JP2002312385A - Document automated dividing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a document automated dividing device capable of omitting a partition sheet and automatically determining a break of a document even when various documents having unclear formats are inputted. SOLUTION: This document automated dividing device is provide with an image reading means 101 reading a plurality of documents and forming a document image, a document image storing buffer 102 storing the read document image, a letter identifying means 103 identifying a letter in the document image, a document dividing information extracting means 104 extracting document dividing information for determining a break of the document from an analysis result and a letter identification result of the document image, a document break determination means 105 determining the break of the document on the basis of the document dividing information, a document break possible selection means 106 displaying the document break determination result by the document break determination means to an operation for correction and confirmation, and a document management system registering means 107 dividing the document image into document units for registration in a document management system.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic document dividing apparatus for automatically judging a document break when registering document images of a plurality of documents collectively in a document management system and dividing the document images into document units. About.

[0002]

2. Description of the Related Art In a conventional document management system, an operator visually checks the contents of a document image group read by a scanner, and designates a document break position using a tool. It was divided into units and registered. For this reason, the operator needs effort to determine the break position of the document, which hinders the registration of the document.

As one of the countermeasures, Japanese Patent No. 2929661 proposes a method of semi-automatically dividing a document by sandwiching a separating paper which can be easily distinguished by a machine when the document is read by a scanner in advance. Have been. Hereinafter, this document division apparatus will be described as a conventional example with reference to FIGS. 14 and 15. FIG.

FIG. 14 is a flowchart of a registration process in a conventional method, and FIG. 15 shows an example of a paper document to be registered. In FIG. 15, reference numeral 1501 denotes a partition sheet; 1502, a series of documents including a plurality of pages; 1503, a specific image area;
Reference numeral 1504 denotes a data description area. Hereinafter, the operation of the conventional example will be described with reference to the flowchart shown in FIG.

At the time of registration, first, a bundle of paper documents to be registered is set in the ADF of the scanner (ST1401). At this time, a partition sheet is attached to the head of each document. For example, when registering two documents 1502 shown in FIG. 15, a partition sheet 1501 is attached to the head of both documents.

Next, a scanner reading process (ST140)
In 2), one paper document is read from the ADF of the scanner and converted into a document image. In the partition paper determination process (ST1403), it is determined whether the read document image is a partition paper or a document to be registered. Specifically, the inside of the specific image area 1503 shown in FIG. 15 is analyzed, and if the character string “{partition sheet}” is entered therein, it is determined that the sheet is a partition sheet. judge.

Here, if it is determined that the sheet is a partition sheet,
A partition paper information reading process (ST1404) is performed. In this process, the data description area 1504 on the partition paper is analyzed, and information on the registration destination of the document described therein is read.
Conversely, if it is determined that the document image is a document to be registered, image storage processing (ST1405) is performed. In this process, the document image is registered in the document management system according to the information of the registration destination of the document already read from the partition sheet.

[0008] Thereafter, it is confirmed whether or not the next sheet is present in the ADF of the scanner (ST1406), and the processing from the scanner reading process (ST1402) is repeated until there is no more sheet.

According to the above-described processing, a partition sheet is added to the head of each document and read from the scanner, and the machine automatically determines whether or not the read document image is a partition sheet.
The document can be registered in the document management system without manually determining the break of the document.

[0010]

However, in the above-described conventional method, it is necessary for a human to insert a partition sheet between documents in advance, and the document cannot be automatically divided without an operation of inserting the partition sheet. is there.

In order to solve this problem, characters are recognized from within a fixed area on a document image, and if a specific character string is included in the recognition result, the document image is determined to be a document break. Unnecessary automatic document dividing device
No. 39465. For example, when a character string “employee list” is entered at the upper end of the document image, this document image is determined to be the first page of the document. But,
This method is based on a fixed-form document whose format is clear, and has a problem that it is difficult to apply when a variety of documents whose format is unknown are input.

The present invention has been made to solve the above problem, and has an automatic document dividing function that can automatically determine a document break even when a variety of documents whose format is unknown are not required. The purpose is to provide the device.

