business card read by the character recognition device. In this case, the business card has items such as company name, name, address, telephone number, and fax number. The character recognition device classifies and recognizes the character strings written on the business card for each of the above items, and outputs the recognition result to the host device.

[0003]

By the way, in the case of managing business card data, the host device generally adopts a method such as sorting in alphabetical order based on the reading of the name. However, with the conventional character recognition device, although the name written on the business card can be read, the reading cannot be estimated. Therefore, it is very troublesome for the user to input the reading after reading the business card.

Some business cards have ruby added to their names in advance. However, the conventional character recognition device can only recognize the ruby as a simple character string and cannot estimate it as the reading of the name.

The present invention has been made in view of the above points, and an object thereof is to provide a character recognition device and a reading estimation method capable of estimating the reading of a name written on a sheet, and in particular, it is written on a business card. The first purpose is to reduce the burden on the user in reading the business card by estimating the reading of the name, and if the business card has ruby added to the name, recognize the ruby and correctly read the name. The second purpose is to estimate, and the third purpose is to correctly estimate the reading of the name even when the business card with ruby added to the name and the business card without ruby are mixed and read. .

[0006]

The character recognition device of the present invention uses a surname and surname dictionary in which a name and its reading are associated with each other, and when the name written on a paper such as a business card is recognized, the name is recognized. The reading is estimated by referring to the surname dictionary.

Further, when the character recognition device of the present invention recognizes a name written on a sheet such as a business card, it detects ruby information added in the vicinity of the name, and this ruby information is used for the name. It is supposed to be read.

Further, the character recognition device of the present invention detects the ruby information by detecting the ruby information added in the vicinity of the name when the name written on the paper such as a business card is recognized. If the ruby information is estimated to be the reading for the name, if the ruby information is not detected, the reading for the name is referred to by referring to the surname and surname dictionary. is there.

[0009]

According to the above construction, when the name written on the paper such as a business card is recognized, the reading corresponding to the name is retrieved from the surname dictionary and output as the estimation result. Therefore, when reading a business card, even if the user does not input the reading of the name, the reading can be automatically registered. This can reduce the burden on the user in the business card reading work.

When a name written on a paper such as a business card is recognized, ruby information near the name is detected,
This ruby information is estimated as the reading for the name. Therefore, when reading a business card in which ruby is added to the name in advance, the reading of the name can be correctly estimated.

Further, when the name written on the form such as a business card is recognized, if there is ruby information near the name, this ruby information is output as a reading for the name, and the ruby information is displayed near the name. If not, the pronunciation for the name is estimated using the surname dictionary. Therefore, even when a business card in which ruby is added to the name and a business card without ruby are mixedly read, the reading of the name can be correctly estimated.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. (First Embodiment) FIG. 1 is a block diagram showing the configuration of a business card management system. The business card management system is a host device 1.
0 and the character recognition device 20, and the host device 10 manages the data of the business card read by the character recognition device 20. Host device 10 and character recognition device 20
Are electrically connected by a cable 30. Business card,
There are items such as company name, name, address, telephone number, and fax number. The character recognition device 20 classifies and recognizes various character strings entered on a business card for each of the above items, and outputs the recognition result to the host device 20.

The host device 10 includes a CPU 11, a floppy disk device (hereinafter referred to as FDD) 12, and a RAM 1.
3, a keyboard 14, and a display device 15. CPU1
1 performs business card management control on the host device 10 side. FD
D12 is a storage device for storing a database (here, a plurality of business card data). RAM13 is
It stores various data required for work on the host device 10 side. The keyboard 14 is an input device for inputting data such as characters. The display device 15 is for displaying data such as characters, and is, for example, a CRT (Cathode R).
ay Tube) Display device or LCD (Liquid Crystal Disp)
lay) device.

The character recognition device 20 includes a CPU 21 and a RAM.
22, a system ROM 23, an image scanner 24, and a character recognition unit 25. The CPU 21 uses the character recognition device 2
Performs business card reading / recognition control on the 0 side. The RAM 22 stores various data necessary for work on the character recognition device 20 side. As shown in FIG. 2, the system ROM 23 includes a program area 23a for storing a business card reading / recognition control program and a dictionary area 23b for storing a first and last name dictionary.
Have.

