JPH0612522A - Character recognizing device - Google Patents

Abstract

PURPOSE:To efficiently process characters, which can not be recognized, even without preserving all the images read from an original. CONSTITUTION:An image read part 3 reads the images of character information such as printed or handwritten characters or symbols as image data, a character extraction part 4 extracts respective character parts from those read image data, and a character position detection part 5 detects the positions of those respective extracted character parts. Then, a character recognition part 6 recognizes characters from the image data of the respective extracted parts, an unrecognizable character information preservation part 9 preserves the image data, which characters can not be recognized, and the position information as unrecognizable character information, and the recognized characters are outputted as recognized character information after being allocated character codes corresponding to those characters by a character code preparation part 7.

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 for reading an image of character information such as characters and symbols printed or handwritten on a recording medium such as a manuscript as image data and performing character recognition from the read image data. Regarding

[0002]

2. Description of the Related Art In recent years, an optical character reading device (OCR) which is used by connecting to various data processing devices (computer body) including a document creating device, an automatic translation device, etc., and a system device such as a filing device has been widely used. There is. Its OCR
Is to read an image of character information such as characters and symbols printed or handwritten on a recording medium such as an original as image data. There is a character recognition device that recognizes characters from image data read by OCR.

For example, there are the following devices. (1) Recognize the characters input by handwriting from the tablet, identify the recognized characters and convert them into an output code of a system that can be distinguished from the output code from the keyboard, and from the output code and the keyboard via the adapter. A character input device that integrates the output code of the above and outputs to the control processing main body (for example, refer to Japanese Patent Laid-Open No. 61-223974).

(2) An information input device in which a handwritten OCR is connected to an information input device capable of inputting information from a voice or a keyboard and when inputting unreadable Chinese characters or the like, handwriting input by the handwritten OCR is possible. (See, for example, Japanese Utility Model Laid-Open No. 62-62368).

(3) Instead of key input on the keyboard or push button input on the panel, a handwritten input character or an instruction input by voice is recognized, and the result is registered in advance in a registration table in accordance with the recognition result. An information processing apparatus for selecting the processing that is performed (see, for example, Japanese Utility Model Laid-Open No. 62-187340).

(4) A keyboard for outputting character and numeral codes by pressing a key, and a hand scanner and an OCR character read by the hand scanner for generating character and numeral codes and outputting from the keyboard. A personal computer input device that has a character and number recognition unit that outputs a logical sum, and handles the input from both the keyboard and the hand scanner as the same input (for example, JP-A-64
-50114).

In the conventional character recognition apparatus, when an unrecognizable character is generated, the operator can manually correct the original image read from the original document so that all the original image read from the original document can be corrected. I had saved it. Alternatively, all original images read from a document are stored on the side of the system device to which the character recognized by the character recognition device is output.

[0008]

However, in order to store all the images read from the original in the character recognition device as described above, it is necessary to provide a memory having a large storage capacity, and the character recognition device is downsized. In addition, there is a problem that hinders cost reduction. Further, if all the images read from the original are saved on the system device side, the storage area of the memory of the system device is used for saving the images, and the system device can use the memory efficiently. There was a problem that I could not do it.

Further, when a character that cannot be recognized occurs during the character recognition processing, a specific character code is made to correspond to the character code of the unrecognizable character, and the specific character code is recognized by the system device instead of the unrecognizable character. It is output as character information, and the system device receives the specific character code and processes it as a character that could not be recognized. Then, the operator had to refer to the original image on the side of the system device and make a manual correction, and there was a problem that the system device could not efficiently process the characters that could not be recognized.

The present invention has been made in view of the above points, and an object of the present invention is to enable efficient processing of characters that cannot be recognized without storing all the images read from a document. .

[0011]

In order to achieve the above-mentioned object, the present invention provides an image reading means for reading an image of character information such as printed or handwritten characters and symbols as image data, and an image read by the means. Character extracting means for extracting each character portion from data, character position detecting means for detecting the position of each character portion extracted by the means, and character recognition for performing character recognition from image data of each portion extracted by the character extracting means Means, a character code creating means for assigning a character code corresponding to the character recognized by the means, a character information output means for outputting the character code assigned by the means as recognized character information, and a character by the character recognizing means. Save unrecognized image data and its position information as unrecognizable character information Providing a character recognition apparatus having a unrecognized character information storage means that.