[0013]

According to the present invention, there is provided an automatic document dividing apparatus, comprising: an image reading means for reading a plurality of documents to create a document image; and a document image storing means for storing the document image created by the image reading means. A buffer, character recognition means for recognizing characters in the document image stored in the document image storage buffer, and a document based on the analysis result of the document image stored in the document image storage buffer and the character recognition result by the character recognition means. Document delimiter information extracting means for extracting document delimiter information for determining a delimiter, document delimiter determining means for determining a document delimiter based on the document delimiter information extracted by the document delimiter information extracting means, A document break candidate selection means for displaying a document break determination result by the document break determination means to correct / confirm the document break; Document management system registration means for dividing a document image into document units and registering the document image in a document management system using a result of determination of a document break by the document break determination means or a break of the document corrected by the document break candidate selection means. Things.

[0014] The document delimiter information extracting means may include a start reserved word which is a frequently-used character string included in the first page of the document and an end reserved word which is a frequently-used character string included in the last page of the document. It is characterized in that it is extracted from the character recognition result of the document image, and the presence or absence of the reserved reserved words and the reserved reserved words is used as document delimiter information.

The document delimiter information extracting means extracts a representative word group representing the description content of the document from the result of character recognition of the document image, and compares it with the word group extracted from the previous page by the same means. Thus, the similarity of the description contents between the document images is obtained, and this is used as document delimiter information.

Further, the document delimiter information extracting means analyzes the document image in an image processing manner, obtains the size of the blank area at the upper end and the lower end thereof, and uses this as the document delimiter information. Things.

Further, the document separation information extracting means obtains the paper size of the document image from the width and height of the document image, and uses the difference between the paper size obtained from the previous page as the document separation information. Things.

The document delimiter information extracting means obtains an average size of characters included in the document image based on the character recognition result of the document image, and determines a difference between the average size of the characters obtained from the previous page and the document. It is characterized in that it is used as delimiter information.

The document delimiter information extracting means checks the directionality (vertical writing / horizontal writing) of a character string included in the document image based on the character recognition result of the document image, and determines the character string obtained from the preceding page. Is used as document delimiter information.

Further, the document delimiter information extracting means uses a difference between a character recognition result of a header area extracted from a target document image and a character recognition result of a header area of the preceding page as document delimiter information. It is a feature.

The document delimiter information extracting means examines a character recognition result for the page number area extracted from the target document image and a character recognition result for the page number area extracted from the document image of the preceding page. It is characterized in that whether page numbers are continuous or discontinuous is used as document delimiter information.

Further, the document break determining means determines whether the target document image is the first page of the document or another page by a discriminant analysis method in which the document break information is input. The document image is divided into document units with the document image determined to be a page as the top page.

Further, the means for selecting a document break candidate includes:
When displaying the result of determination of a document break obtained by the document break determination means on a display, the display method is changed for each document break based on the reliability of the determination result.

[0024]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to FIGS. FIG. 1 is a configuration diagram showing an automatic document dividing apparatus according to an embodiment of the present invention. In FIG. 1, reference numeral 101 denotes an image reading unit that reads a plurality of documents to create a document image; 102, a document image storage buffer that stores the read document image; 103, a character recognition unit that recognizes characters in the document image; Document delimiter information extracting means for extracting document delimiter information for determining a document delimiter from a document image analysis result and a character recognition result,
Reference numeral 105 denotes a document delimiter determining unit that determines a document delimiter based on document delimiter information, 106 denotes a document delimiter candidate selector that displays a document delimitation determination result and corrects / confirms a document delimiter, and 107 denotes a document image. Document management system registration means for registering in the document management system in units of documents.

FIG. 2 is a diagram showing an example of an input document image. FIG. 3 is a block diagram of the document delimiter information extracting means 104, wherein 301 is an item classifying means, 302 is a start / end reserved word extracting means, 303 is a description content extracting means, 304
, 305 is a paper size extracting unit, 306 is a character size / character string direction calculating unit, 30 is
7 is item classification information, 308 is a list of reserved words for start and end, 3
Reference numeral 09 denotes a language dictionary, and reference numeral 310 denotes a document break information generating unit.

FIG. 4 is a diagram showing an example of the result of the item classifying means. Reference numeral 401 denotes a document title area, 402 denotes a header area, and 403 denotes a page number area. FIG. 5 is a diagram showing an example of a start / end reserved word list. FIG.
Is a diagram showing an example of a start / end reserved word extraction result;
1 indicates a start reserved word, and 602 indicates an end reserved word.

FIG. 7 is a diagram showing an example of the result of the description content extracting means. FIG. 8 is a diagram showing an example of a result of the sentence end sentence end blank area extracting means.
802 is a sentence blank area, and 803 is a sentence blank area.
FIG. 9 is a diagram illustrating an example of the result of the sheet size extracting unit.

