JP2003030584A - Document recognizing device, region identifying method for document image, and program and recording medium therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the character extracting precision. SOLUTION: Whether or not a region other than characters is present in a region identified as a character region attribute is judged, and when it is judged that any region other than characters is present in the region identified as the character region attribute, the region identified as the character region attribute is re-divided. Thus, even when any region other than characters such as seal impressions or graphics enters the region judged as characters as the result of the area division, the region identified as the character region attribute is re-divided so that the region other than characters can be eliminated. Thus, the character extracting precision can be improved.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document recognition apparatus, a document image area identification method, a program and a storage medium for identifying and classifying mixed character areas and areas other than characters in document image data.

[0002]

2. Description of the Related Art Conventionally, various methods are known as a method of identifying a character string or a character area (column) in a document image.

For example, in Japanese Patent Laid-Open No. 06-020092, a blank image is extracted from a document image, and a blank separator formed by the connection of the blank regions is treated as a region dividing line to divide the document image. A method of extracting a character string or a character area (column) inside has been proposed.

There is also known a method of using a projection histogram of black pixels to set a portion having a high black pixel distribution as a range of a character string (Akiyama, Masuda "peripheral distribution, linear density, circumscribed rectangle feature are used together. Area Identification of Document Image ", IEICE Transactions 86/8 Vol J69-D)).

[0005]

By the way, according to the conventional method for extracting a character string or a character area (column) in a document image, in the line cutting processing for character recognition performed after the area identification processing, The processing is performed assuming that the stamp area and the drawing area are not mixed in the character area.

However, in reality, an area other than a character such as a seal stamp or a drawing may enter an area determined to be a character as a result of area division. In such a case, the conventional method cannot perform a line cutting process that cuts out a character line from a character string or a character area (column) in a document image, resulting in a problem that the character extraction accuracy is reduced. It was

An object of the present invention is to improve the accuracy of character extraction.

[0008]

According to another aspect of the present invention, there is provided a document recognition device for identifying and classifying a character area and a non-character area mixed in document image data. Circumscribing rectangle extraction means for extracting a rectangle circumscribing the connected component of black pixels contained in the specified region from the document image data, and whether or not a region other than the character exists in the region identified as the character region attribute. And a non-character area determining means for determining based on the information related to the rectangle extracted by the circumscribing rectangle extracting means, and if there is an area other than a character in the area identified as the character area attribute by the non-character area determining means. If it is determined, the subdivision means for subdividing the area identified as the character area attribute is provided.

Therefore, it is determined whether or not an area other than the character exists in the area identified as the character area attribute, and it is determined that an area other than the character exists in the area identified as the character area attribute. The area identified as the character area attribute is subdivided. As a result, even when an area other than a character such as a seal stamp or a drawing enters the area determined to be a character as a result of the area division, the area identified as the character area attribute is divided again to remove the area other than the character. Since it is possible to eliminate the area, it is possible to improve the character extraction accuracy.

According to a second aspect of the present invention, in the document recognition apparatus according to the first aspect, the information regarding the rectangle in the non-character region determining means is assumed to be black pixels in the inside of the rectangle, and every pixel line. Is the maximum value and the minimum value of the black pixel coordinates.

Therefore, it becomes easy to determine whether or not an area other than a character exists in the area identified as the character area attribute.

According to a third aspect of the present invention, in the document recognition apparatus according to the second aspect, the subdivision means sets a character line and a line space according to a maximum value and a minimum value of coordinates of black pixels for each pixel line. It is determined, and the area identified as the character area attribute is re-divided in this line spacing.

Therefore, the subdivision becomes easy.

According to a fourth aspect of the present invention, in the document recognition apparatus according to the third aspect, it is determined whether or not there is an area other than a character by using information related to the rectangle of the area determined to be the line spacing. To do.

Therefore, since there is no character in the area determined to be the space between lines, it becomes easy to determine whether or not there is an area other than the character.

The invention according to claim 5 is the invention according to claims 1 to 4.
In the document recognition device according to any one of,
An index (certainty factor) indicating the certainty of the result of character recognition is calculated, and the portion with low certainty factor is defined as an area other than the character.

