JPH09146947A - Method and device for sequencing character area - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make the passage of the document of a main body based upon, for example, a character recognition result normal by setting the processing order of a footer part to the least. SOLUTION: A document image is inputted and individual character areas are extracted. A character area which is a footer is extracted from the individual extracted character areas (S11). Then, the character areas other than the footer are also given processing order (S12), and the character area judged to be the footer is given order lastly.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character area ordering method and apparatus for extracting a character area in a document image and setting a processing order for the extracted character area.

[0002]

2. Description of the Related Art Generally, a document image is optically read,
In the case of a character recognition device, it is necessary to perform recognition processing along the flow of sentences.

There is no problem when the input document has only one column, but when the text is described in two columns, for example, in the left page and the right field in the same page, character recognition is performed in column units. There is a need to.

[0004]

[Problems to be Solved by the Invention] However, recently,
As shown in FIG. 3, a footer (page number in the figure) is often provided in the document. In this case, the footer cannot be distinguished from the text (the footer is interpreted as the text)
Since the characters are recognized, the flow of the text does not become normal.

[0005]

SUMMARY OF THE INVENTION The present invention intends to provide a character area ordering method and apparatus which solves the above problems by making the processing order of the footer part the last of the other character areas of the original image. It is a thing.

In order to solve such a problem, for example, the character area ordering method of the present invention includes the following steps.
That is, a method for ordering character areas in which a character area in a document image is extracted and a processing order is set for the extracted character area, and a footer character area in each extracted character area is identified. An identification step and a setting step of setting the processing order for the footer area identified in the identification step after the other character areas are provided.

According to a preferred embodiment of the present invention,
The identifying step includes a step of determining whether the target character area is located below a predetermined position in the original image. When the target character area is located below the predetermined position in the determination step, the target character area is set as a footer character area. It is desirable to identify. This makes it possible to determine the character region located in the lower part of the document image as a footer.

Further, the identifying step further includes a step of determining whether or not the target character area is composed of a character string having only one line, and when the character string is determined to be a character string having only one line. It is desirable to identify character areas as footer character areas. This makes it possible to make a determination with higher accuracy than the footer.

Further, it is preferable that the identifying step further includes a step of determining whether the target character area is horizontally written or horizontally long, and when it is determined that the target character area is horizontally written or horizontally long, the target character area is identified as a footer character area. This makes it possible to make a determination with higher accuracy than the footer.

The identifying step further includes a step of determining whether or not another character area exists below the target character area, and when it is determined that there is no other character area below the target character area, the target character area is determined. It is desirable to identify the area as a footer character area. This makes it possible to make a determination with higher accuracy than the footer.

The identifying step further includes a step of determining whether or not the target character area is included in another line drawing,
When included in another line drawing, it is desirable to identify the focused character area as a non-footer character area. by this,
It is possible to make a determination with higher accuracy than the footer.

Furthermore, it is desirable to further include a character recognition step of recognizing a character image in each character area according to the order set in the setting step. This makes it possible to make the flow of sentences resulting from character recognition a natural state.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An example of an embodiment according to the present invention will be described in detail below with reference to the accompanying drawings.

[First Embodiment] Generally, when reading a document image and recognizing a character in the read document image, first, an area in which the character exists is determined.

There are various methods for determining the area where a character exists. For example, a document image is pre-scanned and read at a low resolution so that characters are crushed and are connected to adjacent character patterns. Then, a rectangle circumscribing the area is obtained, the original image is read at the original high resolution, and the bitmap data of the character area is extracted along with the determined rectangle. After this, there is a method of creating a histogram of dots in the horizontal and vertical directions, detecting the existence position and size of each character, and passing the bitmap pattern cut out accordingly to the character recognition processing.

It should be noted that, without pre-scanning, a method of forming a low resolution image by expanding the "1" dots in the read image data in the vertical and horizontal (and diagonal) directions, and in the read image There is also a method of newly creating low-resolution image data with n × m as one dot when at least one “1” is present in the n × m pixel block.