FIG. 10 is a diagram showing a result example of the character size / character string direction calculating means. FIG. 11 is a diagram showing an example of the result of the document break determination unit. Reference numeral 1101 denotes one divided document. FIG. 12 is a diagram showing a display example 1 of a document break candidate. Reference numeral 1201 denotes a mark indicating a document break position. Further, FIG. 13 is a diagram showing a display example 2 of the document break candidates, and 1301 is a mark indicating a break position of the document.

Hereinafter, the operation of this embodiment will be described with reference to FIGS. First, the image reading means 101 of FIG.
A bundle of documents to be registered in the document management system is read by a scanner or the like to create a document image. Then, the created document image is stored in the document image storage buffer 102. Here, the document bundle is a bundle of a plurality of documents having various formats as shown in FIG. 2, and is, for example, a paper bundle obtained by directly extracting a document group closed in a paper file.

Next, the character recognizing means 103 performs a character recognizing process on the document image stored in the document image storage buffer 102, and converts the contents of the document into text. Next, the document delimiter information extracting unit 104
03 and the document image storage buffer 10
The document image stored in the document image 2 is analyzed, and document delimiter information serving as an index for determining a document delimiter is extracted for each document image. Hereinafter, this processing content will be described in detail with reference to the configuration diagram shown in FIG.

First, the item classifying means 301 extracts a document title area, a header area, and a page number area from a document image. This processing is described in "Document Filing Apparatus and Document Filing Method" in JP-A-10-240901.
Can be realized by the method described in (1). More specifically, according to the layout rules of the document structure stored in advance in the item classification information 307 (document title is above the document image), and the description contents of each document item (such as chapters containing numbers),
It is realized by analyzing the character recognition result. For example, FIG.
As a result of applying this processing to the document example of FIG. 4, a document title area 401, a header area 402, and a page number area 403 shown in FIG.
Get.

Next, the start / end reserved word extracting means 302
First, a start reserved word and an end reserved word described in the start reserved end word list 308 are read. Here, the start reserved word is
The character string frequently included in the first page of the document, and the end reserved word is the character string frequently included in the last page of the document.
For example, in the start / end reserved word list 308, as shown in FIG. 5, "beginning" and "abstract" are described as start reserved words, and "summary", "over" and "finally" are described as end reserved words. Thereafter, the start / end reserved word extracting unit 302 searches the text of the character recognition result for a character string that matches the start reserved word or the end reserved word. For example, as a result of applying this processing to the document example in FIG. 2, a character string 601 matching the start reserved word and a character string 602 matching the end reserved word are obtained as shown in FIG.

Next, the description content extracting means 303 extracts a representative word group included in the contents of the document according to the contents of the language dictionary 309 and compares it with the word group extracted from the previous page by the same means. Thus, the similarity of the description contents between the document images is obtained. For example, when the language information 309 is a list of words, it is determined whether or not the words are included in the text of the character recognition result. FIG. 7 shows the result of applying this processing to the document example of FIG. FIG. 7 shows what words are included in the form image, and a circle in the figure means that a word is included. Here, for example, the form image D1 includes “face image”, “recognition”, “survey”, and “result”, and the form image D2 includes “face image”, “recognition”, and “survey”. You can see that it is.

Next, the head-end sentence end blank area extracting means 304 analyzes the document image by image processing and detects blank areas at the upper end and the lower end of the document. Here, the blank area at the upper end of the document is called the head blank area, and the blank area at the lower end is called the end blank area. For example, as shown in FIG. 8, the detection of the head blank region is performed by first creating a histogram 801 in which the number of black pixels on the horizontal line of the document image is counted. Count the number of horizontal lines where is 0. If the number of the horizontal lines is equal to or more than a certain number, the region where the horizontal lines are bundled is determined as the head blank region. Here, the height of the head blank area is represented by the number H of horizontal lines. Similarly, the end-of-sentence blank area is extracted by performing the above processing from the lower end to the upper end of the document. As a result of applying this processing to the example of the document image in FIG.
And a sentence end blank area 803 is obtained.

Next, the paper size extracting means 305 obtains the paper size (A4 length, A4 width, B5 length, etc.) of each document image. This paper size can be easily obtained from the width and height of each document image. For example, by applying this processing to the document image example of FIG. 2, the result shown in FIG. 9 is obtained.