Here, the surname dictionary is as shown in FIG.
This is a dictionary in which a name (surname) is associated with its reading. When there are a plurality of readings for the same name, each reading is registered with a predetermined priority. For example, as shown in FIG. 3, for the name "East", "Higashi",
Three readings of "Azuma" and "Azuma" are registered together with information indicating priority (indicated by numbers in parentheses). This priority order is predetermined according to the frequency of reading.

The image scanner 24 is an image input device for reading characters and figures written on a sheet as image data, and is used here for reading the image data of a business card. The character recognition unit 25 is a device for recognizing characters, and here, the image scanner 2
Various character strings (company name, name, address, telephone number, FAX number, etc.) written on the business card are recognized from the image data of the business card input through 4.

Next, the operation of the first embodiment will be described. Here, on the character recognition device 20 side, a business card is read,
The operation of estimating the reading of the name written on the business card will be described.

FIG. 4 is a flow chart showing the operation of the business card reading process according to the first embodiment. First, the image data of a business card is read through the image scanner 24 (step S11). In this case, the image scanner 24 applies light to the reading surface of the business card and reflects the reflected light on the CCD (C
image data is generated by converting it into a digital signal using a harge coupled device).

In the character recognition unit 25, various character strings (company name,
(Name, address, telephone number, FAX number, etc.) are recognized (step S12). The character recognition method is OC
A method generally used in R (Optical Character Reader) or the like is used, and the description thereof is omitted here.

The CPU 21 understands the item of the character string recognized by the character recognition unit 25 (step S13).
The item comprehension is a work for classifying the recognized character strings into items (company name, name, address, telephone number, FAX number, etc.). This item understanding is performed based on information such as keywords, character string positions, and character sizes. That is, for example, the item is understood based on keywords such as "company", "address", "phone number", "TEL", and "FAX" for fax number. For example, if the business card is written vertically, the name is in the center, the company name is in the upper right, the address or telephone number, the fax number, etc. is the position of the character string, and the name is compared to other items. Understand items based on font size such as is large.

When the name entered on the business card is recognized as a result of such item understanding (Yes in step S14), the CPU 21 estimates the reading of the name (step S15).

Here, the estimation process according to the first embodiment will be described. FIG. 5 is a flowchart showing the operation of the estimation process according to the first embodiment. The CPU 21 accesses the dictionary area 23b of the system ROM 23 and refers to the surname and surname dictionary stored therein to check the first three characters of the name in the surname and surname dictionary (step S21). As a result, if the corresponding name exists in the surname and surname dictionary (Yes in step S22), the CPU 21 retrieves the first candidate reading from the surname and surname dictionary (step S23). If the corresponding name does not exist in the surname dictionary (N in step S22).
o), the CPU 21 does not read the name (step S24).

More specifically, for example, when reading a business card as shown in FIG. 9A, when the character string in the line L1 is recognized as the full name by the item understanding, the surname and surname dictionary shown in FIG. Search for the name "East". In this example, the first reading is "Higashi" for "East".
It is registered as a candidate. Therefore, "Higashi" is obtained as the reading of the name of the business card.

In this way, when the reading of the name for the business card is estimated, the CPU 21 recognizes the reading of the name for each item (company name, name, address, telephone number, F).
It is output to the host device 10 together with the AX number and the like) (step S16). If there is a next business card, the reading process is repeated in the same manner as above (step S17).

On the host device 10 side, these data are registered in the database stored in the FDD 12.
In this case, since the reading of the name of the business card has already been obtained by the above estimation process, it can be automatically registered in the database without the user's own input. Therefore, the burden on the user in the business card reading work can be reduced.

If the reading of the name cannot be estimated (if it is not registered in the surname dictionary) or if it cannot be estimated correctly, the host device 10 side
It is assumed that the correct reading is input through the keyboard 14.

