JP2023005163A - Document processing method and document processor - Google Patents

Abstract

To provide a mechanism for accurately discriminating boundaries of documents of various configurations from image data obtained by reading the documents.SOLUTION: A document creation method comprises steps of acquiring a document as image data page by page, and detecting a sign of a boundary page that satisfies a predetermined condition for each page from the acquired image data. In addition, the document creation method comprises steps of identifying, on the basis of a page containing the sign of detected boundary page and information of adjacent pages adjacent to the page, one or more boundary pages that form boundaries of the document, and creating document data divided at the identified one or more boundary page locations.SELECTED DRAWING: Figure 6

Description

The present invention relates to a document processing method and a document processing apparatus.

In a conventional document management system, an operator visually checks the contents of a group of document images read by a scanner and uses a tool to specify the position of document separation, dividing the group of document images into document units and registering them. Was. For this reason, the operator has to work hard to determine the document delimiter position, which hinders the document registration work.

As one of the countermeasures, in Patent Documents 1 and 2, when a document is read by a scanner, the document is semi-automatically divided by inserting a partition sheet that can be easily distinguished by the machine in advance between the documents. A method is proposed. As another countermeasure, in Patent Document 3, if there is a blank area at the bottom of the page, it is regarded as the last page of the document, and the page following the last page has a header area. A method for judging whether the order is correct or not has been proposed. In Patent Document 4, information in the page of interest (presence or absence of reserved words, height of blank area, character recognition result in header area) and comparison information with the previous page (similarity of words, difference in paper size, average character size , and character string direction) to detect discontinuity of documents. In this way, the first page can be determined by detecting the discontinuity of the document.

Japanese Patent No. 2962961 JP 2005-141425 A Japanese Patent Application Laid-Open No. 2020-120308 JP-A-2002-312385

However, the conventional technology described above has the following problems. For example, in the method of dividing a document using partition sheets, there is a problem that it is necessary to manually insert the partition sheets between documents in advance. In the method of determining the last page of the document when there is a blank area at the bottom of the page, or the method of determining the first page by detecting the discontinuity of the document, part of the table or figure is the next page of the last page of the text There is a problem that the document delimiter is erroneously determined when it is published later. Alternatively, in the case of a bound document, a blank page for page alignment may be inserted after the main text, and in this case also, the document delimiter is erroneously determined.

SUMMARY OF THE INVENTION The present invention has been made in view of at least one of the above problems, and provides a mechanism for accurately discriminating the boundaries of documents of various configurations from image data obtained by reading the documents. do.

The present invention is, for example, a document creation method, comprising: an acquisition step of acquiring image data of a document page by page; an identifying step of identifying one or more boundary pages that form a boundary of a document based on the page including the sign of the boundary page detected in the detecting step and the information of adjacent pages adjacent to the page; and a creating step of creating document data divided at the positions of the one or more boundary pages specified in the specifying step.

According to the present invention, it is possible to accurately determine the boundaries of documents of various configurations from image data obtained by reading the documents.

1 is a diagram showing an example hardware configuration of a document processing apparatus according to an embodiment; FIG. 1 is a diagram showing an example functional configuration of a document processing apparatus according to an embodiment; FIG. 4A and 4B are diagrams for explaining a method of detecting a blank area according to an embodiment; FIG. 4A and 4B are diagrams for explaining a document boundary specifying method according to an embodiment; FIG. 4A and 4B are diagrams for explaining a document boundary specifying method according to an embodiment; FIG. 4 is a flowchart related to processing of a document processing device according to an embodiment;

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. In addition, the following embodiments do not limit the invention according to the scope of claims. Although multiple features are described in the embodiments, not all of these multiple features are essential to the invention, and multiple features may be combined arbitrarily. Furthermore, in the accompanying drawings, the same or similar configurations are denoted by the same reference numerals, and redundant description is omitted.

<First Embodiment>
<Hardware Configuration of Document Processing Device>
An embodiment of the invention is described below. First, with reference to FIG. 1, an example of the hardware configuration of the document processing apparatus according to this embodiment will be described. Here, an image processing apparatus will be described as an example of a document processing apparatus. In the present embodiment, document image page data (hereinafter also simply referred to as image data) obtained by reading document data handled by a document processing apparatus with a scanner will be described as an example. For example, it may process document data input from the outside. Therefore, the document processing apparatus does not need to include a reading unit such as a scanner.