This embodiment may use any method,
The present invention is not limited by the selection of this method.

FIG. 9 shows a block diagram of the character recognition apparatus in the embodiment. In the figure, 1 is a CPU that controls the entire apparatus, 2 is a ROM that stores a boot program and the like, 3 is a RAM that is used by the CPU 1 to develop a program executed by the CPU 1, read images, etc. Is. An input device 4 is composed of a keyboard, a pointing device, and the like.

Reference numeral 5 is an image scanner for reading the original image, and 6 is an interface for taking in the image data read by the image scanner into the inside of the apparatus. 7 is an external storage device such as a hard disk device,
It stores an operating system (OS), programs corresponding to each flowchart described later, a dictionary used for character recognition, and a corresponding program.
Reference numeral 8 is a VRAM that expands an image displayed on the display screen, and 9 is a display device that displays the image expanded in the VRAM 8.

In the above structure, when the power of the device is turned on, the CPU 1 reads the OS program from the external storage device 7 in accordance with the boot program in the ROM 2 and puts the device into an operating state. Then, according to the operator's instruction (or it may be automatic),
It will be loaded into AM3.

In the present embodiment, the character recognition program is activated, and an image is read from the image scanner 5 via the interface 6 in response to an instruction to read the original image (21 in FIG. 2). Character area (Fig. 2
22), the process is performed according to the procedure shown in FIG.

First, in step S11, the character area of the footer is extracted from the character area, and the character areas other than the footer are ordered in step S12. Then, in step S13, the footer character area is assigned to the next sequential position of the last character area.

The extraction processing of the footer character area in step S11 will be described with reference to the flowchart of FIG.

First, in step S51, the position of the focused character area is 1 from the bottom of the document to the height of the document image.
Judge whether or not it is within / n. Note that n is a number that divides the height of the original image into equal parts, and differs depending on, for example, the size of the original image. Since the footer is provided at the lower end of the document, it is determined here whether or not there is a focused character area at that position.

Here, if the attention area does not satisfy this condition, it is determined that the attention character area is not the footer, and, for example, a flag provided in each character area indicates that the area is other than the footer. Set information (step S5)
5).

If it is determined that the character area of interest is near the lower end of the original image, the process proceeds to step S52, and it is determined whether the character area is composed of a character string having only one line. The number of lines can be determined by taking a histogram of dots in the arrangement direction of the character patterns. In any case, when there are two or more lines, the focused character area is determined as a character area other than the footer (step S55).

Further, it is located near the lower end of the original and
If it is determined that it consists of a single line character string,
In step S53, it is determined whether the writing is horizontal or horizontal.

If the result of this determination is "YES", then the flow proceeds to step S54, and the target character area is determined as a footer (the footer is written in the flag).

In step S56, it is determined whether or not all the character areas have been processed. If not, step S5.
The process returns to 1 and the above process is repeated.

As a result of the above, the area is the footer portion,
Since the character string is positioned last by the processing of FIG. 1, the recognition result of the footer portion is not inserted in the middle of the text when the character is recognized.

It should be noted that, before and after the explanation, when the original image is composed of two columns or more, the method itself for discriminating each column is known, and for example, a histogram of dots in the vertical direction of the original image is used. The delimiter between columns should be detected by creating it.

<Second Embodiment> In the first embodiment, the description has been made on the assumption that there is one character area as the footer portion. However, as shown in FIG. A processing example when there are two or more footer character strings will be described as a second embodiment.

In the second embodiment, the character areas are ordered in accordance with the ordering processing procedure shown in FIG.

The steps up to steps S61 and S62 are the same as those in the first embodiment. That is, the area that becomes the footer is extracted from the character area, and the character areas other than the footer are ordered.

In step S63, when there are a plurality of character areas determined to be footers, the one on the left side is given priority based on the coordinate position, and is continued after the character areas other than the footer.