Further, the character code creating means is provided with means for assigning a specific character code including position information to the image data which cannot be recognized by the character recognizing means, and the character information outputting means provides the specific character. The code may also be output as the recognized character information. Further, it is preferable that the character information output means outputs, together with the specific character code, the image data in which the character cannot be recognized as the recognized character information.

[0013]

A character recognition device according to the present invention reads an image of character information such as printed or handwritten characters and symbols as image data, extracts each character portion from the read image data, and extracts each extracted character portion. Detect the position of. Then, character recognition is performed from the extracted image data of each part, and the image data in which character recognition cannot be performed and the position information thereof are saved as unrecognizable character information,
Characters that have been recognized can be assigned a character code corresponding to the character and output as recognized character information. Therefore,
It is possible to distinguish characters that could not be recognized without storing all the read image data.

Further, if a specific character code including the position information is assigned to the image data in which the character cannot be recognized and the specific character code is also output as the recognized character information, the character is output at the output destination. Characters that cannot be recognized can be processed separately. Furthermore, if the image data that cannot be recognized as a character together with the specific character code is output as the recognized character information, more detailed processing can be performed on the character that cannot be recognized.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the present invention will be specifically described below with reference to the drawings. FIG. 1 is a block diagram showing a functional configuration of a character recognition device according to an embodiment of the present invention together with a system device.

The character recognition device 1 is connected to the system device 2 via an interface cable 1a.
The system unit 2 has a built-in microcomputer including a CPU, a ROM, a RAM, etc., and performs various processes such as document creation based on the character information sent from the character recognition unit 1.

The character recognition device 1 includes a CPU, R (not shown).
An image reading unit 3, a character extracting unit 4, a character position detecting unit 5, a character recognizing unit 6, a character code creating unit 7, a character information output unit 8, and an unrecognizable character are built-in with a microcomputer including an OM and a RAM. Each function of the information storage unit 9 is provided.

The image reading unit 3 is an image input means such as a scanner, and optically reads an image of character information such as characters, symbols, numbers printed or handwritten on a recording medium such as a document as image data. The character extracting unit 4 includes the image reading unit 3
Each character portion is extracted from the image data read by.

The character position detecting unit 5 detects the position of each character portion extracted by the character extracting unit 4. Character recognition unit 6
Performs character recognition from the image data of each part extracted by the character extracting unit 4. The character code creation unit 7 assigns a character code corresponding to the character that can be recognized by the character recognition unit 6. In addition, a specific character code is assigned to a character that cannot be recognized.

The character information output unit 8 is a character code creation unit 7
The character code assigned by or the specific character code is output as recognition character information. The unrecognizable character information storage unit 9 stores the image data that cannot be recognized by the character recognition unit 6 and its position information as unrecognizable character information.

Next, the operation between the above-mentioned parts will be described. First, the image reading unit 3 optically reads the image data of the image of the character information on the original, and the dot pattern of the read image data is used by the character extracting unit 4
Send to.

When the character extraction unit 4 extracts a character portion from the dot pattern, it sends the character extraction information to the character position detection unit 5. The character position detection unit 5 detects the position of each character portion based on the character extraction information from the character extraction unit 4, and sends the position information to the character extraction unit 4. Then, the character extraction unit 4 sends the dot pattern for each character and its position information to the character recognition unit 6.

The character recognition unit 5 performs character recognition based on the dot pattern of each character received from the character extraction unit 4, and the dot pattern of the character that cannot be recognized and its position information are unrecognizable character information as unrecognizable character information. The unrecognized character information is stored in the storage unit 9 and sent to the character code creation unit 7.

The character code creating section 7 assigns a specific character code to the image data which cannot be recognized by the unrecognizable character information and sends it to the character recognizing section 5. On the other hand, the dot pattern of the character that has been recognized can be sent to the character code creation unit 7, and the character code creation unit 7 assigns the character code corresponding to the dot pattern of the character that has been recognized and sends it to the character recognition unit 5.

Then, the character recognition unit 5 sends the character code or the specific character code received from the character code creation unit 7 to the character information output unit 8, and the character information output unit 8 sends these character codes via the interface cable 1a. Output to the system unit 2.

Since a known technique may be used for the process of extracting the character portion from the dot pattern and the process of detecting the character position in the character position detecting section 5, detailed description thereof will be omitted.

Next, the character recognition processing in the character recognition unit 6 will be described with reference to the flowchart shown in FIG. First, the dot pattern of a character and its position information are received from the character extraction unit 4 and character recognition processing is performed to determine whether or not the character has been recognized.