The document processing apparatus 100 includes a CPU 101, a RAM 102, a ROM 103, a storage device 104, a scanner 105, and a communication I/F 106, and each device can exchange data via a system bus 111. FIG. The CPU 101 executes programs stored in the ROM 103 and programs such as an OS (Operating System) and applications loaded from the storage device 104 to the RAM 102 . That is, the CPU 101 functions as each processing unit that executes the processing of each flowchart described later by executing the program stored in the readable storage medium. A RAM 102 is a main memory of the CPU 101 and functions as a work area and the like. The scanner 105 reads a document page by page as document image page data, and generates image data for specifying boundary pages between documents. In the present invention, document data for specifying boundary pages between documents is not limited to image data read from a document by the scanner 105, and may be image data received from an external device. A communication I/F 106 is an interface for communicating with an external device via a network, and receives image data to be processed from the external device, for example. A system bus 111 is a communication path for communication between hardware elements.

<Functional block>
Next, the functional configuration of the document processing apparatus 100 according to this embodiment will be described with reference to FIG. Each functional unit in the figure is implemented by the CPU 101 executing programs such as programs stored in the ROM 103 and applications loaded from the storage device 104 to the RAM 102 . Execution results of each process are held in the RAM 102 .

The document processing apparatus 100 includes a page acquisition unit 201, a blank area detection unit 202, a border page identification unit 203, an electronic document creation unit 204, an auxiliary information area detection unit 205, and an additional information confirmation unit 206 as functional configurations. A page acquisition unit 201 acquires page data, which is image data to be processed, from the scanner 105 . Alternatively, the page acquisition unit 201 may acquire image data to be processed from an external device via the communication I/F 106 .

The blank area detection unit 202 detects a blank area for each page of the image data to be processed acquired by the page acquisition unit 201 . Also, the auxiliary information area detection unit 205 detects an auxiliary information area on the page corresponding to the image data. In this embodiment, the auxiliary information area includes a footer area, a footnote area, and a page number area. may be The blank area detection unit 202 also uses the detection result of the auxiliary information area detection unit 205 to detect blank areas.

A boundary page identification unit 203 identifies a boundary page of a document based on the detection result of the blank area detection unit 202 for image data to be processed. The additional information confirmation unit 206 confirms whether the image data to be processed (document image page data) is additional information in the document content. The additional information is information indicating contents continued from the previous page, such as "column", "reference information", and "index". The boundary page specifying unit 203 also uses the check result of the additional information checking unit 206 to specify the border page. Details of the identification method will be described later. The electronic document creation unit 204 creates document data including boundary information based on the identification result of the boundary page identification unit 203 .

<Blank area detection>
Next, a blank area detection method of the document processing apparatus 100 according to the present embodiment will be described with reference to FIG. Reference numeral 300 denotes an image corresponding to one page of document image page data read from the document.

301 indicates the document area and 302 indicates the detected blank area. The blank area 302 includes a footnote area 303 as an auxiliary information area. The auxiliary information area detection unit 205 according to this embodiment analyzes the document image page data and detects the footnote area 300 as an auxiliary information area. Here, the auxiliary information area detection unit 205 detects ruled lines, character strings, and the like existing in the area of interest as the auxiliary information area based on the position information of the area of interest being processed (for example, a predetermined area at the bottom of the page). The blank area detection unit 202 controls the footnote area 300 detected by the auxiliary information area detection unit 205 as a blank area. Next, the blank area detection unit 202 detects a blank area from the lower end of the document while regarding the footnote area 300 as a blank area. In the example of FIG. 3, a blank area 302 is detected and its height is indicated at 304 .

More specifically, the blank area detection unit 202 creates a histogram that counts the number of black pixels on the horizontal line of the document image page data, and counts the number of horizontal lines with a histogram value of 0 from the lower end to the upper end of the document. is counted, and the count result is used as the height of the blank area. Note that the present invention is not limited to this, and any other method may be used to detect the height of the blank area. Also, the area of the blank area may be obtained instead of the height. The height of the detected blank area 302 is used to identify the boundaries of the document, as will be described later.

<Identification of document boundaries>
Next, a document boundary specifying method of the document processing apparatus 100 according to the present embodiment will be described with reference to FIGS. 4 and 5. FIG. Reference numerals 400, 410, 420, and 430 in FIG. 4 denote images corresponding to the document image page data of each page read from the document. Each page is provided with a page number area in the lower right to indicate the number of each page. On each page, areas 401, 412, 421, and 431 indicate document areas, respectively. Area 411 indicates the drawing area. Areas 402 and 422 indicate blank areas.