Therefore, it is possible to further improve the division accuracy.

A document image area identification method according to a sixth aspect of the present invention is a document image area identification method for identifying and classifying a mixed character area and non-character area in document image data. A circumscribing rectangle extraction step of extracting a rectangle circumscribing a connected component of black pixels included in the area identified as from the document image data, and whether an area other than a character exists in the area identified as the character area attribute A non-character area determination step of determining whether or not based on the information related to the rectangle extracted by the circumscribing rectangle extraction step, and an area other than a character in the area identified as the character area attribute by the non-character area determination step And a subdivision step of subdividing the area identified as the character area attribute when it is determined to exist.

Therefore, it is determined whether or not an area other than the character exists in the area identified as the character area attribute, and it is determined that the area other than the character exists in the area identified as the character area attribute. The area identified as the character area attribute is subdivided. As a result, even when an area other than a character such as a seal stamp or a drawing enters the area determined to be a character as a result of the area division, the area identified as the character area attribute is divided again to remove the area other than the character. Since it is possible to eliminate the area, it is possible to improve the character extraction accuracy.

According to a seventh aspect of the present invention, in the area identification method for a document image according to the sixth aspect, the information regarding the rectangle in the non-character region determination step is assumed to be black pixels inside the rectangle. It is the maximum value and the minimum value of the coordinates of the black pixel for each pixel line.

Therefore, it becomes easy to determine whether or not an area other than a character exists in the area identified as the character area attribute.

According to an eighth aspect of the invention, in the method of identifying the area of a document image according to the seventh aspect, the subdivision step is 1
A character line and a line space are determined according to the maximum value and the minimum value of the black pixel coordinates for each pixel line, and the region identified as the character region attribute is redivided in the line space portion.

Therefore, the subdivision becomes easy.

According to a ninth aspect of the present invention, in the document image area identification method according to the eighth aspect, it is determined whether or not an area other than a character exists by using information about the rectangle of the area determined to be the line spacing. To determine.

Therefore, since there is no character in the area determined to be the space between lines, it becomes easy to determine whether or not there is an area other than the character.

The invention according to claim 10 is an index (confidence) indicating the certainty of the result of character recognition after subdivision in the area identification method of the document image according to any one of claims 6 to 9. And the part with low confidence is
Area other than characters.

Therefore, it is possible to further improve the division accuracy.

The program according to the invention of claim 11 is a program for causing a computer to perform classification and classification of a character area and a non-character area that are mixed in document image data. A circumscribing rectangle extraction function for extracting a rectangle circumscribing the connected component of black pixels contained in the area identified as the above from the document image data, and whether an area other than a character exists in the area identified as the character area attribute A non-character area determination function that determines whether or not there is a non-character area determination function based on the information related to the rectangle extracted by the circumscribing rectangle extraction function, and an area other than a character exists in the area identified as the character area attribute by the non-character area determination function. If it is determined that the character area attribute exists, a re-division function of re-dividing the area identified as the character area attribute is executed.

Therefore, it is determined whether or not an area other than the character exists in the area identified as the character area attribute, and it is determined that the area other than the character exists in the area identified as the character area attribute. The area identified as the character area attribute is subdivided. As a result, even when an area other than a character such as a seal stamp or a drawing enters the area determined to be a character as a result of the area division, the area identified as the character area attribute is divided again to remove the area other than the character. Since it is possible to eliminate the area, it is possible to improve the character extraction accuracy.

According to a twelfth aspect of the present invention, in the program according to the eleventh aspect, the information regarding the rectangle in the non-character area determining function assumes that the inside of the rectangle is all black pixels, and black for each pixel line. It is the maximum value and the minimum value of the pixel coordinates.

Therefore, it becomes easy to determine whether or not an area other than a character exists in the area identified as the character area attribute.

According to a thirteenth aspect of the present invention, in the program according to the twelfth aspect, the subdivision function determines a character line and a line space according to a maximum value and a minimum value of coordinates of black pixels for each pixel line. , The area identified as the character area attribute is re-divided in the space between the lines.

Therefore, the subdivision becomes easy.