If the character can be recognized, the character code creating section 7
The character code for the recognized character is assigned by and output to the character information output unit 8. If the character cannot be recognized, the character position is determined based on the position information from the character position detection unit 5, and the character position (position information) and the dot pattern are stored in the unrecognizable character information storage unit 9 and the character cannot be recognized. A specific character code is assigned to the selected character by the character code creating unit 7, and the specific character code is output to the character information output unit 8. Then, the presence or absence of the next data (dot pattern) is determined, and if there is, the process returns to the first process and the above process is repeated. If not, the process of this character recognition ends.

Next, a specific example of character recognition processing will be described. FIG. 3 is a diagram showing an example of a manuscript in which a plurality of English characters are handwritten. In this manuscript St, 21 English characters are arranged in 3 rows and 7 columns. This manuscript S
When the image of t is read, the character recognition unit 6 performs character recognition based on the dot pattern for each character from the character extraction unit 4.

When the character recognition unit 6 normally recognizes the alphabetic characters "A" to "J" in order, each character code is assigned by the character code creation unit 7, and those character codes are output through the character information output unit 8. Output to the system unit in order.

Thereafter, for example, if the character in the broken line frame Im cannot be recognized, the position of the character that cannot be recognized based on the position information from the character position detection unit 5, that is, the 4th column and the 2nd line in the manuscript. Determine that it is a character. Next, the position information is stored in the unrecognizable character information storage unit 9 together with the image data in the broken line frame Im, a specific character code is assigned to the unrecognizable character, and the system information is sequentially output through the character information output unit 8. Output to. And
The recognition process of the English characters after "L" is continued.

For example, the character code creating section 7 in this embodiment is provided with means for assigning a specific character code including position information to image data which cannot be recognized by the character recognizing section 6, and a character information output section is provided. The specific character code is also output by 8 as the recognized character information.

In this case, the format of a specific character code for a character that cannot be recognized is as follows. Specific character code: "FFXY" This character code is expressed in hexadecimal notation, "FF" is a specific code for a character that cannot be recognized, "X" is the number of columns of position information of the character that cannot be recognized, and "Y" is Represents the number of lines of position information of a character that cannot be recognized.

Therefore, for example, when the original document shown in FIG. 3 is read and character recognition is performed, the character code of the character that is recognized is output to the system unit. On the other hand, if the character cannot be recognized, the positional information is stored together with the image data, and the specific character code including the positional information for the unrecognizable character is output.

For example, when the character within the broken line frame Im cannot be recognized, a specific code "FF" indicating that the character cannot be recognized and a code "4" and 2nd column indicating the position information thereof are displayed. A specific character code "FF42" in which "42" is combined is assigned and output to the system device.

Further, the character information output section 8 may output the specific character code as well as the image data whose character cannot be recognized as the recognized character information. The data format including a specific character code for a character that cannot be recognized and image data that cannot be recognized in that case is shown. Format: "FFXY, xxyy, (image data)"

This data is represented by a hexadecimal number and is "FF".
Is a specific code for characters that cannot be recognized, "X"
Represents the number of columns of position information of characters that cannot be recognized, and "Y" represents the number of lines of position information of characters that cannot be recognized. Further, "xxyy" represents the image data size of a character that cannot be recognized, and the image data is added after that. In other words, if the character cannot be recognized, the position information is stored together with the image data, and the specific character code including the position information for the unrecognizable character and the image data thereof are output.

FIG. 4 is a diagram showing a dot pattern of image data of characters in the broken line frame Im shown in FIG. In this case, a vertical and horizontal 32 × 32 bit area, that is, 1 to 12
A dot pattern of a character that cannot be recognized is developed in each byte up to 8, and the image data size is “20H × 20H”.

For this character, a specific code "FF" indicating that the character cannot be recognized, a code "42" indicating the 4th row and 2nd column of the position information, and the image data size of the character that cannot be recognized The code “2020” and the image data “0000, ..., 0000” are combined. Therefore, "FF42,202,0000, ...
.., "0000" data is output to the system unit.

Next, the effects of the above-described embodiments will be listed. (1) By storing not only the image data of the character that could not be recognized but also the position information of the character, it is possible to distinguish multiple image data that could not be recognized. You don't have to save it. Therefore, the memory capacity of the unrecognizable character information storage unit can be small, and the device can be downsized and the cost can be reduced.

