WO2014050480A1

WO2014050480A1 - Document image processing device, method for controlling operation thereof, and program for controlling operation thereof

Info

Publication number: WO2014050480A1
Application number: PCT/JP2013/073885
Authority: WO
Inventors: 浩教矢野
Original assignee: 富士フイルム株式会社
Priority date: 2012-09-26
Filing date: 2013-09-05
Publication date: 2014-04-03

Abstract

The present invention performs Japanese line breaking processing in a document image. A character image representing a character is extracted from the document image (step 31). If the extracted character image contains a character prohibited at the beginning of a line, same is joined to the character image immediately preceding the character prohibited at the beginning of a line, generating a joined character image (step 33). If the extracted character image contains a character prohibited at the end of a line, same is joined to the character image immediately following the character prohibited at the end of a line, generating a joined character image (step 34). The generated joined character images and extracted character images are positioned and displayed at a desired display region in accordance with the arrangement in the document image (steps 35 and 36).

Description

Document image processing apparatus, operation control method thereof, and operation control program thereof

The present invention relates to a document image processing apparatus, its operation control method, and its operation control program.

When a document image or fixed layout document file is viewed on a mobile terminal, the paragraph size in the document is larger than the display screen size of the mobile terminal, so the scrolling of the paragraph area is required for continued browsing. For this reason, when browsing a document on a portable terminal, it is necessary to be aware of the browsing action and the terminal operation action alternately, and it becomes impossible to perform comfortable document browsing obtained by continuing only the browsing action. In order to solve such a problem, “development of document image layout reconstruction technology“ GT-Layout ”for portable terminals” (Non-Patent Document 1) is known. This GT-Layout makes it possible to view documents by scrolling in one direction by rearranging the character positions according to the display screen from the document image and character position information, and configuring the document image to match the display screen size. To do.

The display order of the character images is determined, and the character images are displayed on the display screen according to the order. When displaying normal text represented by text data that is not a document image on the screen, it is necessary to perform a prohibition process to adjust the balance between characters so that the punctuation in the text does not come to the beginning of the line. .

In addition, when a punctuation mark follows, a line break is not allowed (Patent Document 1), a punctuation mark cannot be positioned at the beginning of a line by prohibition processing (Patent Document 2), and a prohibition processing code is assigned to a character image What is to be performed (Patent Document 3) and not a document image, but in the case of forbidden characters, there are also those that are synthesized with the previous character (Patent Document 4).

Japanese Unexamined Patent Publication No. 5-266168 JP 2004-280418 A JP 2005-267129 A JP-A-6-236372

However, even when prohibition processing is performed, in the case of a document image, since characters are represented as images, the same prohibition processing as in the case of a document represented by text data cannot be performed. For example, in Patent Document 4, in the case of a prohibited character, a part of the character area one character before is cut out, and the prohibited character is combined with the cut out part, but if a part of the character image is cut off, There is also a possibility that the image portion representing is cut off.

This invention is intended to allow forbidden processing on document images.

A document image processing apparatus according to the present invention includes a character image cutout unit that cuts out a character image representing a character included in an image from a document image obtained by converting the document into an image, and a character that is represented by the character image cut out by the image cutout unit. A prohibition character judging means for determining whether or not a forbidden character or an end-of-line prohibition character is included, and a forbidden character according to whether the prohibition character judging means judges that a forbidden character image representing a forbidden character is included. When a character image is combined with a character image immediately before the forbidden character image to generate a combined character image, and the forbidden character judging means determines that the forbidden character image representing the forbidden character is included. A combined character image generator that generates a combined character image by combining a non-end-of-line character image with the character image immediately after the non-end-of-line character image. Characterized in that it comprises a.

The present invention also provides an operation control method suitable for a document image processing apparatus. That is, in this method, the character image cutout means cuts out a character image representing the character included in the image from the document image obtained by imaging the document, and the forbidden character determination means uses the character image cut out by the image cutout means. It is determined whether the displayed character includes a prohibited character or a prohibited character, and the combined character image generation means determines that the prohibited character determination means includes a prohibited character image representing the prohibited character. In response to this, the forbidden character image is combined with the character image immediately before the forbidden character image to generate a combined character image, and the forbidden character judging means includes a forbidden character image representing the forbidden character at the end of the line. In accordance with the determination, the line end prohibited character image is combined with the character image immediately after the line end prohibited character image to generate a combined character image. .