According to a fourteenth aspect of the present invention, in the program according to the thirteenth aspect, it is determined whether or not there is an area other than characters by using the information related to the rectangle of the area determined to be the line spacing.

Therefore, since there is no character in the area determined to be the space between lines, it becomes easy to determine whether or not there is an area other than the character.

According to a fifteenth aspect of the present invention, in the program according to any one of the eleventh to fourteenth aspects, after subdivision, an index (confidence) indicating the certainty of the result of character recognition is calculated, The part with low confidence is an area other than characters.

Therefore, it is possible to further improve the division accuracy.

A computer-readable storage medium according to a sixteenth aspect of the present invention stores the program according to any one of the eleventh to fifteenth aspects.

Therefore, by installing this storage medium in a computer, the storage medium according to claims 11 to 15 can be obtained.
It is possible to obtain the same operation as that of the program described in any one of 1.

[0040]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described with reference to FIGS.

FIG. 1 is a block diagram schematically showing the hardware configuration of the document recognition device 1. As shown in FIG. 1, a document recognition device 1 includes a CPU (Central Processing Unit) 2 that centrally controls each unit of the document recognition device 1.
The CPU 2 has a read-only memory (ROM) that is a read-only memory that stores a BIOS and the like.
y) 3 and a RAM (R
andom Access Memory) 4 is connected by a bus 5. Further, the bus 5 has an HDD (HardDi
sk Drive) 6, a CD-ROM drive 8 for reading a CD (Compact Disc) -ROM 7, a communication control device 10 for controlling communication between the document recognition device 1 and the network 9, a keyboard, a mouse, etc. functioning as an input unit. Input device 1
1, CRT (Cathode Ray Tube), LCD (Liquid C)
An output device 12 such as a crystal display) and an image input device 13 such as a scanner that functions as an image input unit are connected via an I / O (not shown).

The RAM 4 functions as a work area of the CPU 2 because it has a property of rewritably storing various data.

Further, the HDD 6 stores program files for storing various programs.

The CD-ROM 7 shown in FIG. 1 embodies the storage medium of the present invention and stores a predetermined program. The CPU 2 reads the program stored in the CD-ROM 7 with the CD-ROM drive 8 and installs it in the HDD 6. As a result, the document recognition device 1 is in a state in which it is possible to perform various types of processing described below.

A CD-ROM 7 is used as a storage medium.
In addition, various types of media such as various optical disks such as DVDs, various magneto-optical disks, various magnetic disks such as floppy (registered trademark) disks, and semiconductor memories can be used. Alternatively, the program may be downloaded from the network 9 such as the Internet via the communication control device 10 and installed in the HDD 6. In this case, the storage device that stores the program in the server on the transmission side is also the storage medium of the present invention. In addition, the program is a predetermined OS (Operating Syst
em) may be used, or in this case, the OS may be allowed to execute a part of various processes described below, or a predetermined application software such as word processing software or an OS may be used. It may be included as a part of a group of program files constituting the program.

Next, various functions realized by the CPU 2 of the document recognition apparatus 1 being controlled based on a program will be described. FIG. 2 is a functional block diagram of the document recognition device 1.

The area identification section 14 is, for example, the image input device 1.
The document image input from the memory 3 and stored in the memory (RAM 4 or the like) is identified, and the document image is classified into a character region, a table region, a drawing region, a photograph region, and the like. Note that the area attribute of a document can be obtained by a method such as using the density of black runs to make a determination, but since this method has been publicly known, its description is omitted.

The drawing area extracting unit 15 determines whether or not a solid line (seal stamp, drawing, etc.) is mixed in the area classified as a character area by the area identifying unit 14. A method for determining whether or not a solid line (seal stamp, drawing, etc.) is mixed is conventionally known, and therefore its description is omitted.

When it is judged that the solid line (seal stamp, drawing, etc.) is mixed in the area classified as the character area by the figure area extracting section 15, the area dividing section 16 re-divides the target character area, It is passed to the character recognition unit 17.

The character recognition unit 17 cuts out one character by the line cutout process and the character cutout process, and selects a character candidate by the matching process of the character recognition process for the cut out character.