As a result, as shown in FIG. 7, the sequence number of the footer character area 72 is "11" and that of the footer character area 73 is "12".

As described above, according to the embodiment of the subject 2,
It is possible to perform natural character recognition when there are a plurality of footer portions.

<Third Embodiment> In the first and second embodiments, first, the footer portion is taken out, then the character areas other than the footer are ordered, and the footer character area is added to the end. However, this does not limit the present invention. For example, the processing may be performed as shown in FIG.

First, in step S81, all character areas including the footer are ordered. After this, in step S82,
The footer is taken out from each character area, and the order of the footer taken out in step S83 is added to the last order.

The same effect can be obtained by the above processing.

<Fourth Embodiment> In the above embodiment, the footer is judged based on the height of the original, but the judgment as to whether the footer is a footer is not limited to this.

In particular, in the above embodiment, the height of the document is 1
The position of / n was determined as the boundary position, and the portion below it was determined as the possibility of footer. However, for example,
As shown in FIG. 15, if the boundary position is at the position indicated by reference numeral 102 and the character area 103 of the text exists below the boundary position, that position is also determined as a footer.

Therefore, in the fourth embodiment, even in such a case, the character area 104 in FIG. 15 is determined as a footer. Note that the device configuration is the same as that in FIG. 9, and the character region extraction processing is also the same as in the first embodiment, and a description thereof will be omitted.

FIG. 10 is a flowchart of the footer determination processing in the fourth embodiment, which is an alternative to FIG.

First, in step S101, it is determined whether the character area is only one line. If it is determined that there is only one line, the process proceeds to step S102, and it is determined whether the character area is horizontally written or horizontally long. If it is determined to be either, the process proceeds to step S103, and an area below the coordinate position of the focused character area (for example, when the focused character area is the character area 103 in FIG. It is determined whether there is another character area in the area 105).

As a result, if it is determined that the target character area is located at the bottom, it is determined in step S104 that the target character area is the footer. If any of the conditions in steps S101 to S103 is not satisfied (for example, in the case of the character area 103), the character area is determined as a character area other than the footer.

Thus, until it is determined in step S106 that the processing for all the character areas is completed,
The above process is repeated.

As a result, even in the situation as shown in FIG. 15, only the character area 104 is determined to be the footer, and the character area 103 can be determined as the character area (text) other than the footer. Become.

<Fifth Embodiment> In the fourth embodiment, whether or not there is another character area on the lower side (the area having the same width as the character area and the lower side) of the character area of interest. It was determined whether the focused character area was the footer. For example,
When a document is read, the user sets a document with a footer with a keyboard or the like, and if there is such a setting, there is no problem if there is no footer although there is no problem if the above-mentioned fourth processing is performed. When the character recognition is performed as, the character area at the bottom of the body is recognized as a footer, and as a result, the recognition result is positioned at the end. It cannot be denied that the recognition result may come at the end of.

Therefore, in the embodiment of the present subject 5, as in step S111 of FIG.
As with the first embodiment, the problem is solved by including the position information.

Since the other parts are the same as those in the fourth embodiment, the description thereof will be omitted.

As a result, according to the fifth embodiment, it is possible to correctly identify the footer as compared with the fourth embodiment.

<Sixth Embodiment> A sixth embodiment will be described. In the sixth embodiment, as shown in, for example, FIGS. 16 and 17, frames (for example, tables) 112 and 122 are included in the text.
Therefore, even if it happens that the character areas 113 and 123 are located on the lower side of the table, the character areas 113 and 123 are not determined as footers.
In addition to the rectangular frame, for example, as shown in FIG.
Even if there is a character area 133 in the line drawing 132, the character area 133 is not determined as a footer.

FIG. 12 shows the processing procedure of the sixth embodiment. The difference from the fifth embodiment is that step S
This is the point where 124 is added.

That is, the condition is that whether or not the target character area belongs (is included) other than the character area such as a frame (frame), table, or figure.