(Second Embodiment) Next, a second embodiment of the present invention will be described. In addition, in the second embodiment, the configurations shown in FIGS. 1 to 4 are the same as those in the first embodiment, and therefore only the name reading estimation process will be described here.
In the second embodiment, similar to the first embodiment, the reading of the name is estimated using the surname dictionary, but at that time, the readings of all the candidates registered in the surname dictionary are extracted. And

FIG. 6 is a flow chart showing the operation of the estimation processing according to the second embodiment. In the second embodiment,
First, the CPU 21 first determines the dictionary area 2 of the system ROM 23.
3b is accessed, and the first three characters of the name are checked in the first and last names dictionary by referring to the first and last names dictionary stored therein (step S31). As a result, if the corresponding name exists in the surname and surname dictionary (Yes in step S32), the CPU 21 retrieves all readings from the surname and surname dictionary (step S33).
If the corresponding name does not exist in the surname dictionary (No in step S32), the CPU 21 processes it without reading the name (step S34).

More specifically, for example, when reading a business card as shown in FIG. 9A, when the character string in the line L1 is recognized as the full name by the item understanding, the surname and surname dictionary shown in FIG. Search for the name "East". In this example, the first reading is "Higashi" for "East".
Candidate, "Azuma" is the second candidate, "Azuma"
Is read as the third candidate. Therefore, as the reading of the name of the business card, there are three readings of "Higashi", "Azuma", and "Azuma".

The name reading thus obtained is output to the host device 10. In this case, since a plurality of readings are output as candidates, on the host device 10 side,
It is necessary to select a desired reading from each of the candidates with the keyboard 14, but it is not necessary to input the reading itself. Therefore, the burden on the user in the business card reading work can be reduced.

As a method of displaying each candidate, first,
After displaying the first-ranked candidate, there is a method of displaying the other candidates one by one in the priority order according to the next candidate instruction, or a method of displaying all the candidates collectively in the priority order.

When the reading of the name cannot be estimated (when it is not registered in the surname dictionary) or when the desired reading is not found in each candidate, the keyboard 14 on the host device 10 side. Enter the correct reading through.

(Third Embodiment) Next, a third embodiment of the present invention will be described. Since the configuration shown in FIGS. 1 to 4 in the third embodiment is the same as that in the first embodiment, only the name reading estimation process will be described here. The third embodiment is characterized in that the reading of the name is estimated by recognizing a character string near the name as ruby.

FIG. 7 is a flow chart showing the operation of the estimation processing according to the third embodiment. In the third embodiment,
First, the CPU 21 searches for a character string consisting of hiragana, katakana, or romaji within a predetermined range around the recognized name (step S41). As a result, when there is a corresponding character string (Yes in step S42), the CPU 21 determines the character string as ruby added to the name, and determines the reading according to the type of the character string as follows. Yes (step S43).

If the type of character string is hiragana, the CPU
21 directly processes the character string as the reading of the name (step S46). If the type of the character string is katakana, the CPU 21 converts the katakana into hiragana (step S44), and then uses this character string as the reading of the name (step S46). In this case, the conversion from katakana to hiragana is performed by exchanging the codes of the upper two digits. That is, for example, the hexadecimal code of “A” is “2
522 ”, the hexadecimal code for“ A ”is“ 2 ”
422 ”, and the codes of the upper two digits are different. Therefore, katakana can be easily converted into hiragana simply by changing the code of the upper two digits from "25" to "24".

If the type of character string is Roman letters,
The CPU 21 converts the Roman character into hiragana (step S45), and then uses this character string as the reading of the name (step S46). In this case, the conversion from romaji to hiragana is performed using a correspondence table of hiragana and romaji (not shown).

On the other hand, if there is no corresponding character string in step S42 (No in step S42),
The CPU 21 processes as if the name is not read (step S47).

More specifically, for example, when reading a business card as shown in FIG. 9B, when the character string of the line indicated by L1 is recognized as the name by the item understanding, it is in the vicinity of the line L1. The character string in the line indicated by L2 is recognized as ruby. In this example, the character string in the row L2 is written in hiragana as "Azuma". Therefore, "Azuma" is obtained as the reading of the name of the business card.