(2) Since the specific character code corresponding to the character that cannot be recognized includes the position information of that character, when the character that cannot be recognized on the system unit side is corrected, the character recognition unit It is not necessary to send all the image data before the character extraction to the system device, and the system device can receive and correct only the image data of the character that cannot be recognized by the distinguished character. Therefore, the data transfer time from the character recognition device to the system device can be shortened, and the storage area of the memory on the system device side can be used efficiently.

(3) Since character position information is included in a specific character code corresponding to a character that could not be recognized, and the dot pattern amount and dot pattern information of the image data are added, the character cannot be recognized. Detailed information about the selected characters can be output to the system unit. Therefore, the system device can register the information whose character cannot be recognized as an external character, and process it in the same manner as the character code that can be recognized by the external character code registered.

That is, even when the character recognition device outputs information of a character that cannot be recognized to the system device, it is not necessary to process the recognized character data after the operator corrects the character data. Character data can be efficiently processed by the processing.

Next, an embodiment of a device to which the character recognition device according to the present invention is applied will be described. FIG. 5 is an external perspective view of a data processing device using an input device including the character recognition device described above. In this data processing device, the input device 11 is a keyboard interface (I / F) cable 1
2 is connected to a system device 13 similar to the system device 2 of FIG. 1 having a document creation function, a data processing function, etc., and the input device 11 has a sheet (paper or the like) inserted through the insertion port 11a. ) It has a function of optically reading information such as characters printed or handwritten on it and converting it into a keyboard signal.

FIG. 6 is a block diagram showing the functional configuration of the input device 11 shown in FIG. 5 together with the system unit 13.
The input device 11 is the same as the character recognition device 1 shown in FIG. 1, except that the keyboard operation code creation unit 16, the keyboard operation code creation information storage unit 17, the keyboard signal creation unit 18, the keyboard signal creation information storage unit 19, and the keyboard signal output are provided. The part 20 is added.

The keyboard operation code creation unit 16 uses the information stored in the keyboard operation code creation information storage unit 17 to recognize each recognition code input from the character recognition device 1 as a keyboard operation code (for key operation). It is converted into a corresponding code) and sent to the keyboard signal creation unit 18.

Keyboard operation code creation information storage unit 17
Stores the information referred to by the keyboard operation code creation unit 16 in a table format. For all information such as characters recognized by the recognition code from the character recognition device 1, these recognition codes are used as indexes. The corresponding keyboard operation code is registered.

The keyboard signal creation unit 18 uses the keyboard operation code sent from the keyboard operation code creation unit 16 based on the information stored in the keyboard signal creation information storage unit 19 to generate a keyboard signal (key on the keyboard). The key code at the time of pressing and the corresponding ON / OFF signal are converted) and sent to the keyboard signal output unit 20.

The keyboard signal creation information storage unit 19 stores the information referred to by the keyboard signal creation unit 18, and the keyboard signal data corresponding to each keyboard operation code (one key input) on the keyboard device is stored in a table. It is registered in the format.

The keyboard signal data is, for example,
When the character "A" is input from the keyboard device to the system device, the keyboard device inputs the key to the system device along with the physical position (key code) of the character "A" in the key arrangement on the keyboard device. Data that indicates that the key is pressed (on) and released (off) is notified, and is notification information regarding key press and key release determined for each key unit on the keyboard device.

The keyboard signal output unit 20 outputs the keyboard signal data received from the keyboard signal creation unit 18 to the system unit 13 via the keyboard I / F cable 12 for exchanging data between the system unit 13 and the keyboard unit. To do.

Further, the character recognition apparatus 1 assigns a specific character code (for example, "FFXY" described above) including the position information to a character that cannot be recognized and outputs it as a recognition code. The code creating unit 16 creates a keyboard operation code for the recognition code, and the keyboard signal creating unit 18 creates a corresponding keyboard signal by the keyboard operation code,
The keyboard signal is output to the system unit 13. Therefore, the system device 13 side can know the existence of the unrecognizable character and its position information.

FIG. 7 is a diagram showing an example of a storage format of keyboard operation code creation information, and FIG. 8 is a diagram showing an example of a storage format of keyboard signal creation information. The keyboard operation code creation information consists of recognition data and keyboard operation data. The recognition data is information such as characters and symbols corresponding to each key on the keyboard device (for example,
"0", "A", "A", etc.), and the keyboard operation data is information on keyboard operation for inputting the respective information.