If it is determined that the solid line (seal stamp, drawing, etc.) is not included in the area classified as the character area by the drawing area extracting unit 15, the area is classified as the character area by the drawing area extracting unit 15. The area that has been opened is passed to the character recognition unit 17 as it is.

Here, the flow of processing in the drawing area extracting unit 15 and the area dividing unit 16 that exerts the characteristic function of this embodiment will be described in detail with reference to FIG. First, in step S1, for the area classified as a character area by the area identification unit 14, area coordinate data (X, Y of the start point and the end point extracted by the 1/4 compressed image obtained by compressing the input image into 1/4) is used. (Coordinates) is used to determine whether or not the corresponding area is a vertically long area. If the corresponding area is a vertically long area, the corresponding area is excluded (the subsequent processing is not performed).

In addition, in step S2, it is determined whether or not the row direction of the relevant area is "vertical". If the row direction of the relevant area is "vertical", the relevant area is excluded ( Do not perform subsequent processing).

Next, in step S3, candidate areas are detected. More specifically, first, rectangle extraction processing is performed on the original image of the character area detected by the above processing to obtain rectangle coordinate data. Here, the function of the circumscribing rectangle extracting means is executed. Here, the reason why the 1/4 compressed image is not used is that when the compressed image is used, the rectangles come into contact with each other and become a large rectangle. This is because the forced division position is estimated from the rectangular coordinate information in the subsequent process, so that if the rectangles come into contact with each other to form a large rectangle, the estimation accuracy will not increase. Then, it is assumed that all the rectangles in the character area obtained as a result of the rectangle extraction processing are black pixels, and the smallest X coordinate in the character area is calculated for each line (for each Y coordinate).
The thick solid line: minXs (y)) and the largest X coordinate (thick broken line: maxXe (y) shown in FIG. 4) are obtained.

The values of y of minXs (y) and maxXe (y) take from the start point (area.Ys) to the end point (area.Ye) of the coordinates of the relevant area, and the above range is 0 to. Normalize to 100 data of 99. minXs (y), maxXs (y) → (normalization) → minXs
(Y), maxXe (Y) where Y = (y-area.Ys) / (area.Ye-area.Ys) × 10
From the above, the maximum value and the minimum value of the X coordinate of the rectangle in the area are obtained for each pixel row.

Then, it is determined whether or not a plurality of lines are in contact with each other from the maximum and minimum values of the X coordinate of the rectangle in this area. In the example in which the left sides of a plurality of lines are in contact (see FIG. 4A), attention is paid to the minimum value of the X coordinate, and the line where the minimum value continuously approaches the start point Xs of the area is set. Also,
Since the minimum value of the X coordinate is close to the end point Xe of the area, the space between lines is defined as the space between successive lines. Then, it is determined that a plurality of lines are in contact with each other in the area where the combination of “line-line spacing-line” is detected (that is, a seal or a figure is mixed), and as shown in FIG. Forcibly divide the area at the center. In addition, in the example in which the right sides of a plurality of lines are in contact (see FIG. 4B), the same process is performed paying attention to the maximum value of the X coordinate. As a result, the character area is subdivided, and the candidate area is detected (step S3).
Ends. Here, the function of the non-character area determining means and the function of the re-dividing means are executed.

Finally, the process proceeds to step S4 to execute the final judgment process. The final judgment is to determine whether or not a solid line (seal stamp, figure, etc.) is mixed in the re-divided character area, and the solid line (seal stamp or figure) is added to the rectangle existing in the area determined to be the line space. Etc.) is present. If a solid line (seal stamp, drawing, etc.) exists in the rectangle, the rectangle is used as the drawing area.

In the above description, an example in which a line is divided horizontally based on the unevenness information of the coordinates has been described. However, by using the unevenness information, a figure area portion which is considered to be a figure or a photograph is vertically formed by a convex portion as shown in FIG. You may make it divide | segment into a direction.

After these divisions, an index (certainty) indicating the certainty of the result of character recognition is calculated, and the portion with low certainty (low certainty) is set as the drawing area. It is also possible to further improve the division accuracy.

Although the present embodiment has been described with respect to a horizontal writing document having a horizontal orientation in the row direction, the present invention is not limited to this and can be applied to a vertical writing document having a vertical orientation in the row direction.