In order to determine whether a frame, a table, or a figure (these are collectively referred to as a line drawing), for example, a continuous circumscribing rectangle of dots forming the line drawing is detected and the circumscribing rectangle is determined. It suffices to use a method such as whether the ratio of the number of black dots to the size of is less than or equal to a predetermined value.

As a result, according to the sixth embodiment, even if the characters included in the line drawing are in the lower part of the original image, the character string is not judged to be the footer. It is possible to make the order of the results normal.

<Seventh Embodiment> Further, by adding step S134 of FIG. 13 (the same as step S124 of FIG. 12) to FIG. 10, the same operational effect as that of the sixth embodiment can be obtained. It will be possible.

<Eighth Embodiment> Further, as shown in FIG. 14, the processes in the sixth and seventh embodiments may be integrated. In this case, it is possible to more accurately distinguish the footer from the character area that does not.

As described above, according to the first to eighth embodiments, the footer portion in the original image can be identified.
Further, since the character recognition processing of the identified footer portion is performed after the text, it is possible to at least make the flow of text in the text that is the recognition result natural.

Even when the present invention is applied to a system composed of a plurality of devices (eg, host computer, interface device, reader, printer, etc.), a device composed of one device (eg, copier, facsimile, etc.) May be applied to.

Further, an object of the present invention is to supply a storage medium storing a program code of software for realizing the functions of the above-described embodiments to a system or apparatus, and to supply the computer (or CPU or M of the system or apparatus).
It goes without saying that it is also achieved by the PU) reading and executing the program code stored in the storage medium.

In this case, the program code itself read from the storage medium realizes the novel function of the present invention, and the storage medium storing the program code constitutes the present invention.

A storage medium for supplying the program code is, for example, a floppy disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD.
-R, magnetic tape, non-volatile memory card, ROM, etc. can be used.

Further, by executing the program code read by the computer, not only the functions of the above-described embodiment are realized, but also the OS or the like running on the computer is actually executed based on the instruction of the program code. It goes without saying that a case where a part or all of the processing of (1) is performed and the function of the embodiment is realized by the processing is also included.

Further, after the program code read from the storage medium is written in the memory provided in the extended function board inserted into the computer or the extended function unit connected to the computer, based on the instruction of the program code, It goes without saying that a case where a CPU or the like included in the function expansion board or the function expansion unit performs a part or all of the actual processing and the processing realizes the functions of the above-described embodiments is also included.

When the present invention is applied to the above-mentioned storage medium, the storage medium stores the program code corresponding to one of the above-mentioned flowcharts. Briefly, the memory shown in FIG. Each module shown in the map example will be stored in the storage medium. That is, at least the identification module for identifying the footer character area in each character area and the setting module for setting the processing order for the identified footer area after the other character areas are stored in the storage medium. Good.

[0068]

As described above, according to the present invention, with respect to the footer portion, if the processing order is set to the end of other character areas of the original image, for example, if character recognition processing is performed, recognition is performed. It is possible to make the flow of the resulting sentence natural.

[0069]

[Brief description of the drawings]

FIG. 1 is a flowchart of a character area ordering process according to a first embodiment.

FIG. 2 is a diagram showing a character area extracted from a document image.

FIG. 3 is a diagram showing a conventional ordering of character areas.

FIG. 4 is a diagram showing an ordering result in the first embodiment.

FIG. 5 is a flowchart showing a footer identification process in the first embodiment.

FIG. 6 is a flowchart showing an ordering process of the second embodiment.

FIG. 7 is a diagram showing an ordering result in the second embodiment.

FIG. 8 is a flowchart showing an ordering process in the third embodiment.

FIG. 9 is a block configuration diagram of an apparatus according to an embodiment.

FIG. 10 is a flowchart showing footer identification processing according to the fourth embodiment.

FIG. 11 is a flowchart showing footer identification processing according to the fifth embodiment.

FIG. 12 is a flowchart showing footer identification processing according to the sixth embodiment.