According to the present invention, a character image is cut out from a document image, and the characters represented by the cut-out character image are prohibited characters (characters that are not suitable as characters appearing at the beginning of a row or column) or prohibited characters (end of a row or column). It is determined whether or not a character that is not suitable as the last appearing character is included. If a forbidden character is included, the forbidden character image is combined with the immediately preceding character image to generate a combined character image. If a line-end prohibited character image is included, the line-end prohibited character image is combined with the character image immediately after that to generate a combined character image. Since there is no line-breaking character image or line-breaking character image alone, it is possible to prevent the line-beginning or line-ending character image from becoming a line-breaking-character image or line-breaking-character image.

Positioning means for positioning the character image clipped by the character image cutout means and the combined character image generated by the combined character image generation means in the display area of the display screen according to the character arrangement in the document image, the last character image in the row or column Is a combined character image, and storage determination means for determining whether the combined character image does not fit in the display area, and the storage determination means determines that the combined character image does not fit in the display area. You may make it further provide the reduction means to reduce all the character images of the row | line | column or column in which a combined character image is contained.

The reduction means includes, for example, reduction determination means for determining whether or not the character image in the row or column including the combined character image is fit in the display area of the combined character image at a predetermined reduction ratio. Also good. In response to determining that the combined character image fits in the display area by the reduction determination unit, all the character images in the row or column including the combined character image are reduced. In this case, the positioning unit determines whether the combined character image is not within the display area by the reduction determination unit, so that the combined character image is positioned at the head of the next row or column next to the row or column in which the combined character image was last positioned. Preferably, the combined character image is positioned at the position.

It is a block diagram which shows the electrical structure of a document image processing apparatus. It is an example of a document image. It is a flowchart which shows the process sequence of a document image processing apparatus. It is an example of a document image. It shows how a combined character image is generated. It is an example of a character information table. It is an example of a character information table. It is an example of the character image positioned in the display area. It is an example of a document image. It shows how a combined character image is generated. It is a flowchart which shows a character image positioning process procedure. It is an example of the character image positioned in the display area. It is an example of the character image positioned in the display area.

FIG. 1 shows an embodiment of the present invention, in which an electric image of a document image processing apparatus 1 for shaping a document image (an imaged document will be referred to as a document image) so that it can be displayed in a desired display area. It is a block diagram which shows a typical structure.

The overall operation of the document image processing apparatus 1 is controlled by the control apparatus 2.

The document image processing apparatus 1 includes an input device 3 such as a keyboard for inputting various commands, a communication device 4 for communicating with other client terminal devices, mobile phones, etc., a display device 5 for displaying a document image, etc. A memory 6 for storing the data is provided. The document image processing apparatus 1 is provided with a CD (compact disk) driver 7. When a compact disk 8 storing a program for controlling operations to be described later is loaded into the CD driver 7 and the program stored in the compact disk 8 is read by the CD driver 7, the program is processed by the document image processing apparatus. 1 installed. However, the communication device 4 may be used to receive a program, and the received program may be installed in the document image processing device 1.

Furthermore, the document image processing apparatus 1 includes a character area acquisition device 11, a prohibited character extraction device 12, an area synthesis device 13, and a shaped image creation device 14. The character area acquisition device 11 detects and extracts a character image area from a document image. Extraction of character images can use the function of OCR (Optical Character Reader). The coordinate position of the character image in the document image, the character type represented by the character image, the order of the characters, and whether the character is written horizontally or vertically are also detected. The prohibited character extraction device 12 extracts a character image when the character represented by the character image is a prohibited character. Since the character type detection in the character area acquisition device 11 is not necessarily accurate, a forbidden character is extracted with reference to the relative positions of the character image immediately before and after the character image. be able to. For example, a quotation mark called double quotation can be determined by whether the image is a square shape in the upper 10 percent of the character represented by the character image immediately preceding the quotation mark. Of course, it is also possible to extract a character image representing a prohibited character using pattern matching. The area synthesizer 13 combines a character image representing a prohibited character with a character image immediately before or after it to generate a combined character image. When the forbidden character image represents the end-of-line prohibited character, it is combined with the immediately following character image, and when the forbidden character image represents the forbidden character, it is combined with the immediately preceding character image. The shaped image creation device 14 positions the character image obtained by the character region acquisition device 11 and the combined character image obtained by the region synthesis device 13 so that it can be displayed on a display screen having a desired display region. is there. Details of these processes will be described later.