Here, it is determined whether or not an area other than the character exists in the area identified as the character area attribute, and it is determined that the area other than the character exists in the area identified as the character area attribute. In this case, the area identified as the character area attribute is redivided. As a result, even when an area other than a character such as a seal stamp or a drawing enters the area determined to be a character as a result of the area division, the area identified as the character area attribute is divided again to remove the area other than the character. Since it is possible to eliminate the area, it is possible to improve the character extraction accuracy.

[0062]

According to the document recognition apparatus of the first aspect of the invention, in the document recognition apparatus for distinguishing and classifying the mixed character area and non-character area in the document image data, the character area attribute is identified. A circumscribing rectangle extracting means for extracting a rectangle circumscribing a connected component of black pixels contained in the region from the document image data, and determining whether or not a region other than a character exists in the region identified as the character region attribute. A non-character area determining unit that determines based on the information about the rectangle extracted by the circumscribing rectangle extracting unit, and a non-character area is determined to exist in the area identified as the character area attribute by the non-character area determining unit. In the case where the area identified by the character area attribute is re-divided, it is determined whether or not an area other than the character exists in the area identified by the character area attribute. If it is determined that there is an area other than characters in the area identified as the area attribute, the area identified as the character area attribute is re-divided so that the area other than the character such as a seal or a figure is divided as a result of the area division. Even when it enters the area determined to be a character, the area other than the character can be excluded by dividing the area identified as the character area attribute again, so that the character extraction accuracy can be improved. it can.

According to the second aspect of the present invention, in the document recognition apparatus according to the first aspect, the information regarding the rectangle in the non-character area determining means is assumed to be a black pixel inside the rectangle, With the maximum value and the minimum value of the coordinates of the black pixel for each line, it is possible to easily determine whether or not there is an area other than the character in the area identified as the character area attribute.

According to the third aspect of the present invention, in the document recognition apparatus according to the second aspect, the subdivision means determines a character line and a character line according to the maximum value and the minimum value of the coordinates of the black pixel for each pixel line. Re-division can be easily performed by determining the space between lines and re-dividing the region identified as the character region attribute in this space between lines.

According to the fourth aspect of the invention, in the document recognition apparatus according to the third aspect, it is determined whether or not there is an area other than a character by using the information related to the rectangle of the area determined to be the line spacing. By determining, since there is no character in the area determined to be the line spacing, it can be easily determined whether or not there is an area other than the character.

According to the invention described in claim 5, in the document recognition apparatus according to any one of claims 1 to 4, after the subdivision, an index (confidence) indicating the certainty of the result of character recognition is obtained. It is possible to further improve the division accuracy by calculating and calculating the portion with low confidence as a region other than the character.

According to the document image area identification method of the sixth aspect of the present invention, there is provided a document image area identification method for distinguishing and classifying mixed character areas and non-character areas in document image data. A circumscribing rectangle extraction step of extracting a rectangle circumscribing a connected component of black pixels included in the area identified as the area attribute from the document image data, and an area other than a character exists in the area identified as the character area attribute. A non-character area determination step of determining whether or not to perform based on the information on the rectangle extracted by the circumscribing rectangle extraction step, and a character other than a character in the area identified as the character area attribute by the non-character area determination step. If it is determined that the area exists, a subdivision step of subdividing the area identified as the character area attribute,
If it is determined that an area other than the character exists in the area identified as the character area attribute and the area other than the character exists in the area identified as the character area attribute, the character By redistributing the area identified as the area attribute, even if a non-character area such as a seal or a figure enters the area determined to be the character as a result of the area split, the area is identified as the character area attribute. Since the area other than the character can be excluded by dividing the area that has been
The character extraction accuracy can be improved.

According to the seventh aspect of the present invention, in the document image area identification method according to the sixth aspect, the information regarding the rectangle in the non-character area determining step is assumed to be all black pixels inside the rectangle. With the maximum and minimum values of the black pixel coordinates for each pixel line, it is possible to easily determine whether or not an area other than a character exists in the area identified as the character area attribute. .

