JPH0546804A - Character recognition system - Google Patents

Abstract

PURPOSE:To provide a character recognition system which can obtain a considerably high character recognition rate in the actual situations in which the character recognition rate does not become 100%. CONSTITUTION:The character recognition system consists of a communication network having a terminal and a character recognition device 106. A means which recognizes the character of a character recognition sheet called from the terminal and identifies the character which cannot be recognized is provided for the character recognition device 106. A means which selects the plural candidate characters of the character which cannot be recognized is provided for the character recognition device 106. A means which returns the plural candidate characters to a user and makes the user decide one of the plural candidate characters is provided for the communication network. A means which reedits the decided character as the recognized character in the character recognition sheet is provided for the character recognition device 106.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition system provided with a character recognition device in a communication network, and more particularly, when an unrecognizable character exists in a read character recognition sheet,
The present invention relates to a character recognition method in which candidate characters of unrecognizable characters are presented to a user so that the user can determine an appropriate character.

[0002]

2. Description of the Related Art In a conventional character recognition system in which a character recognition device is provided in a communication network, especially when there is a character that cannot be recognized by the character recognition device, nothing is done as an unrecognizable character or a user is informed of that fact. One of the ways to report was done.

[0003]

However, in the above conventional technique, even if an unrecognizable character occurs, the system does not have a learning function, so that the user is only notified that there is an unrecognizable character each time. , Recognition rate is 100
In the actual situation of the character recognition technology that does not reach the%, no measures have been taken to process unreadable characters to increase the recognition rate of unrecognizable characters.

The present invention has been made to solve the above problems, and an object of the present invention is to achieve a character recognition rate of 10
An object of the present invention is to provide a character recognition system capable of obtaining a very high character recognition rate in the actual situation where the character recognition rate is not 0%.

[0005]

In order to achieve the above object, the present invention provides a character recognition system comprising a communication network having a terminal and a character recognition device, wherein a character recognition sheet called from the terminal is provided. The character recognition device is provided with means for performing character recognition and for identifying unrecognizable characters, and means for selecting a plurality of candidate characters of the unrecognizable characters is provided in the character recognition device, and the plurality of candidate characters are returned to the user and sent to the user. Mostly, the means for determining one of the plurality of candidate characters is provided in the communication network, and the means for re-editing the determined character as a character that can be recognized by the character recognition sheet is provided in the character recognition device. It is a characteristic.

[0006]

According to the above-mentioned means, in the character recognition system, character recognition is performed on the character recognition sheet called from the terminal, unrecognizable characters are identified, and a plurality of candidate characters of the unrecognizable characters are selected. The candidate character is returned to the user, the user is made to decide one from the plurality of candidate characters, and the decided character is re-edited as a character that can be recognized by the character recognition sheet, so that the character recognition rate is 100%. It is possible to obtain a very high character recognition rate in a situation that is not the case.

[0007]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a schematic configuration of a character recognition system including a communication network according to an embodiment of the present invention. In FIG. 1, 101 and 102 are subscriber telephones,
103 is a facsimile terminal, 104 is an electronic computer, 10
Reference numeral 5 is an exchange, 106 is a character recognition device, and 107 is a communication line connecting them.

In the character recognition system comprising the communication network of this embodiment, for example, as shown in FIG. 2, the character recognition sheet is transmitted from the subscriber telephone 101 and the facsimile terminal 103,
By performing character recognition by the exchange 105 and the character recognition device 106 and transmitting the character recognition result to the electronic computer 104, it is possible to perform communication for performing the character recognition function in the communication network.

FIG. 2a shows a character recognition sheet in which characters that the user wants to recognize are entered in the entry frame.
3456, the second column is labeled ABCDEF.

FIG. 2b is a selection sheet returned to the user as a result of character recognition. A dotted frame indicating that 1 and B are unrecognizable characters is described, and candidate characters up to are described below them. One candidate character that could not be recognized is described as l (el), 1 (ichi), i, | (vertical bow), I (eye). Similarly, the candidate characters of B that could not be recognized are described as 8, &,%, B, and R.