For example, the keyboard operation data for inputting “0” corresponding to the recognition data “0” has the number of key operations = 2 (indicating that two key operations are required) as the operation data control information. , A pointer pointing to the "alphanumeric" key in the keyboard signal creation information indicating the operation of the "alphanumeric" key, and a pointer pointing to the "0" key in the keyboard signal creation information indicating the operation of the "0" key.

Further, the keyboard operation data for inputting "A" corresponding to the recognition data "A" is key operation number = 2 as operation data control information and keyboard signal creation information indicating operation of the "Katakana" key. From the pointer that points to the "Katakana" key and the pointer that points to the "3" key in the keyboard signal creation information that indicates the operation of the "3" key (since the "A" key also serves as the "3" key) Become.

The keyboard signal generation information is composed of keyboard signals when each key is pressed and released, as shown in FIG.

Next, the operation of each part of the input device 11 shown in FIG. 6 will be described. First, the character recognition unit 15 outputs a pointer pointing to keyboard operation data corresponding to the corresponding character in the keyboard operation code creation information storage unit 17 as a result of the recognized character.

As shown in FIG. 9, this keyboard operation data has the number of key operations as the operation data control information and the operation information of the key as the operation data. It is the number of key operations when inputting from the device, and the operation data corresponds to the input operation procedure, and the key operations (types of input keys) are arranged by the number of key operations in the order of input. A pointer pointing to the corresponding key stored in the keyboard signal creation information storage unit 19 is set in the key operation information.

Upon receiving the pointer from the character recognition device 1, the keyboard operation code creation unit 16 retrieves the number of key operations and operation data corresponding to the pointer, and operates the key operation information (keyboard in order from the operation data by the number of key operations). A pointer to signal creation information) is input and output to the keyboard signal creation unit 18.

The keyboard signal generation unit 18 sequentially outputs the code of the corresponding key and the keyboard signal indicating the pressing and releasing of the key to the keyboard signal output unit 20 by the pointer from the keyboard operation code generation unit 16. And
The keyboard signal output unit 20 outputs the keyboard signal to the system unit 13 via the keyboard I / F cable 12.

Next, the process of inputting information by the input device 11 will be described with reference to the flowchart of FIG. First, a character is input by the character recognition device 1, the character recognition processing is performed, and the keyboard operation code creation unit 16
Send the pointer to.

Upon receiving the pointer, the keyboard operation code creating section 16 obtains the number of key operations Kt of the corresponding data pointed to by the pointer from the keyboard creating code information,
The number Kn of key operations taken out is set to zero. Compare the number Kt of key operations with the number Kn of key operations taken out,
If they are equal, it is determined whether or not the input is completed, and if YES, the process ends, but if NO, the character input process is repeated again.

On the other hand, if Kt and Kn are not equal, 1 is added to Kn by the keyboard signal creating unit 18 to obtain the operation of the Kn-th key from the operation data, and the pointer pointed to is pressed from the keyboard signal creation information. When the keyboard signal is obtained, the keyboard signal output unit 20 outputs the keyboard signal.

Further, when a keyboard signal for release is obtained from the keyboard signal creation information by the pointer pointed by the operation of the Kn-th key and output by the keyboard signal output unit 20, the process returns to the process of comparing Kt and Kn, and Kt and Kn. The process is repeated until

Next, an example of actual character input processing in the input device 11 will be described with reference to the block diagram of FIG. Here, the case of inputting the katakana “A” will be described. The character recognition device 1 reads the katakana “A” written on the sheet P, and the katakana “A” is read.
Is recognized and the pointer pointing to the recognition data "A" of the keyboard operation creation information is moved to the keyboard operation code creation unit 1
Send to 6.

Then, the keyboard operation code creating section 16
Includes the data of the number of times the key is operated as the keyboard operation data corresponding to the recognition data “A” in the information stored in the keyboard operation code creation information storage unit 17. 'The pointer for key operation is sent to the keyboard signal creation unit 18.

In the keyboard signal creation unit 18, the keyboard signal when the "Katakana" key is pressed in the information stored in the keyboard signal creation information storage unit 19 by the pointer, the keyboard signal when the "Katakana" key is released, and " The keyboard signal when the 3'key is pressed and the keyboard signal when the '3' key is released are output to the keyboard signal output unit 20.