Next, character size / character string direction calculating means 3
In step 06, the average size of the characters included in the document image and the direction of the character string (vertical writing, horizontal writing) are obtained. The average size of a character can be calculated by averaging the width and height of a character image recognized as one character during the character recognition process. Also,
The direction of the character string is determined as vertical writing when there are many vertical writing character strings in the document image, and as horizontal writing when there are many horizontal writing character strings.
Since the vertical and horizontal writing of each character string is determined in the course of general character recognition processing, this result is used.

Finally, the document break information generating means 310
Generates document delimiter information included in each document image according to the presence / absence of the start / end reserved words obtained in the above process and the description contents. Here, the document separation information S of the i-th document image
i is represented by a 10-dimensional vector shown in the following equation (1). Document delimiter information Si = {x1, x2,..., X10} (1)-x1: Presence or absence of reserved reserved word-x2: Presence or absence of reserved reserved word-x3: Similarity of description contents-x4: Blank of sentence head Height x5: Height of sentence end blank area x6: Difference in paper size x7: Difference in average character size x8: Difference in character string direction x9: Difference in character recognition result in header area x10: Page number Continuity

In equation (1), the information amount x1 is a value indicating the presence or absence of a reserved reserved word, for example, the number of reserved reserved words included in the document image i. Similarly, the information amount x2 is the number of end reserved words included in the previous page.

The information amount x3 is a value indicating the similarity of the description contents between the document images, and has a larger value as the description contents are more similar between the document image i and the previous page. As the value, for example, a word group extracted from the document image i by the description content extraction unit 303 and a word group extracted from the previous page.
The number of words included in both of the word groups extracted in step (1) can be used. In this case, for example, in the example of the result of the description content extracting means shown in FIG.
3. x3 = 0 for the document image D4.

Next, the information amount x4 is a value indicating the presence or absence of the head blank area, for example, the height H of the head blank area included in the document image i. The information amount x5 is a value indicating the presence or absence of the end-of-sentence blank area, for example, the height H of the end-of-sentence blank area included in the document image i.

The information amount x6 is a value indicating the difference in paper size between document images. For example, if the document image i and the previous page have the same paper size, the value is 0; Specifically, in the example of the result of the paper size extracting unit 305 shown in FIG. 9, the document image D2 has the same paper size as the document image D1, so x5 = 0, and the document image D7 has a different paper size from the document image D6. X5 = 1.

The information amount x7 is a value indicating the difference in the average character size between the document images, and is the difference between the average character size of the character image i and the average character size of the previous page. For example,
In the processing result example of the character size / character string direction calculating unit 306 shown in FIG. 10, since the document image D2 has the same average character size as the document image D1, x7 = 0, and the document image D4
Since the average character size is 2 larger than the document image D3, x7 = 2.

The information amount x8 is a value indicating the difference in the character string direction between the document images. The information amount x8 has the value 0 when the character string direction is the same between the document image i and the previous page, and has the value 1 when the document image i is different. For example, the character size / character string direction calculating means 30 shown in FIG.
6, the document image D2 is the document image D1
And the character string direction is the same, x8 = 0, the document image D6
Since the character string direction is different from that of the document image D5, x8 = 1.

The information amount x9 is a value indicating the difference in the character recognition result in the header area between the document images. When the character recognition result in the header area is the same between the document image i and the previous page, the value 0 is set.
If different, it has the value 1. For example, in the example of the result of the item classifying means shown in FIG. 4, the same character recognition result as "report A" is obtained in the header area of the document image D2 and the document image D1, so that the document image D2 has x9 = 0. Become. Although the document image D4 has no header area, the document image D4
In No. 3, since the character recognition result “Report A” is selected from the header area, the document image D4 has x9 = 1.

Finally, the information amount x10 is a value indicating the continuity of the page number between the document images. When the value of the character recognition of the page number area between the document image i and the preceding page is a continuous value, It has a value of 0, and a value of 1 if it is discontinuous. For example, in the result example of the item classifying means shown in FIG. 4, the page number of the document image D9 is 7
32, the page number of the document image D8 is 731 and the page numbers are continuous. Therefore, the document image D9 has x10 = 0. Conversely, the page number of the document image D10 is 760, and the page number of the document image D9 is discontinuous with 731. Therefore, the document image D10 has x10 = 1. Similarly, the page number 731 is obtained in the document image D8, but since there is no page number in the document image D7, the page numbers are discontinuous and the document image D8 has x10 = 1. The above is the document separation information extracting means 10
This is the operation of No. 4.

