JP2009141525A - Apparatus and method of processing image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To embed information in a text region while being capable of extracting the embedded information more accurately. <P>SOLUTION: An apparatus 110 for processing an image embeds watermark information 105 into a document image 100. The apparatus 110 includes: an image input section 101 to input the document image 100; a division section 102 for dividing the input document image 100 into a text region composed of characters and a non-text region excepting the text region; a circumscribing-rectangle extracting section 103 for extracting a circumscribing rectangle circumscribed with characters contained in the text region; a region-information embedding section 104 for embedding mutually different region information in at least adjacent text regions as information indicating a text region to which each character belongs into each character according to the circumscribing rectangle extracted by the circumscribing-rectangle extracting section 103; and a watermark-information embedding section 106 for embedding the watermark information 105 in each text region embedding the region information by the region-information embedding section 104 according to an embedding argorithm different from that of the region-information embedding section 104. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a technique for embedding information in a document image and extracting the embedded information.

There is a technique for dividing an input document image into a text region composed of characters and a non-text region composed of figures other than characters (Patent Document 1). In addition, there is a technique for embedding information in units of characters by changing the interval or size of characters included in a text area (Patent Documents 2 and 3).
US patent application Ser. No. 08 / 17,172 JP 2002-232679 A JP 2003-259112 A

  However, since the techniques described in Patent Documents 2 and 3 described above change the feature amount such as the character spacing and size in character units, the character position is shifted when information is embedded. For this reason, the division result of the document image may be different before and after the information is embedded. For example, two or more text areas are extracted as one text area, or one text area is extracted as a plurality of text areas. For this reason, there is a difference between the number of divided areas and the number of extracted areas, and thus it may be difficult to extract embedded information as embedded.

  Accordingly, an object of the present invention is to embed information in a text region while enabling the embedded information to be extracted more accurately.

  In the present invention, an image processing apparatus for embedding watermark information in a document image, the input means for inputting the document image, a text area composed of characters in the input document image, and other than the text area Information indicating which text area each character belongs to, a dividing unit that divides into non-text areas, an extraction unit that extracts a circumscribed rectangle circumscribing characters included in the text area, and at least adjacent to each other A first embedding unit that embeds different region information in each character in each character according to the circumscribed rectangle extracted by the extracting unit, and each text region in which the region information is embedded by the first embedding unit. And second embedding means for embedding the watermark information in accordance with an embedding algorithm different from that of the first embedding means. The image processing apparatus comprising: a is provided.

  The present invention is also an image processing apparatus for extracting the watermark information from a document image in which watermark information is embedded, the input means for inputting the document image, and the input document image made up of characters. Area information indicating which text area each character belongs to, for each circumscribed rectangle circumscribing a character included in the text area, and a dividing means for dividing the text area into a non-text area other than the text area When the different region information is extracted from the target text region, the target text region is re-established so that only the same region information is included in the target text region. And when the same area information is extracted from a text area adjacent to the text area of interest, the text area of interest is extracted. Correction means for correcting so as to merge the text area from which the same area information is extracted, and second extraction means for extracting the watermark information embedded for each text area, An image processing apparatus is provided.

  Further, in the present invention, there is provided an image processing method for embedding watermark information in a document image, the input step for inputting the document image, the text region composed of characters in the input document image, and the text region A division step of dividing into a non-text region other than, an extraction step of extracting a circumscribed rectangle circumscribing characters included in the text region, and information indicating which text region each character belongs to, In the adjacent text region, different region information is embedded in each character according to the circumscribed rectangle extracted in the extraction step, and each region information is embedded in the first embedding step. Second embedding the watermark information in a text area according to an embedding algorithm different from the first embedding step Image processing method characterized by comprising: a degree, is provided.