In this way, information such as characters input by means other than key input is output to the system unit as a keyboard signal as if the key input was made from the keyboard unit. Therefore, the system device can process input information from other than the keyboard device as the same data as the key operation input from the keyboard device.

The effects of the input device 11 of this embodiment will be listed below. First of all, as to the improvement of the man-machine interface, (1) it is not necessary to remember the key arrangement, so that the input becomes easy. (2) The skill of key operation becomes unnecessary. (3) Even without using an unfamiliar keyboard device, not only characters but also various control information (for example, half-width, double-width, line feed, page break, conversion, etc.) by handwriting input, character combinations, symbols, You can quickly enter by using the mark. (4) There is no risk of erroneous input unless you make a mistake in entering the characters. (5) It is possible to easily perform a fixed-form input operation such as starting a business program that is used daily, by reading a card on which a series of operation information is printed and starting the apparatus. (6) Even persons with disabilities can easily input characters and the like.

With regard to hardware, (7) Since all input information is input as keyboard signals from the keyboard I / F, there is an increase in the load of different recognition processing on the system unit side depending on the input means. Absent. (8) It can be used as it is for most data processing devices using a general keyboard device and is versatile. (9) The size can be easily reduced without lowering the operability. (10) When applied to a filing device, the amount of data can be small because a code can be input instead of an image in the filing process, and resources such as a memory in the system device can be effectively used.

[0071]

As described above, according to the character recognition apparatus of the present invention, it is possible to efficiently process characters that cannot be recognized without storing all the images read from the document, so that the storage capacity can be increased. Since it is not necessary to provide a large memory, it is possible to reduce the size and cost of the character recognition device. Further, the amount of use of the memory area of the system device, which is the output destination of the character code, can be reduced, and the system device can efficiently use the memory. Further, it is possible to efficiently process characters that cannot be recognized by the system device.

[Brief description of drawings]

FIG. 1 is a block diagram showing a functional configuration of a character recognition device according to an embodiment of the present invention together with a system device.

FIG. 2 is a flowchart showing a character recognition process in a character recognition unit shown in FIG.

3 is a diagram showing an example of a manuscript handwritten with a plurality of English characters to be read by the character recognition device shown in FIG.

4 is a diagram showing a dot pattern of image data of a character within a broken line frame Im shown in FIG.

FIG. 5 is an external perspective view of a data processing device using an input device including a character recognition device according to the present invention.

6 is a block diagram showing an example of a functional configuration of the input device shown in FIG. 5 together with a system device.

FIG. 7 is a diagram showing an example of a storage format of keyboard operation code creation information.

FIG. 8 is a diagram showing an example of a storage format of keyboard signal creation information.

FIG. 9 is a diagram showing an example of a storage format of keyboard operation data.

FIG. 10 is a flowchart showing information input processing of the input device.

FIG. 11 is a block diagram for explaining an actual example of character input processing by the input device.

[Explanation of symbols]

1 character recognition device 1a interface cable 2 system device 3 image reading unit 4 character extraction unit 5 character position detection unit 6 character recognition unit 7 character code creation unit 8 character information output unit 9 unrecognizable character information storage unit St manuscript 11 input device 12 Keyboard interface (I / F) cable 13 System unit 16 Keyboard operation code creation unit 17 Keyboard operation code creation information storage unit 18 Keyboard signal creation unit 19 Keyboard signal creation information storage unit 20 Keyboard signal output unit

Claims (3)

[Claims] 1. An image reading unit for reading an image of character information such as printed or handwritten characters and symbols as image data, a character extracting unit for extracting each character portion from the image data read by the unit, and the unit. The character position detecting means for detecting the position of each character portion extracted by the above, the character recognizing means for recognizing the character from the image data of each portion extracted by the character extracting means, and the character recognized by the means. Character code creating means for allocating a character code, character information output means for outputting the character code assigned by the means as recognition character information, and image data which cannot be recognized by the character recognition means and its position information cannot be recognized. Characteristic information storage means for storing as character information is provided. Character recognition device. 2. The character recognition device according to claim 1, wherein the character code creating means assigns a specific character code including position information to image data that cannot be recognized by the character recognition means. Have,
A character recognition device, wherein the character information output means also outputs the specific character code as recognized character information. 3. The character recognition device according to claim 2, wherein the character information output means outputs, together with the specific character code, image data in which the character cannot be recognized as recognized character information. Character recognizer.