FIG. 2 is an example of an imaged document image 20.

The document image 20 includes characters (INVENTION!) Represented by the image. These characters are not represented by text data, but are represented by images. In this embodiment, the document image 20 is shaped.

FIG. 3 is a flowchart showing the processing procedure of the document image processing apparatus 1.

It is assumed that document image data representing the document image 20 is stored in the memory 6. As shown in FIG. 2, processing such as extraction of a character image from the document image 20 is performed (step 31).

FIG. 4 shows a state in which character images 21-30 are extracted from the document image 20. The extraction of the character image 21-30 uses the OCR function as described above. The extracted character image 21-30 is surrounded by a rectangle. When the upper left vertex of the document image 20 is the origin (X0, Y0), the upper left coordinates of these rectangles are the coordinate positions of the character images 21-30. For example, the positions of the

character images

21, 22, and 23 are represented by coordinates (x1, y1), (x2, y2), and (x3, y3). Similarly, the position of the character image 30 is represented by coordinates (x10, y10). Further, the width and height of the character image 21-30 are also detected. The coordinates and the like of the detected character image 21-30 are stored in the character information table.

FIG. 6 is an example of a character information table.

The character information table shown in FIG. 6 is for the document image 20.

The character information table stores, for each ID for identifying a detected character image, the X coordinate, Y coordinate, width, height, and character type represented by the character image. . ID1 to ID10 stored in the character information table correspond to character images 21 to 30, respectively. For example, the ID of the character image 21 is ID1, the X coordinate is x1, the Y coordinate is y1, the width is 0.5w, the height is h, and the character type is “I”. The ID of the character image 30 is ID10, the X coordinate is x10, the Y coordinate is y10, the width is 0.5w, the height is h, and the character type is "!".

Referring back to FIG. 3, it is confirmed whether or not a character image representing a forbidden character or a forbidden character is detected from the extracted character image (step 32). The beginning of a line prohibition character or the end of a line prohibition character is predetermined. For example, exclamation marks, question marks, commas, periods, end parentheses, etc. are forbidden characters, and opening parentheses are forbidden characters.

When a bullet-inhibited character image representing a bullet-inhibited character image is detected, a character image immediately before the bullet-inhibited character image and the bullet-inhibited character image are attached to generate a combined character image (step 33). ).

FIG. 5 shows how a combined character image is generated.

The character image 30 detected in the document image 20 shown in FIG. 4 represents a question mark, and is a forbidden character image 30. For this purpose, the character image 29 immediately before the forbidden character image 30 and the forbidden character image 30 are attached to generate one combined character image 30A. When the combined character image 30A is generated, the character information table described above is also corrected.

FIG. 7 shows an example of the corrected character information table.

As the ID of the combined character image 30A generated as described above, ID9 which is the ID of the character image 29 before combining is used. Since the combined character image 30A has

character images

29 and 30 attached thereto, the width is changed from w to 1.5w, and the character type is changed from “N” to “N!”. The X coordinate, Y coordinate, and height are not changed.

Returning to FIG. 3, when a line-end prohibited character image representing a line-end prohibited character image is detected, the character image immediately after the line-end prohibited character image and the line end-prohibited character image are combined to form a combined character image. Is generated (step 34). The process of combining line-end prohibited character images will be described later (see FIGS. 9 and 10).

If neither the forbidden character image nor the forbidden character image is detected, the processing in

step

33 or 34 is skipped.

When a prohibited character image or a prohibited character image is detected and a combined character image is generated, or when neither a prohibited character image nor a prohibited character image is detected, the detected character image or the like is displayed. Positioning is performed in the display area of the display screen (step 35). As a result, processing for creating a shaped image is performed.