In the first page 400, the blank area detection unit 202 detects a blank area 402 with a height indicated by 403 at the bottom. Here, the boundary page specifying unit 203 determines whether the blank area 402 detected by the blank area detecting unit 202 has a predetermined size (for example, a predetermined height), and determines if the blank area 402 has a predetermined size. page is identified as a boundary candidate. Once identified as a boundary candidate, the boundary page identification unit 203 acquires the document image page data of the next page and detects additional information indicating that the page is a page whose content continues from the previous page.

In this embodiment, it is determined whether the acquired document image page data contains additional information when at least one of the following two conditions is satisfied. The first condition is when the acquired document image page data is a sentence that begins with a predetermined character string. For example, "sample", "column", "reference information", "appendix", and "appendix" can be set as the predetermined character string. The setting of the character string is arbitrary and can be changed according to user input. The second condition is when a drawing area or a table area is detected at the top of the acquired document image page data. Note that the control may be such that the character strings listed as the first condition are regarded as boundaries. That is, the boundary may be determined between the page containing the character string and the previous page. These settings can be arbitrarily set by the user.

In the example of FIG. 4, the drawing area 411 exists in the upper part of the second page 410 . In other words, it can be inferred that the figure drawn in the drawing area 411 is shifted to the top of the second page 410, which is the next page, without fitting into the blank area 403 at the bottom of the first page 400. can be done. That is, the boundary page identifying unit 203 determines that the first page 400 registered as a boundary candidate of the document is not actually a boundary candidate, but continues to the second page 410 and subsequent pages.

Since there is no blank area on the second page 410, the processing target is shifted to the third page 420, which is the next page. A blank area 422 is identified on the third page 420, and this page is set as a new boundary candidate. Subsequently, the boundary page specifying unit 203 acquires and analyzes the document image page data of the next page. The fourth page 430 has a document area continuing from the top, and a predetermined character string or chart is not detected. Therefore, the boundary page specifying unit 203 determines that a document boundary 440 exists between the third page 420 and the fourth page 430 set as boundary candidates, and registers the boundary.

As described above, the document processing apparatus 100 according to the present embodiment detects a blank area of a predetermined size existing at the bottom, a page in which a blank area of a predetermined size is detected, and subsequent pages (hereinafter referred to as adjacent pages). ) to identify document boundaries. That is, in addition to blank areas, document boundaries are identified based on adjacent page information.

FIG. 5 shows an example of a contiguous page containing additional information. 500 includes a character string 501 “sample” and a document area 502 . 510 indicates a blank page (blank paper). Note that the blank page is an adjustment page for page alignment. Therefore, the boundary page identification unit 203 determines that both 500 and 510 contain additional information. When the boundary page identifying unit 203 detects a page including these additional information, it determines that the page identified as the boundary candidate is not a boundary and that there is a continuation page.

<Processing procedure>
Next, with reference to FIG. 6, a processing procedure for identifying boundary pages of document image page data in this embodiment will be described. The processing described below is realized by, for example, the CPU 101 reading a program stored in the ROM 103 or the storage device 104 into the RAM 102 and executing the program.

In S100, the page acquisition unit 201 uses the scanner 105 or the like to read a bundle of documents to be automatically divided page by page as document image page data. Here, a bundle of documents is a bundle of documents having miscellaneous formats, for example, a paper bundle obtained by directly extracting a group of documents bound in a paper file. Note that the document image data read by the external device may be acquired using the communication I/F 106 . Subsequently, in S111, the CPU 101 determines whether all pages (document image page data) of the document image data read in S100 have been processed. If the processing has been completed, the process proceeds to S112, and if the processing has not been completed, the process proceeds to S101.

In S101, the CPU 101 acquires the unacquired first page from the read document image data as document image page data. Here, an unacquired page indicates a page for which processing has not been completed. When the page is acquired, in S102 the auxiliary information area detection unit 205 detects the auxiliary information area of the acquired page. This processing can be realized by analyzing the layout using, for example, layout rules and ruled line information of the document. Alternatively, the auxiliary information area may be detected when the positional information of the character string obtained from the character recognition result, the font information of the character string, or the predetermined character symbol is included in the character string. Furthermore, it is also possible to use the positional information of the detected auxiliary information area (region of interest) to select the correctly detected auxiliary information area. Here, the auxiliary information area detection unit 205 uses the processing described above to detect auxiliary information areas such as the footer area, footnote area, and page number area positioned at the bottom of the page.