In this way, a character string having a name close to it is recognized as a ruby, and the same character string is output to the host device 10 as a reading of the name. Therefore, if it is a business card in which ruby is added to the name in advance, the reading of the name can be correctly estimated.

When the reading of the name cannot be estimated (when the ruby cannot be recognized), the host device 10
On the side, the correct reading should be entered through the keyboard 14.

(Fourth Embodiment) Next, a fourth embodiment of the present invention will be described. Since the configuration shown in FIGS. 1 to 4 in the fourth embodiment is the same as that in the first embodiment, only the name reading estimation process will be described here. In the fourth embodiment, similar to the third embodiment, the character string near the name is recognized as ruby, but when ruby cannot be recognized, the reading of the name is estimated using the surname and surname dictionary. It is characterized by

FIG. 8 is a flow chart showing the operation of the estimation processing according to the fourth embodiment. In the fourth embodiment,
First, the CPU 21 searches for a character string consisting of hiragana, katakana, or romaji within a predetermined range around the recognized name (step S51). As a result, when there is the corresponding character string (Yes in step S52), the CPU 21 determines the character string as ruby attached to the name, and determines the reading as follows according to the type of the character string. Yes (step S53).

If the type of character string is hiragana, the CPU
21 directly processes the character string as the reading of the name (step S56). If the type of the character string is katakana, the CPU 21 converts the katakana into hiragana (step S54), and then makes this character string the reading of the name (step S56). If the type of the character string is Roman letters, the CPU 21 converts the Roman letters into Hiragana (step S55), and then makes this character string the reading of the name (step S56).

If there is no corresponding character string in step S52 (N in step S52).
o), the CPU 21 is the dictionary area 23 of the system ROM 23.
b is accessed, and the reading of the name is estimated by referring to the surname dictionary stored therein (step S57). In this case, the first three characters of the name are checked in the surname and surname dictionary, and if the corresponding name is present in the surname and surname dictionary, the first candidate reading is retrieved from the surname and surname dictionary as in the first embodiment, or It is assumed that the readings of all the candidates are extracted as in the second embodiment.

In this way, first, the ruby is recognized and the reading of the name is estimated, but if the ruby cannot be recognized,
Estimate the reading of a name using a surname dictionary. Therefore,
Even when a business card in which ruby is added to the name and a business card without ruby are mixedly read, the reading of the name can be correctly estimated.

If the reading of the name cannot be estimated or cannot be correctly estimated using the surname dictionary, the correct reading is input through the keyboard 14 on the host device 10 side.

In each of the above-mentioned embodiments, the case has been described in which a business card is set as an object to be read and the reading of the name written on the business card is estimated. However, the present invention is not limited to the business card, and any form in which the name is written can be used. By using the sheet as the reading target, the reading of the name can be estimated by the same method as described above.

In the case of estimating the reading of the name, the reading was estimated for the "surname" of the name, but the reading of the "name" as well as the "surname" should be estimated. Can be easily realized by using a surname and surname dictionary in which both “surname” and “name” are registered, or by recognizing ruby added to both “surname” and “name”.

Although the reading of the name is estimated by "Hiragana", the present invention is not limited to the character type of the reading, and it may be estimated by other character types such as "Katakana" and "Romaji". It is possible.

[0050]

As described above, according to the present invention, the name written on the paper such as a business card is recognized, and the reading corresponding to the name is taken out from the surname dictionary and output as the estimation result. Even when the user does not input the reading of the name when reading, the reading can be automatically registered. This can reduce the burden on the user in the business card reading work.

Further, since the name written on the paper such as a business card is recognized, the character string near the name is recognized as ruby information, and the same character string is estimated as the reading for the name. When reading a business card to which is added, the reading of the name can be correctly estimated.

When a name written on a paper such as a business card is recognized and a character string is detected near the name, the same character string is output as a reading for the name, and the character string is detected near the name. When it is not possible, the full name dictionary is used to estimate the reading of the full name, so even when reading both a business card with ruby added to the name and a business card without ruby, the correct reading of the full name is estimated. can do.