Next, the operation of the document break determination means 105 will be described. The document break determination unit 105 estimates a break position of a document based on the document break information S obtained by the document break information extraction unit 104. Specifically, a document image i
Is classified as a first page of the document, or a non-separator class, which means that it is not the first page of the document.The document image classified into the first page is classified as a first page. Split the document. For example, as a result of applying this processing to the document image example of FIG. 2, as shown in FIG. 11, the documents D1, D4, D6, D7, D8, and D10 are classified into the division page class, and these are divided as the first page. A document group 1101 is obtained.

Here, the classification into the break page class and the non-break page class is described in, for example, "PC Statistical Analysis Handbook II".
Multivariate Analysis ”(Kyoritsu Shuppan Co., Ltd.). The discriminant analysis method is one of multivariate analysis methods, and includes a set of document delimiter information S included in a document image belonging to a preparatory page class and a set of document delimiter information S included in a document image belonging to a non-separable page class. Is used as learning data to generate an identification function, and using this to determine to which class the unknown document image i belongs.
Specifically, when the document separation information Si obtained from the document image i is input to the generated identification function, a distance value Gi indicating a deviation between the document image i and each class is calculated, and the distance value G is calculated.
Classify into the class with the smaller i.

Next, the document break candidate selection means 106
The result of the automatic division obtained by the document break determination means 105 is displayed to the operator for confirmation and correction. However, when it is not necessary for the operator to confirm / correct the document, the document separation candidate selection unit 106 may be skipped and the document may be registered in the document management system by the document management system registration unit 107.

The document segment candidate selection means 106 displays the document image and the result of the document segment decision on a display.
For example, as a result of applying this processing to the document image example shown in FIG. 2, the screen of FIG. 12 is displayed on the display. FIG.
In the figure, reference numeral 1201 denotes a mark indicating a document break position automatically determined. The operator looks at this display,
Check whether there is an error in the determination result of the document break, and if there is an error, correct it by operating a keyboard or the like. Specifically, when an extra portion is determined as a document break, the document break is deleted. Conversely, if the correct document break has not been extracted, the document break is added thereto.

In this case, when displaying the determination result of the document break on the display, the display method may be changed for each document break based on the reliability of the determination result. Here, the value of the degree of reliability can be calculated, for example, from the distance value Gi between the document image i obtained by the document break determination means 105 and the page break class. Since the smaller the distance value Gi is, the more likely it is that the document image i is a document break, the reliability can be defined as 1 / Gi. For example, the display screen shown in FIG. 13 displays the line width of the mark 1301 indicating the document break position thicker as the reliability is higher. As a result, the operator only needs to check the places with low reliability, and the work efficiency can be improved.

Finally, the document management system registration means 107
Divides the document image into document units by using the document break determination result obtained by the document break determination means 105 or the document break corrected by the document break candidate selection means 106;
Register in the document management system.

The above is the operation of the present apparatus. In the above embodiment, the scanner is described as an example of the image reading unit 101, but another device such as a facsimile may be used.

According to the embodiment, the document image is analyzed to extract the document separation information, and the separation of the document is automatically determined based on this information. Therefore, there is no need to manually insert a partition sheet between documents in advance, and the operation can be simplified. In addition, the document delimiter information uses various information that can be extracted from a general document, such as the presence or absence of a reserved reserved word, the similarity of the description content, the height of a blank area at the beginning and end of a sentence, the difference in paper size, and the like.
The format is not specific to known fixed-form documents. Therefore,
Even when a variety of unknown documents are input, the separation of the documents can be automatically determined.

[0054]

As described above, according to the present invention, image reading means for reading a plurality of documents to create a document image,
A document image storage buffer for storing a document image created by the image reading unit, a character recognition unit for recognizing characters in the document image stored in the document image storage buffer, and a character image stored in the document image storage buffer. Document delimiter information extracting means for extracting document delimiter information for determining a document delimiter from the analysis result of the document image and the character recognition result by the character recognizing means, and the document delimiter information extracted by the document delimiter information extracting means. Document delimiter judging means for judging a document delimiter based on the source, document delimiter candidate selecting means for displaying a document delimitation judgment result by the document delimiter and correcting and confirming the document delimiter; The document is determined using the determination result of the document break according to the document break or the document break corrected by the document break candidate selecting means. Document management system registration means for dividing images into document units and registering them in the document management system eliminates the need for partition paper and allows for the separation of documents even when a variety of unclear documents are input. Can be determined automatically.