  Further, in the present invention, there is provided an image processing method for extracting the watermark information from a document image in which watermark information is embedded, the input step for inputting the document image, and the input document image being composed of characters. Area information indicating which text area each character belongs to, for each circumscribed rectangle circumscribing a character included in the text area, and a division step for dividing the text area into a non-text area other than the text area And when the different region information is extracted from the target text region, the target text region is regenerated so that only the same region information is included in the target text region. And when the same area information is extracted from a text area adjacent to the text area of interest, the text area of interest is extracted. And a correction step for correcting so as to merge the text region from which the same region information is extracted, and a second extraction step for extracting the watermark information embedded in each text region, An image processing method is provided.

  According to the present invention, it is possible to embed information in a text area while allowing the embedded information to be extracted more accurately.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

<First Embodiment>
In the present embodiment, the configuration and operation procedure of the information embedding device 110 for embedding information in the document image 100 and the information extracting device 610 for extracting information embedded in the document image 100 will be described in order. Information to be embedded in the document image 100 is hereinafter referred to as watermark information.

[Configuration of Information Embedding Device 110]
FIG. 1 is a block diagram showing a functional configuration of an information embedding device 110 according to an embodiment of the present invention. The image input unit 101 receives a document image 100 that is a target for embedding the watermark information 105. The document image 100 may be a printed matter input from a scanner or the like and converted into a bitmap, or may be electronic data created using a document editing application program. Alternatively, various electronic data including a format unique to an application program, a text format, and the like stored in a storage medium may be converted by an image processing software and converted into a bitmap.

  The dividing unit 102 divides the input document image 100 into a text area composed of characters and an area other than the text area (that is, a non-text area composed of figures or photographs other than characters). The circumscribed rectangle extracting unit 103 extracts a circumscribed rectangle circumscribing the characters included in the text area divided by the dividing unit 102.

  The circumscribed rectangle originally indicates a region where character recognition is performed. However, in the digital watermark technique, the circumscribed rectangle indicates a region to be embedded. In other words, each pixel value of the document image 100 is projected onto the vertical coordinate axis, and a blank portion (a portion without a black character) is searched to determine a row and perform line division. Thereafter, the document image 100 is projected on the horizontal coordinate axis line by line, and a blank portion is searched and divided into character units. Thereby, each character can be cut out as a circumscribed rectangle.

  The area information embedding unit 104 is information indicating which text area each character belongs to, and at least adjacent text areas are different from each other according to the circumscribed rectangle extracted by the circumscribed rectangle extracting unit 103. Embed in characters. The area information embedding unit 104 functions as a first embedding unit.

  The watermark information embedding unit 106 embeds watermark information in each text region in which the region information is embedded by the region information embedding unit 104 according to an embedding algorithm different from that of the region information embedding unit 104 to generate a watermark image 108. The watermark information embedding unit 106 functions as a second embedding unit. The image output unit 107 outputs the watermark image 108 generated by the watermark information embedding unit 106.

[Operation Procedure of Information Embedding Device 110]
2 is a flowchart showing an operation procedure of the information embedding device 110, FIG. 3 is a flowchart showing a detailed operation procedure of step S203, and FIG. 4 is a diagram showing an example of processing in the dividing unit 102. is there. 5 and 9 are diagrams illustrating an example of processing in the region information embedding unit 104. FIG.

  First, in step S <b> 200, the document image 100 to be embedded with the watermark information 105 is input to the dividing unit 102 via the image input unit 101.

  In step S201, the dividing unit 102 divides the input document image 100 into a plurality of areas. Thereby, a circumscribed rectangle of the text area and the non-text area is extracted. For this extraction method, a known region extraction technique (for example, US Patent Application No. 08/17172) is used. For example, as shown in FIG. 4, when the document image 401 is divided into a plurality of regions, an extraction result 402 is obtained. In step S202, the circumscribed rectangle extracting unit 103 extracts circumscribed rectangles of characters included in the text area. In step S203, the region information embedding unit 104 embeds region information in characters (or character images) in each circumscribed rectangle included in the text region.

  Here, the detailed operation procedure of step S203 is demonstrated using FIG. First, in step S203a, region information is assigned to the text region extracted in step S201 according to the order of raster scanning. When assigning region information, in this embodiment, among the integers 0 to 3, the numbers of regions adjacent to each other in the vertical and horizontal directions are different, and the smallest number is selected and assigned.