Next, in S103, the blank area detection unit 202 registers the auxiliary information area detected by the auxiliary information area detection unit 205 as a blank area. In other words, when the document image data is monochrome image data, the blank area detection unit 202 performs control so that the colored pixels in the area detected as the auxiliary information area are recognized as white pixels. Here, there is no need to actually change the pixel values of the document image data, and the area is recognized as a blank area and processed. For example, the coordinate position of the auxiliary information area is set and used in S104, which will be described later.

In S104, the blank area detection unit 202 analyzes the document image page data in terms of image processing to detect a blank area (page bottom blank area) at the bottom of the document image page data. Subsequently, the blank area detection unit 202 creates a final lower blank area by combining areas overlapping and adjacent to the lower blank area of the page, if any, in the auxiliary information area detected in S103. do.

Next, in S105, the boundary page specifying unit 203 determines whether the current page satisfies the document boundary candidate conditions, and branches the processing based on the determination result. In this embodiment, the conditions for document boundary candidates are determined by the position and size of the blank area at the bottom of the page. That is, when the bottom blank area of the page is in contact with the bottom edge of the page and the height of the bottom blank area of the page is equal to or greater than a predetermined height H, it is determined as a border page candidate (border page symptom) of the document. Here, the lower page blank area being in contact with the bottom edge of the page means that there is no area other than the blank area under the page lower blank area. The value of H is set in advance. It should be noted that it is also possible to dynamically set the value of H instead of setting it in advance. For example, the result of statistically processing the height of the blank area at the bottom of the page up to the previous page (for example, the average value) is set as the value of H, or the height of the blank area at the bottom of the page on the previous page is set as H It is also possible to configure to If the current page satisfies the document boundary candidate conditions, the process proceeds to S106; otherwise, the process returns to S111.

Next, in S106, the CPU 101 acquires the next document image page data from the document image data read in S100 in order to allow the additional information confirmation unit 206 to confirm whether additional information is written on the next page. Subsequently, in S107, the additional information confirmation unit 206 determines whether the document image page data acquired in S106 includes additional information regarding the previous page of the boundary candidate. Since the determination method has already been described with reference to FIGS. 4 and 5, details thereof will be omitted. Branch the processing based on the determination result. If it is determined that the acquired document image page data includes additional information, the process proceeds to S102. After that, the process from S102 to S105 is executed to search for the final page candidate (document boundary candidate) of the additional information. On the other hand, if it is determined that the acquired document image page data does not include additional information, or if the next document image page data cannot be acquired in S106, the process proceeds to S108.

In S108, the CPU 101 acquires the next document image page data from the document image data read in S100 in order to confirm whether the next page is a blank page. Subsequently, in S109, the boundary page specifying unit 203 determines whether the document image page data acquired in S108 is a blank page, and branches the processing based on the determination result. Blank page determination in this embodiment is performed by detecting a blank area from document image page data. A blank page is determined when the detected blank area is in contact with the top and bottom edges of the page. If the obtained document image page data is determined to be a blank page, the process proceeds to S108, and if not determined to be a blank page or if the next document image page data cannot be obtained in S108, the process proceeds to S110. .

In S110, the boundary page specifying unit 203 sets the document boundary assuming that the previously processed document image page data is the boundary page of the document, and returns the processing to S111. After that, if it is determined in S111 that all pages have been processed, the process proceeds to S112, and the electronic document creation unit 204 creates and processes an electronic document according to one or more document boundaries identified by the boundary page identification unit 203. exit.

As described above, in the document creation method according to the present embodiment, a document is acquired as image data on a page-by-page basis, and signs (blank areas) of boundary pages that satisfy predetermined conditions are identified for each page from the acquired image data. To detect. Further, the document creation method identifies one or more boundary pages that serve as boundaries of the document based on the page containing the detected blank area and information on adjacent pages adjacent to the page, and identifies the identified one or more boundary pages. Create document data divided at the position of the boundary page. As a result, if there are footnotes or page numbers at the bottom of the boundary page of the document, if figures or tables related to the document are placed on the following pages as separate tables, or if information that complements the content of the document is Document boundaries can be correctly identified even if they are added after the page. Furthermore, even if a blank page is added at the end of the document due to circumstances such as bookbinding, the document boundary can be specified correctly. As described above, according to the present embodiment, it is possible to accurately determine the boundaries of documents having various configurations from the image data obtained by reading the documents.