According to the eighth aspect of the present invention, in the document image area identification method according to the seventh aspect, the subdivision step is performed according to the maximum value and the minimum value of the coordinates of the black pixel for each pixel line. Re-division can be easily performed by determining a character line and a line space, and re-dividing the region identified as the character region attribute in the line space portion.

According to the invention of claim 9, in the area identification method of the document image according to claim 8, there is an area other than a character by using the information related to the rectangle of the area determined to be the line spacing. By determining whether or not there is no character in the area determined to be the line spacing, it is possible to easily determine whether or not there is an area other than the character.

According to the invention of claim 10, claim 6
In the document image area identification method according to any one of 1 to 9, an index (confidence factor) indicating the certainty of the result of character recognition is calculated after the subdivision, and the portion with low confidence factor is not a character. By setting the area to, the division accuracy can be further improved.

According to the program of the eleventh aspect of the present invention, there is provided a program for causing a computer to perform classification / discrimination between a character area and a non-character area mixed in the document image data. A circumscribing rectangle extraction function for extracting a rectangle circumscribing a connected component of black pixels included in the area identified as the area attribute from the document image data, and an area other than a character exists in the area identified as the character area attribute. A non-character area determination function for determining whether or not to perform based on the information related to the rectangle extracted by the circumscribing rectangle extraction function, and a character other than a character in the area identified as the character area attribute by the non-character area determination function. When it is determined that the area exists, the subdivision function of subdividing the area identified as the character area attribute is executed, and the area identified as the character area attribute is executed. If there is an area other than a character in the area, and if it is determined that an area other than a character exists in the area identified as the character area attribute, the area identified as the character area attribute should be subdivided. Even if an area other than a character, such as a seal stamp or a drawing, enters an area determined to be a character as a result of area division, the area other than the character is divided by dividing the area identified with the character area attribute again. Can be eliminated, so that the character extraction accuracy can be improved.

According to the invention of claim 12, claim 1
In the program according to 1, the information regarding the rectangle in the non-character area determination function is a maximum value and a minimum value of the coordinates of the black pixel for each pixel line, assuming that the inside of the rectangle is all black pixels. It is possible to easily determine whether or not an area other than a character exists in the area identified as the character area attribute.

According to the invention of claim 13, claim 1
In the program described in 2, the subdivision function determines a character line and a line space according to a maximum value and a minimum value of a black pixel coordinate for each pixel line, and an area identified as a character area attribute in the line space portion. Can be easily re-divided.

According to the invention of claim 14, claim 1
In the program described in 3, the information regarding the rectangle of the area determined to be the line spacing is used to determine whether or not there is an area other than the character, so that a character is present in the area determined to be the line spacing. Therefore, it is possible to easily determine whether or not there is an area other than the character.

According to the invention of claim 15, claim 1
In the program according to any one of 1 to 14,
After the redivision, an index (confidence factor) indicating the certainty of the result of character recognition is calculated, and the part with a low confidence factor is an area other than the character, so that the division accuracy can be further improved. .

According to the computer-readable storage medium of the sixteenth aspect of the present invention, the eleventh to fifteenth aspects are provided.
By storing the program described in any one of,
By installing this storage medium in a computer, it is possible to obtain the same actions and effects as the program according to any one of claims 11 to 15.

[Brief description of drawings]

FIG. 1 is a block diagram schematically showing a hardware configuration of a document recognition device according to an embodiment of the present invention.

FIG. 2 is a functional block diagram of a document recognition device.

FIG. 3 is a flowchart showing a flow of processing in a diagram area extracting unit and an area dividing unit.

FIG. 4 is a diagram showing an X of an in-region rectangle based on an in-region rectangle extraction result.
It is explanatory drawing which shows the example which calculated | required the maximum value and minimum value of a coordinate.

FIG. 5 is an explanatory diagram showing an example of a forced division position.

FIG. 6 is an explanatory diagram showing another example of the forced division position.