  For example, as shown in FIG. 5, raster scan reading is performed from the upper left to the right of the page. When the end of the line is reached, the line moves to the next line and again from the left end toward the right end. Shall be done. In this case, reading by raster scanning is performed in the order of 1-8. That is, the minimum area information 0 is assigned to the text area of order 1 in raster scanning, and the area information 0 assigned to the text area of order 1 is assigned to the text area of order 2 adjacent to the text area of order 1. Minimum area information 1 excluding is assigned. The processing proceeds as described above, and the text area of order 8 is the minimum area excluding the area information 3, 1, and 2 assigned to the text areas of order 4, 6, and 7 that are adjacent vertically and horizontally. Information 0 is assigned. Through the above processing, different area information can be assigned to adjacent areas. Although an example using four numbers from 0 to 3 is shown here, the number of integers to be allocated may be adjusted as necessary.

  Next, in step S203b, one text area is selected from the document image and set as the attention area. Thereafter, in step S203c, the area information assigned in step S203a is embedded in each character in the attention area. That is, the same area information is embedded in each character included in one text area. For example, as shown in FIG. 9, when the area information assigned in step 203a is embedded in the text area of the document image 901, a document image 902 is obtained. When embedding region information, in this embodiment, characters in each circumscribed rectangle are configured by selecting a mask pattern corresponding to the region information from among four different types of mask patterns prepared in advance. Thus, the area information is embedded. FIG. 9 schematically shows a state in which the region information is embedded by describing a number in each circumscribed rectangle. However, in actuality, the region information is obtained by displaying characters in a fine mask pattern. Will be embedded.

  In step S203d, it is determined whether or not the attention area is the last text area included in the document image. If it is determined in step S203 that the attention area is not the last text area, the process returns to step S203b to select the next text area as the attention area. On the other hand, if it is determined in step S203 that the attention area is the last text area, the processing in step S203 is terminated.

Here, it returns to description of FIG. 2 again. As described above, after embedding the area information for each text area, in step S204, similarly, one text area is selected from the document image as an attention area, and the watermark information 105 is embedded in the attention area. In embedding the watermark information 105, in the present embodiment, a technique (embedded between characters) is used in which information is embedded by shifting the circumscribed rectangle of interest in the horizontal direction and providing a size relationship between the circumscribed rectangles before and after. Use.
The order in which the watermark information 105 is embedded follows the order assigned to each area, that is, the order of raster scanning.

  In this way, in step S205, the watermark image 108 in which the watermark information 105 is embedded is generated, and in the present embodiment, the watermark image 108 is output as a paper medium from the image output unit 107 such as a printer.

  In the present embodiment, the area information is embedded by masking each character with a different mask pattern, and the watermark information 105 is embedded by changing the character spacing. However, the present invention is not limited to this embedding method. The watermark information 105 may be embedded by masking each character with a different mask pattern, and the area information may be embedded by changing the front-rear interval between circumscribed rectangles. Further, as an alternative embedding method, various methods such as an interline watermark or a background pattern watermark may be used. Here, the interline watermark is a technique for embedding information by changing the vertical spacing between lines composed of circumscribed rectangles, and the background pattern watermark is a blank area when a document is printed on paper with a laser printer or the like. This is a technique for embedding information by drawing a fine dot pattern.

  In this embodiment, the watermark image 108 is printed and output as a paper medium. However, the watermark image 108 may be stored as image data in a storage device or the like. Furthermore, it may be transmitted to another terminal or the like via a network.

[Configuration of Information Extraction Device 610]
FIG. 6 is a block diagram illustrating a functional configuration of the information extraction device 610 according to an embodiment of the present invention. The watermark image 108 generated by the information embedding device 110 is input to the image input unit 600. The dividing unit 601 divides the input watermark image 108 into a text area composed of characters and a non-text area other than the text area. The circumscribed rectangle extraction unit 602 extracts a circumscribed rectangle of characters in the extracted text area. The area information extraction unit 603 extracts area information indicating which text area each character belongs to for each circumscribed rectangle circumscribing the character included in the text area. The area information extraction unit 603 functions as a first extraction unit.