FIG. 8 shows a state in which the character image is positioned in the display area 50 corresponding to the desired display screen.

The width of the display area 50 is narrower than the width of the document image 20. For this reason, in the document image 20, all of the character images 21 to 30 (30A) are displayed in one line, but all of the character images 21 to 30 (30A) are displayed in one line of the display area 50. I can't do it.

In the first line of the display area 50, the character images 21 to 25 are positioned, and in the second line of the display area 50, the character images 26 to 30A are positioned. Needless to say, the positioning of the character images 21 to 30A is performed using the character information table shown in FIG. For example, in the example shown in FIG. 8, when the character images 21 to 26 are stored in the first line, the character image 26 protrudes from the display area 50, so that the character image 26 is placed at the head of the second line. It is positioned.

The character image thus positioned in the display area 50 is displayed on the display screen 6 of the display device 5.

FIG. 9 shows another example of the document image 40.

The document image 40 includes character images 41 to 49. Although the character images 41 to 46 and 48 are not prohibited character images, the character image 47 is a forbidden character image 47 indicating the beginning of parentheses, and the character image 49 is a prohibited character image 49 indicating the end of parentheses.

FIG. 10 shows a state where a combined character image is generated from the end-of-line prohibited character image 47 and the end-of-line prohibited character image 49.

As described above, in the case of the line end prohibited character image 47, it is combined with the character image 48 immediately after that. Further, in the case of the forbidden character image 49, it is combined with the character image 48 immediately preceding it. As a result, a combined character image 49A is obtained.

11 to 13 show other embodiments.

In this embodiment, as described above, when a combined character image is generated by combining a forbidden character image and a character image, the combined character image is positioned at the end of the line so that the combined character image does not fit in the display area. When it runs out, it reduces all the character images in the line to fit.

FIG. 11 is a flowchart showing a character image positioning process procedure. The processing procedure of step 35 of FIG. 3 is shown. 12 and 13 show a state in which the character images 61 to 67, 71 to 77, 81 to 86, and the combined character image 87 are positioned in the display area 50. FIG.

First, the number parameter n and the line parameter m are each reset to 1 (step 41). When the nth character image is read out of the character images extracted as described above (NO in step 42 and step 43), the nth character image is positioned in the display area of the mth row. (Step 44). For example, if it is the first character image, the character image is positioned at the first position on the first line, and the character image 61 is positioned as shown in FIG. If the nth character image fits in the display area 50 (NO in step 45), the number parameter n is incremented (step 46), and the processes of

steps

42 and 44 are repeated again. As a result, the character images are sequentially positioned in the m-th line according to the character arrangement of the document image.

If the nth character image does not fit in the display area (YES in step 45), it is determined whether or not the nth character image is a combined character image (step 47). If it is not a combined character image (NO in step 47), the line parameter m is incremented (step 49), and the nth character image is positioned in the display area 50 of the mth row (step 44). For example, as shown in FIG. 12, the character images 61 to 67 are positioned on the first line, and the next read character image 71 does not enter the display area 50 when attempting to position on the first line. Since the character image 71 is not a combined character image, it is positioned at the beginning of the second line.

If the nth character image that does not fit in the display area 50 is a combined character image (YES in step 47), all the character images in the mth line that do not fit in the display area 50 are reduced by a predetermined reduction ratio (for example, reduced by 90%). It is confirmed whether or not it can fit in the display area 50 when it is reduced at (rate) (step 48). If it does not fit (NO in step 48), the line parameter m is incremented (step 49) and the combined character image is positioned at the beginning of the next line (step 44). When it is within the range (YES in step 48), all the character images in the m-th line including the combined character image are reduced (step 50). For example, as shown in the third line of FIG. 12, the combined character image 87 is positioned at the end of the third line, and the combined character image 87 protrudes from the display area 50, and the character images 81 to 86 in the third line. When it is determined that all of the combined character images 87 are reduced within the display area 50 by being reduced at a predetermined reduction rate, as shown in the third row of FIG. 86 and the combined character image 87 are reduced. The character images 81 to 86 and the combined character image 87 in the third line will fit in the display area 50. If all of the character images 81 to 86 and the combined character image 87 in the third line are reduced within the display area 50 even if they are reduced at a predetermined reduction ratio, the combined character image 87 is the first in the fourth line. Is positioned.