[Brief description of drawings]

FIG. 1 is a block diagram showing a device configuration according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of a ROM shown in FIG.

FIG. 3 is a diagram showing a configuration of a first and last name dictionary stored in the ROM.

FIG. 4 is a flowchart showing an operation of a business card reading process according to the first embodiment.

FIG. 5 is a flowchart showing an operation of an estimation process according to the first embodiment.

FIG. 6 is a flowchart showing an operation of estimation processing according to the second embodiment.

FIG. 7 is a flowchart showing the operation of the estimation process according to the third embodiment.

FIG. 8 is a flowchart showing the operation of the estimation process according to the fourth embodiment.

FIG. 9 is a diagram showing an example of a business card used as a reading target in each of the embodiments.

[Explanation of symbols]

10 ... Host device, 11 ... CPU, 12 ... FDD, 13
... RAM, 14 ... Keyboard, 15 ... Display device, 20 ...
Character recognition device, 21 ... CPU, 22 ... RAM, 23 ... System ROM, 23a ... Program area, 23b ... Dictionary area, 24 ... Image scanner, 25 ... Character recognition unit, 3
0 ... Cable.

Claims (11)

[Claims] 1. A sheet on which a name is written, a character recognition means for recognizing the name written on this sheet, a dictionary storage means for storing a first and last name dictionary in which the name and its reading are associated, and this dictionary. Referring to the first and last name dictionary stored in the storage means,
A character recognition device, comprising: an estimation unit that estimates the reading for the name. 2. A plurality of readings for the same name are registered in the surname and surname dictionary with a predetermined priority order, and the estimating means reads out the first candidate from the surname and surname dictionary according to the priority order. 2. The character recognition device according to claim 1, wherein 3. A plurality of readings for the same name are registered in the surname and first name dictionary with a predetermined priority, and the estimating means reads all candidates from the surname and first name dictionary according to the priority. 2. The character recognition device according to claim 1, wherein 4. A sheet on which a name is written, a character recognition unit for recognizing the name written on the sheet, a ruby detecting unit for detecting ruby information added in the vicinity of the name, and a ruby detecting unit. A character recognition device comprising: an estimation unit that estimates the ruby information detected by the unit as a reading for the name. 5. A sheet on which a name is entered, a character recognition means for recognizing the name on the sheet, a ruby detecting means for detecting ruby information added to the name, and a name and its reading. When the ruby information is detected by the dictionary storage means that stores the first and last name dictionary corresponding to, the ruby information is estimated as the reading for the name, and the ruby information is not detected. In this case, the character recognition device is provided with an estimating means for estimating a reading for the name by referring to the surname dictionary stored in the dictionary storage means. 6. The character recognition device according to claim 5, wherein the ruby detecting means detects, as ruby information, a character string that is in the vicinity of the name and consists of only hiragana, katakana, or romaji. 7. The surname and surname dictionary stores a plurality of readings for the same name with a predetermined priority, and the estimating means stores the surname and surname dictionary in the surname and surname dictionary when the ruby information is not detected. The character recognition device according to claim 5, wherein the reading of the first candidate rank is taken out from the inside according to the priority order. 8. A plurality of readings for the same name are registered in the surname and surname dictionary with a predetermined priority, and the estimating means stores the surname and surname dictionary in the surname and surname dictionary when the ruby information is not detected. The character recognition device according to claim 5, wherein all candidate readings are taken out from the inside according to the priority order. 9. A reading estimation method, wherein when a name written on a form is recognized, a reading for the above name is estimated by using a surname and first name dictionary in which the name and its reading are associated with each other. . 10. When the name written on the form is recognized, ruby information added in the vicinity of the name is detected, and this ruby information is estimated as a reading for the name. Reading estimation method. 11. When recognizing a name written on a form, the ruby information added to the name is detected, and when the ruby information can be detected, the ruby information is estimated as a reading for the name. Then, when the ruby information cannot be detected, a reading estimation method for estimating the reading for the name by using a surname and first name dictionary in which the name and the reading are associated with each other.