[Explanation of symbols]

1 Document recognition device 7 storage media

Claims (16)

[Claims] 1. In a document recognition device for distinguishing and classifying mixed character areas and areas other than characters in document image data, a rectangle circumscribing a connected component of black pixels included in the area identified as a character area attribute. Circumscribing rectangle extracting means for extracting from the document image data, and whether or not a region other than a character exists in a region identified as a character region attribute, based on the information relating to the rectangle extracted by the circumscribing rectangle extracting device. Based on the non-character area determination means based on the non-character area determination means based on the non-character area determination means based on the non-character area determination means based on the non-character area determination means based on the non-character area determination means based on the non-character area determination means based on the non-character area determination means. A document recognition device, comprising: a re-division unit for dividing. 2. The information relating to the rectangle in the non-character region determining means is a maximum value and a minimum value of coordinates of black pixels for each pixel line, assuming that the inside of the rectangle is all black pixels. The document recognition device according to claim 1. 3. The subdivision means determines a character line and a line space according to the maximum value and the minimum value of the coordinates of the black pixel for each pixel line, and determines the region identified as the character region attribute in the line space portion. The document recognition device according to claim 2, wherein the document recognition device is divided again. 4. The document recognition apparatus according to claim 3, wherein it is determined whether or not there is an area other than a character by using information about the rectangle of the area determined to be the line spacing. 5. After the subdivision, an index (certainty factor) indicating the certainty of the result of character recognition is calculated, and the portion with low certainty factor is an area other than the character. The document recognition device according to any one of 1 to 4. 6. A document image area identification method for identifying and classifying a character area and a non-character area mixed in document image data, wherein black pixels included in an area identified as a character area attribute are concatenated. A circumscribing rectangle extracting step of extracting a rectangle circumscribing a component from the document image data; and a step of extracting whether a region other than a character exists in a region identified as a character region attribute by the circumscribing rectangle extracting process. A non-character area determination step that is determined based on the information related to the rectangle, and if it is determined that an area other than a character exists in the area that is identified as a character area attribute by this non-character area determination step, it is identified as the character area attribute. And a subdivision step of subdividing the created area. 7. The information regarding the rectangle in the non-character region determining step is a maximum value and a minimum value of coordinates of black pixels for each pixel line, assuming that the inside of the rectangle is all black pixels. The area identification method of the document image according to claim 6. 8. The subdivision step determines a character line and a line space according to the maximum value and the minimum value of the coordinates of black pixels for each pixel line, and determines an area identified as a character area attribute in the line space portion. The area identification method of the document image according to claim 7, wherein the area division is performed again. 9. The area identification of a document image according to claim 8, wherein it is determined whether or not an area other than a character exists by using information about the rectangle of the area determined to be the line spacing. Method. 10. An index (certainty factor) indicating the certainty of the result of character recognition after re-division, and the portion with low certainty factor is an area other than the character. 6. The document image area identification method according to any one of 6 to 9. 11. A program for causing a computer to identify and classify a mixed character region and non-character region in document image data, the program being included in a region identified as a character region attribute by the computer. A circumscribed rectangle extraction function for extracting a rectangle circumscribing a connected component of black pixels from the document image data, and a circumscribed rectangle extraction function for determining whether or not a region other than a character exists in a region identified as a character region attribute. A non-character area determination function that is determined based on the information related to the rectangle extracted by, and if it is determined that an area other than a character exists in the area that is identified as a character area attribute by the non-character area determination function, A program for executing a subdivision function for subdividing an area identified as a character area attribute. 12. The information relating to the rectangle in the non-character region determining function is a maximum value and a minimum value of coordinates of black pixels for each pixel line, assuming that the inside of the rectangle is all black pixels. The program according to claim 11, wherein: 13. The subdivision function determines a character line and a line space according to a maximum value and a minimum value of a black pixel coordinate for each pixel line, and determines an area identified as a character area attribute in the line space portion. 13. The program according to claim 12, wherein the program is divided again. 14. The program according to claim 13, wherein it is determined whether or not an area other than a character exists by using information about the rectangle of the area determined to be the line spacing. 15. An index (certainty factor) indicating the certainty of the result of character recognition after the subdivision is calculated, and the portion with the low certainty factor is an area other than the character. 15. The program according to any one of 11 to 14. 16. A computer-readable storage medium storing the program according to any one of claims 11 to 15.