As shown in FIG. 2c (selection sheet returned to the character recognition device), the user encloses the character desired by the user, for example, 1 (one) and B among the candidate characters. Mark is added, and the selection sheet with the circle mark is sent again. In addition, # 22518 at the upper right end of each of FIGS. 2b and 2c shows an example of a symbol added by the communication network side for identifying the selection sheet, and the symbol should be recognized by the communication network side. Due to
It can be identified on the communication network side that it is the selection sheet of FIG. 2a.

FIG. 3 is a diagram for explaining the sequence control of the character recognition system of this embodiment shown in FIG. First, the user sends the character recognition sheet to the communication network. In the communication network, character recognition is performed on the received character recognition sheet, and the unrecognized characters are selected,
The candidate characters of the unrecognized characters are selected and displayed in FIG. 2b.
Create a sheet for selecting candidate characters such as and send it back to the user. The user determines the candidate character and transmits the selection sheet, as shown in FIG. 2c. In the communication network, the identification number # 22518 of the selection sheet can be recognized by character to identify it as the response of the selection sheet of FIG. 2a.
Count the number of times the selected character is selected, record and store it. This is so that when the next candidate character is output, it can be printed on the selection sheet in an order that is suitable as a candidate character and easy to select. Next, in the communication network, the character recognition sheet is re-edited with the determined characters to complete the character recognition. It is possible to save the recognition result of the completed character recognition sheet, send it to the electronic computer 104, or process it in a communication network as shown in FIG.

FIG. 4 is a block diagram showing the configuration of the character recognition device of this embodiment. Character recognition device 106 of the present embodiment
Is a facsimile (signal) receiving unit 10 as shown in FIG.
61, character cutout unit 1062, character recognition processing unit 106
3, a dictionary 1064 for character recognition processing, an unrecognizable character editing unit 1065, a facsimile (signal) transmission unit 1066,
The recognition character output unit 1067 is provided. The character recognition processing unit 1063 includes a character recognition unit 10631, a recognized character management unit 10632, an unrecognizable character management unit 10633,
A candidate character selection unit 10634, a determined character determination counter unit 10635, and a determined character re-editing unit 10636 are provided.

When the user inputs a character recognition sheet from the facsimile terminal 103, the facsimile receiving unit 106
Receive with 1. The facsimile signal is sent to the character slicing unit 1062.

The character cutout unit 1062 cuts out a character portion from the received facsimile signal. Various methods have been devised for extracting characters. As an example, the region segmentation method (Akiyama, Masuda: “Region segmentation of document image using marginal distribution, line density, and circumscribed rectangle feature”), SI (d), J69-D, no.8 (1986/8) )).

The character recognition processing unit 1063 receives the facsimile signal cut out by the character cutting unit 1062. Character recognition unit 106 of character recognition processing unit 1063
31 performs character recognition from the extracted characters and the information in the dictionary 1064. The recognized characters are managed by the recognized character management unit 10632. However, if there are unrecognizable characters, the unrecognizable character management unit 1063
Manage with 3. The unrecognized character management unit 10633
Regarding unrecognizable characters, candidate character selection unit 10634
, The candidate characters of unrecognized characters are selected, and the characters to be presented to the user are arranged. The selected candidate character and the received character recognition sheet are sent to the unrecognizable character editing unit 1065.

The unrecognizable character editing unit 1065 uses the selected candidate characters and the received character recognition sheet to make it easy for the user to edit as shown in FIG. Send to.

The user who has received the selection sheet shown in FIG. 2B adds the circles to the characters determined from the candidate characters of this selection sheet, as shown in the example, 1 (ichi) and B. , Retransmit to the communication network. The re-transmitted selection sheet shown in FIG.
1 and the character slicing unit 1062, and is sent to the re-editing unit 10636 by the determined character of the character recognition processing unit 1063. The characters that could not be recognized by the determined character re-editing unit 10636 are replaced with the determined characters, the characters are recognized (coded), and the result of re-editing is added to the recognized character management unit 10632 and managed. in this way,
After allowing the user to determine the unrecognized character, the management of the determined character shifts from the unrecognizable character management unit 10633 to the recognized character management unit 10632. In this way
The recognized character management unit 10632 considers that all the selection sheets shown in FIG. 2A have been recognized, and the recognized characters are all recognized by the recognized character management unit 10632.
067, and the series of character recognition processing ends.