FIG. 13 is a flowchart showing footer identification processing according to the seventh embodiment.

FIG. 14 is a flowchart showing footer identification processing in the eighth embodiment.

FIG. 15 is a diagram illustrating the contents of operation processing according to the fourth embodiment.

FIG. 16 is a diagram showing an example of a non-footer character area.

FIG. 17 is a diagram showing an example of a non-footer character area.

FIG. 18 is a diagram showing an example of a non-footer character area.

FIG. 19 is a diagram showing a program storage state in a storage medium.

[Explanation of symbols]

 21, 71 Original image 22 Character area 23, 72, 73 Footer character area

Claims (15)

[Claims] 1. A method for ordering character areas in which a character area in an original image is extracted and a processing order is set for the extracted character area, wherein a footer character area in each extracted character area is provided. And a setting step of setting a processing order for the footer area identified in the identification step after the other character areas. 2. The identifying step includes a step of determining whether the target character area is located below a predetermined position in the original image, and the target character area is located below the predetermined position in the determining step. The character area ordering method according to claim 1, wherein is identified as a footer character area. 3. The identifying step further includes a step of determining whether or not the character area of interest is composed of a character string having only one line, and when it is determined that the character area is a character string having only one line. The method for ordering character areas according to claim 2, wherein the character areas are identified as footer character areas. 4. The identifying step further includes a step of determining whether the target character area is horizontally written or horizontally long, and when it is determined to be horizontally written or horizontally long, the target character area is identified as a footer character area. The method for ordering character areas according to claim 3. 5. The identifying step further includes a step of determining whether or not there is another character area below the target character area, and when it is determined that there is no other character area below, the target character area is determined. The method for ordering character areas according to claim 4, wherein the areas are identified as footer character areas. 6. The identifying step further includes a step of determining whether or not the target character area is included in another line drawing. When the target character area is included in another line drawing, the target character area is set as a non-footer character area. The character area ordering method according to claim 5, wherein the character areas are identified. 7. The ordering of character areas according to claim 1, further comprising a character recognition step of recognizing a character image in each character area according to the order set in the setting step. Method. 8. A character area ordering device for extracting a character area from a document image and setting a processing order for the extracted character area, wherein the footer character area is included in each extracted character area. An apparatus for ordering character areas, comprising: identification means for identifying the character area; and setting means for setting a processing order for the footer area identified by the identification means after other character areas. 9. The identifying means includes means for determining whether the character area of interest is located below a predetermined position in the original image, and the character area of interest is located when the character area is located below the predetermined position by the determining means. 9. The character area ordering device according to claim 8, wherein is identified as a footer character area. 10. The identifying means further includes means for determining whether or not the target character area is composed of a character string having only one line, and is discriminated when it is determined that the character string has only one line. 10. The character area ordering device according to claim 9, wherein the character area is identified as a footer character area. 11. The identifying means further includes means for determining whether the target character area is horizontally written or horizontally long, and when it is determined to be horizontally written or horizontally long, the target character area is identified as a footer character area. The character area ordering device according to claim 10. 12. The identifying means further includes means for determining whether or not there is another character area below the target character area, and when it is determined that there is no other character area below, the target character area is identified. The character area ordering device according to claim 11, wherein the area is identified as a footer character area. 13. The identifying means further includes means for determining whether or not the target character area is included in another line drawing. When the target character area is included in another line drawing, the target character area is set as a non-footer character area. 13. The character area ordering device according to claim 12, wherein the character area is identified. 14. The character area ordering device according to claim 8, further comprising character recognition means for recognizing a character image in each character area according to the order set by the setting means. apparatus. 15. A computer-readable memory storing a program code for extracting a character area from a document image and setting a processing order for the extracted character area, the program being a computer-readable memory. Of the footer character area in the character area of the character identification area, and a code of a setting step of setting the processing order after the other character areas for the footer area identified in the identification step. A computer-readable memory characterized by the following.