In the embodiment described above, in the document image processing apparatus 1, processing for extracting a character image from the document image, processing for determining whether the extracted character image is a prohibited character, processing for generating a combined character image, processing for creating a shaped image In addition, display processing on the display device 5 is performed. Data representing the created shaped image is transmitted from the document image processing device 1 to another terminal device such as a mobile phone, and the display processing is performed in the terminal device. It may be performed. Further, the shaping image creation process may be performed in another terminal device. Furthermore, the processing in the document image processing apparatus 1 may be executed by software using a server instead of a dedicated apparatus, or may be executed by a mobile phone such as a smartphone.

Furthermore, in the above-described embodiment, the horizontally written document image has been described. However, the embodiment can be similarly applied to a vertically written document image instead of horizontally written. In the case of vertical writing, it may be read as a column instead of a row.

1 Document Image Processing Device 2 Control Device
11 Character area acquisition device
12 Forbidden character extraction device
13 area synthesizer
14 Shaped image creation device
20, 40 Document image

Claims

A character image cutout means for cutting out a character image representing a character included in an image from a document image in which the document is imaged;
A forbidden character determining means for determining whether a character represented by the character image cut out by the image cutout means includes a forbidden character or an end-of-line prohibited character, and a forbidden character representing a forbidden character by the forbidden character determining means. In response to the determination that the character image is included, the forbidden character image is combined with the character image immediately before the forbidden character image to generate a combined character image. A combined character image generating means for generating a combined character image by combining the line end prohibited character image with the character image immediately after the line end prohibited character image in response to the determination that the line end prohibited character image is represented;
A document image processing apparatus comprising:
Positioning means for positioning the character image cut out by the character image cut-out means and the combined character image generated by the combined character image generating means in the display area of the display screen according to the character arrangement in the document image;
If the last character image in the row or column is a combined character image, the storage determining means for determining whether the combined character image does not fit in the display area, and if the combined character image does not fit in the display area by the storage determining means A reduction means for reducing all character images in the row or column including the combined character image according to the determination,
The document image processing apparatus according to claim 1, further comprising:
The reduction means is
A reduction determination means for determining whether or not the character image in the row or column including the combined character image is reduced by a predetermined reduction ratio to determine whether or not the combined character image is displayed in the display area;
In response to determining that the combined character image fits in the display area by the reduction determination unit, all the character images in the row or column including the combined character image are reduced.
The positioning means is
In response to determining that the combined character image does not fit in the display area by the reduction determination means, the combined character image is displayed at the beginning of the next row or column of the row or column where the combined character image was last positioned. Positioning
The document image processing apparatus according to claim 1.
A character image cutout means cuts out a character image representing a character included in the image from the document image obtained by converting the document into an image.
A forbidden character judging means judges whether or not the character represented by the character image cut out by the image cutting out means includes a forbidden character or a forbidden character at the end of the line;
The combined character image generating means converts the forbidden character image to the character image immediately before the forbidden character image in response to the prohibition character determining means determining that the forbidden character image representing the forbidden character is included. A combined character image is generated by combining the line-end prohibited character image immediately after the line-end prohibited character image in response to determining that the line-end prohibited character image representing the line end prohibited character is included by the prohibited character determining means. Combine with a character image to generate a combined character image,
An operation control method for a document image processing apparatus.
A computer-readable program for controlling a computer of a document image processing apparatus,
A character image representing characters included in the image is cut out from the document image in which the document is imaged.
Determine whether the character represented by the cut-out character image contains a forbidden character or an end-of-line character,
In response to the determination that a forbidden character image representing a forbidden character is included, the forbidden character image is combined with the character image immediately before the forbidden character image to generate a combined character image;
A document image so as to generate a combined character image by combining the forbidden character image with the character image immediately after the forbidden character image when it is determined that the forbidden character image representing the forbidden character is included. A program for controlling a computer of a processing device.