<Modification>
The present invention is not limited to the above embodiment, and various modifications are possible. In the above embodiment, the blank area detection unit 202 detects a blank area that satisfies a predetermined condition, and the document image page data having the blank area and a predetermined number of adjacent page data are analyzed to specify the page boundary. However, without being limited to this, when a sign other than a blank area is detected, a page boundary may be determined by analyzing a page having the sign and a predetermined number of pages. For example, the blank area detection section may be a boundary page sign detection section, and the boundary page sign detection section may detect a predetermined character string from the document image page data. The predetermined character string is, for example, "〆" or "above". Alternatively, the boundary page sign detection unit may be configured to detect a blank area in addition to detecting a predetermined character string from the document image page data. In this case, the page boundary can be specified more accurately by analyzing the document image page data having the blank area and the character string and the predetermined number of adjacent page data.

<Other embodiments>
The present invention supplies a program that implements one or more functions of the above-described embodiments to a system or device via a network or a storage medium, and one or more processors in the computer of the system or device reads and executes the program. It can also be realized by processing to It can also be implemented by a circuit (for example, ASIC) that implements one or more functions.

The invention is not limited to the embodiments described above, and various modifications and variations are possible without departing from the spirit and scope of the invention. Accordingly, the claims are appended to make public the scope of the invention.

101: CPU; 102: RAM; 103: ROM; Electronic document creation department

Claims (16)

A document creation method comprising:
an acquisition step of acquiring the document as image data page by page;
a detection step of detecting a sign of a boundary page that satisfies a predetermined condition for each page from the acquired image data;
an identifying step of identifying one or more boundary pages serving as document boundaries based on the page containing the sign of the boundary page detected in the detecting step and the information of adjacent pages adjacent to the page;
and a creating step of creating document data divided at the positions of the one or more boundary pages specified in the specifying step. 2. The document creation method according to claim 1, wherein said detection step detects a blank area that satisfies said predetermined condition as said boundary page sign. further comprising a confirmation step of confirming additional information indicating content continued from the previous page of the adjacent page in the adjacent page;
3. The document creation method according to claim 2, wherein, in said identifying step, said one or more boundary pages are identified using a confirmation result obtained in said confirming step. 4. The document creation method according to claim 3, wherein, in said identifying step, if said adjacent page contains additional information, said boundary page is identified from adjacent pages after said adjacent page. 5. The document creation method according to claim 3, wherein, in said confirmation step, when a predetermined character string or chart is included at the head of said adjacent page, it is determined that said additional information is included. 6. The document according to any one of claims 2 to 5, wherein in said detection step, an auxiliary information area containing auxiliary information of said document is detected, and said detected auxiliary information area is treated as a blank area. How to make. 7. The document creation method according to claim 6, wherein the auxiliary information area includes at least one of a footer area, a footnote area, and a page number area. 8. The document creation method according to claim 6, wherein in said detecting step, said auxiliary information area is detected based on positional information of an area of interest on a page. 9. The document creation method according to claim 8, wherein in said detecting step, said auxiliary information area is detected based on font information of characters included in said focused area. 10. The document creation method according to claim 8, wherein, in said detection step, if said area of interest contains a predetermined symbol, said area of interest is detected as said auxiliary information area. 11. The document creation method according to any one of claims 2 to 10, wherein in said detecting step, at least one of a position on a page and a size indicating said detected blank area is detected. 12. A condition according to any one of claims 2 to 11, wherein in said detecting step, when detecting said blank area, there is no area different from said blank area below said blank area on a page. or the document creation method according to item 1. 2. The document creation method according to claim 1, wherein, in said detecting step, a page including a predetermined character string at the bottom of the page as said predetermined condition is detected as a sign of said boundary page. 14. The document creation method according to any one of claims 1 to 13, wherein in said acquisition step, image data read from a document by a scanner is acquired. 14. The document creation method according to any one of claims 1 to 13, wherein in said acquisition step, image data read from a document is acquired from an external device. A document production device,
Acquisition means for acquiring a document as image data on a page-by-page basis;
detection means for detecting a sign of a boundary page that satisfies a predetermined condition for each page from the acquired image data;
identifying means for identifying one or more boundary pages serving as boundaries of a document based on the page containing the sign of the boundary page detected by the detecting means and information on adjacent pages adjacent to the page;
and a creating unit that creates document data divided at the positions of the one or more boundary pages specified by the specifying unit.

Images