Further, the document delimiter information extracting means may include a start reserved word which is a character string frequently included in the first page of the document and an end reserved word which is a character string frequently included in the last page of the document. Since it is extracted from the character recognition result of the document image and the presence or absence of the reserved reserved word and the reserved end word is used as document delimiter information, it is possible to determine that accurate document delimiter information is to be extracted.

The document delimiter information extracting means extracts a representative word group representing the description content of the document from the result of character recognition of the document image, and compares it with the word group extracted from the previous page by the same means. Thus, the similarity of the description contents between the document images is obtained and used as the document delimiter information, so that it is possible to determine that accurate document delimiter information is extracted.

Further, the document separation information extracting means analyzes the document image by image processing, finds the size of the blank area at the upper end and the lower end, and uses this as the document separation information. It can be determined that accurate document separation information is extracted.

Further, the document separation information extracting means obtains the paper size of the document image from the width and height of the document image and uses the difference from the paper size obtained from the previous page as the document separation information. It can be determined that accurate document separation information is extracted.

Further, the document delimiter information extracting means calculates the average size of the characters included in the document image based on the character recognition result of the document image, and determines the difference from the average size of the characters obtained from the previous page. Since it is used as delimiter information, it can be determined that accurate document delimiter information is extracted.

The document delimiter information extracting means checks the directionality (vertical writing / horizontal writing) of a character string included in the document image based on the character recognition result of the document image, and determines the character string obtained from the previous page. Is used as document delimiter information, so that accurate document delimiter information can be determined to be extracted.

Further, the document delimiter information extracting means uses the difference between the character recognition result of the header area extracted from the target document image and the character recognition result of the header area of the previous page as document delimiter information. Therefore, it can be determined that accurate document separation information is extracted.

The document delimiter information extracting means checks the character recognition result for the page number area extracted from the target document image and the character recognition result for the page number area extracted from the previous page document image. Since whether the page numbers are continuous or discontinuous is used as document delimiter information, it can be determined that accurate document delimiter information is to be extracted.

The document break determining means determines whether the target document image is the first page of the document or another page by a discriminant analysis method using the document break information as an input. Since the document image is divided into document units with the document image determined to be a page as the top page, the break of the document can be accurately determined.

Further, the document segmenting candidate selection means includes:
When displaying the determination result of the document break obtained by the document break determination means on the display, the display method is changed for each document break based on the reliability of the determination result.
It is possible to accurately select a document break candidate.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing an automatic document dividing apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing an example of an input document image.

FIG. 3 is a configuration diagram of a document separation information extraction unit 104 of FIG. 1;

FIG. 4 is a diagram showing an example of a result of an item classifying unit.

FIG. 5 is a diagram showing an example of a start / end reserved word list.

FIG. 6 is a diagram illustrating an example of a result of a start / end reserved word extracting unit;

FIG. 7 is a diagram illustrating an example of a result of a description content extracting unit.

FIG. 8 is a diagram showing an example of a result of a sentence end sentence end blank area extracting unit.

FIG. 9 is a diagram illustrating an example of a result of a paper size extracting unit;

FIG. 10 is a diagram illustrating an example of a result of a character size / character string direction calculating unit.

FIG. 11 is a diagram illustrating an example of a result of a document break determination unit.

FIG. 12 is a diagram showing a display example 1 of a document break candidate.

FIG. 13 is a diagram showing a display example 2 of a document break candidate.

FIG. 14 is a processing flowchart of a conventional method.

FIG. 15 is a diagram showing an example of a paper document registered by a conventional method.