The determined character determination counter unit 1063
5 has a function of counting and storing the number of determined characters designated by the user. This is to determine the order of candidate characters so that the unrecognizable character candidate characters are as easy to understand and easy to use as possible for the user according to the frequency of use, the frequency of unrecognizable characters, and the selected frequency. Is used for. This is because, by presenting the candidate character with high frequency to the user, the probability that the user selects is likely to be in the order from the left, for example.

In the method of this embodiment described above, the selection sheet is sent to the user and the facsimile terminal 10 is sent to the user.
This is a method of specifying in 3. However, the method specified by the user is not limited to this method. As another example, the facsimile signal is the character recognition device 106.
It is also conceivable to specify by voice for a character that cannot be recognized by the character recognition device 106 when the character reaches. For example, "I can't read the characters in the 3rd row and 13th column from the communication network.
Select from the numbers below. l (el), 1
(Ichi), i, | (vertical bow), I (eye),
It is also possible that the user replies "" when said "Other" and recognizes it by voice recognition. At this time, it can be inferred that the character recognition device has substantially the same configuration.

In the above embodiment, the case where the character recognition communication of the facsimile terminal 103 is performed using the public telephone network has been described, but the present invention is the case where the exchange 105 is not used and the character recognition communication is performed using the dedicated line. Needless to say, this can also be applied to.

Although the present invention has been specifically described based on the embodiments, it is needless to say that the present invention is not limited to the above embodiments and various modifications can be made without departing from the scope of the invention. Absent.

[0024]

As described above, according to the present invention, in the actual situation where the character recognition rate is not 100%,
Since an unrecognizable character is returned to the user and the correct character is designated, a very high character recognition rate can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a character recognition system including a communication network according to an embodiment of the present invention,

FIG. 2a is a character recognition sheet in which characters that the user of this embodiment wants to recognize are entered in an entry frame;

FIG. 2b is a selection sheet returned to the user as a result of character recognition in this embodiment;

FIG. 2c is a selection sheet returned to the character recognition device of the present embodiment,

FIG. 3 is a diagram for explaining sequence control of the character recognition system of the present embodiment shown in FIG.

FIG. 4 is a block diagram showing the configuration of a character recognition device of this embodiment.

[Explanation of symbols]

101, 102 ... Subscribed telephone, 103 ... Facsimile terminal, 104 ... Electronic computer, 105 ... Exchange, 106 ... Character recognition device, 107 ... Communication line, 1061 ... Facsimile (signal) receiving section, 1062 ... Character cutting section, 1063
... Character recognition processing unit, 10631 ... Character recognition unit, 1063
2 ... Recognized character management unit, 10633 ... Unrecognizable character management unit, 10634 ... Candidate character selection unit, 10635 ... Determined character determination counter unit, 10636 ... Redetermined character re-editing unit, 1064 ... Dictionary for character recognition processing, 1065
... Unrecognizable character editor, 1066 ... Facsimile
(Signal) transmission unit, 1067 ... Recognized character output unit.

Front Page Continuation (72) Inventor Kobayashi Public notice 1-6 Uchisaiwaicho, Chiyoda-ku, Tokyo Nihon Telegraph and Telephone Corporation

Claims (1)

[Claims] 1. A character recognition system comprising a communication network having a terminal and a character recognition device, wherein the character recognition device is provided with means for performing character recognition on a character recognition sheet called from the terminal and identifying unrecognizable characters. A means for selecting a plurality of candidate characters of the unrecognizable character is provided in the character recognition device, a means for returning the plurality of candidate characters to the user and allowing the user to determine one of the plurality of candidate characters is provided in the communication network. A character recognition system characterized in that the character recognition device is provided with means for re-editing the determined character as a character that can be recognized in the character recognition sheet.