[Explanation of symbols]

101 image reading means, 102 document image storage buffer, 103 character recognizing means, 104 document break information extracting means, 105 document break determining means, 106 document break candidate selecting means, 107 document management system registration means,
301 item classification means, 302 start / end reserved word extraction means, 303 description content extraction means, 304 sentence end sentence blank area extraction means, 305 paper size extraction means, 306 character size / character string direction calculation means, 307 item classification information, 308 start End reserved word list, 309 language dictionary, 310 document break information generation processing, 401 document title area, 402 header area, 403 page number area, 601 start reserved word, 602 end reserved word, 801
Histogram, 802 blank area at the beginning of sentence, 803 blank area at the end of sentence, 1101 delimited document, 1201 mark indicating document break position, 1301 mark indicating document break position, 1501 partition paper, 1502 one document consisting of multiple pages , 1503 Specific image area, 1
504 Data description area.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04N 1/387 H04N 1/387 F-term (Reference) 5B050 BA06 BA10 BA16 EA17 GA08 5B064 AA01 AA07 BA01 5B075 ND03 NR03 PP04 UU06 5C076 AA01 AA36 BA03 BA06 CA10

Claims (11)

[Claims] An image reading unit that reads a plurality of documents to create a document image; a document image storage buffer that stores a document image created by the image reading unit; and a document image stored in the document image storage buffer. A character recognizing unit for recognizing a character in the document, and a document for extracting document delimiter information for determining a document delimiter from the analysis result of the document image stored in the document image storage buffer and the character recognition result by the character recognizing unit. Displaying a delimiter information extracting unit, a document delimiter determining unit that determines a document delimiter based on the document delimiter information extracted by the document delimiter information extracting unit, and displaying a document delimiter determination result by the document delimiter determining unit. A document break candidate selection unit for correcting and confirming a document break, and a determination result of a document break by the document break determination unit Alternatively, an automatic document dividing apparatus comprising: a document management system registration unit that divides a document image into document units and registers the document image in a document management system by using a document segment corrected by the document segment candidate selection unit. 2. The automatic document dividing apparatus according to claim 1, wherein the document delimiter information extracting unit includes a start reserved word that is a frequently-used character string included in a first page of the document and a start reserved word included in a last page of the document. An automatic document segmentation apparatus characterized in that an end reserved word that is a frequently-used character string is extracted from a character recognition result of a document image, and the presence or absence of the start reserved word and the end reserved word is used as document delimiter information. 3. The automatic document dividing device according to claim 1, wherein the document delimiter information extracting unit extracts a representative word group representing the description content of the document from a character recognition result of the document image,
An automatic document segmentation apparatus characterized in that similarity of description content between document images is obtained by comparing with a word group extracted by a similar means from the previous page, and this is used as document delimiter information. 4. The automatic document dividing apparatus according to claim 1, wherein said document delimiter information extracting means analyzes the document image in an image processing manner, and extracts a blank area at an upper end and a lower end thereof. An automatic document dividing apparatus for determining a size and using this as document delimiter information. 5. The automatic document dividing apparatus according to claim 1, wherein the document delimiter information extracting unit obtains a paper size of the document image from a width and a height of the document image, and obtains a paper size from a previous page. An automatic document dividing apparatus characterized in that a difference from the paper size is used as document separation information. 6. The automatic document dividing apparatus according to claim 1, wherein the document delimiter information extracting unit determines an average size of characters included in the document image based on a character recognition result of the document image. An automatic document dividing apparatus characterized in that a difference between the calculated character and the average size of characters obtained from the previous page is used as document delimiter information. 7. The automatic document dividing device according to claim 1, wherein the document delimiter information extracting unit determines a direction of a character string included in the document image based on a character recognition result of the document image. An automatic document dividing apparatus which checks (vertical writing / horizontal writing) and uses the difference from the directionality of the character string obtained from the previous page as document separation information. 8. The automatic document dividing apparatus according to claim 1, wherein said document delimiter information extracting means includes a character recognition result of a header area extracted from a target document image, An automatic document dividing apparatus, wherein a difference from a character recognition result for a header area is used as document delimiter information. 9. The automatic document dividing apparatus according to claim 1, wherein said document delimiter information extracting means includes: a character recognition result for a page number area extracted from a target document image; An automatic document dividing apparatus for examining a character recognition result for a page number area extracted from a document image of a document, and using whether the page number is continuous or discontinuous as document delimiter information. 10. The automatic document segmentation apparatus according to claim 1, wherein the document segment determination unit determines a target document image by a discriminant analysis method using document segment information as an input. An automatic document dividing apparatus which determines whether a page is a page or another page, and divides the document image into document units using the document image determined as the first page of the document as the first page. 11. The automatic document dividing apparatus according to claim 1, wherein the document segment candidate selecting unit performs a decision when displaying a document segment decision result obtained by the document segment determining unit on a display. An automatic document dividing apparatus, wherein a display method is changed for each document break based on the reliability of a result.