  When different region information is extracted from the target text region, the region correction unit 604 subdivides the target text region so that only the same region information is included in the target text region. Further, when the same region information is extracted from a text region adjacent to the target text region, the region correction unit 604 corrects the target text region and the text region from which the same region information is extracted to be merged. To do. The watermark information extraction unit 605 extracts and outputs watermark information 606 embedded for each text area of the watermark image 108. The watermark information extraction unit 605 functions as a second extraction unit.

[Operation Procedure of Information Extraction Device 610]
FIG. 7 is a flowchart showing an operation procedure of the information extraction apparatus 610, and FIG. 8 is a flowchart showing a detailed operation procedure of step S704. First, in step S <b> 700, the watermark image 108 is input to the dividing unit 601 via the image input unit 600. In step S <b> 701, the dividing unit 601 extracts a text area and a non-text area from the input watermark image 108.

  In step S <b> 702, the circumscribed rectangle extracting unit 602 extracts the circumscribed rectangles of the characters included in the text area extracted by the dividing unit 601. In step S703, the area information extraction unit 603 extracts area information embedded in each character.

  In step S704, the region correction unit 604 corrects the text region extraction result based on the extracted region information.

  Here, the detailed operation procedure of step S704 will be described with reference to FIG. First, in step S704a, a text region in the watermark image 108 is selected as a region of interest. Next, in step S704b, the area information embedded in each character in the attention area is extracted by identifying the mask pattern.

  In step S704c, it is determined whether or not all area information extracted from each circumscribed rectangle in the attention area is the same. If it is determined in step S704c that all area information is not the same, in step S704d, the text area is corrected so that only the circumscribed rectangle having the same area information is included in the attention area.

  Next, in step S704e, it is determined whether or not the attention area is the last text area included in the watermark image 108. If it is determined in step S704e that the attention area is not the last text area, the process returns to step S704a to select the next text area as the attention area.

  On the other hand, if it is determined in step S704e that the attention area is the last text area, in step S704f, the text area in the watermark image 108 is selected again to be the attention area. Next, in step S704g, it is determined whether the attention area and the area information embedded in the adjacent area are the same. If it is determined in step S704g that the region information of the attention region and the adjacent region is the same, the attention region and the adjacent region are merged in step S704h.

  Next, in step S704i, it is determined whether or not the attention area is the last text area included in the watermark image 108. If it is determined in step S704i that the attention area is not the last text area, the process returns to step S704f, and the next text area is selected as the attention area. If it is determined in step S704i that the attention area is the last text area, the series of processes in step S704 is terminated.

  Here, it returns to description of FIG. 7 again. In step S705, the watermark information extraction unit 605 extracts the embedded watermark information 606 by determining the size relationship between the circumscribed rectangles. Note that the watermark information 606 is extracted according to the order indicated by the region information extracted from the text region corrected by the region correction unit 604.

  As described above, according to the present embodiment, since region information different between adjacent regions is embedded as a mask pattern in characters included in each divided region, the region information is embedded without shifting the character position. Can do. On the other hand, even if the text area is not correctly divided, the division result can be corrected based on the area information embedded in each character. Therefore, it is possible to embed information in the text area while making it possible to extract the embedded information more accurately.

  In the present embodiment, the area information and the watermark information are embedded only in the text area. However, the area information and the watermark information may be embedded in the non-text area.

<Second Embodiment>
In the above-described embodiment, the region information is embedded in all characters in the text region. However, the present embodiment is different in that the region information is embedded only in characters located on the outer periphery of the text region. Note that the information embedding device 110 further includes an order information embedding unit that embeds order information indicating the order in which watermark information is read for each text region, and the information extraction device 610 includes order information embedded for each text region. Is further provided with an order information extraction unit. The order information embedding unit functions as a third embedding unit, and the order information extracting unit functions as a third extracting unit. In addition, regarding the operation procedure, only the operation procedure of step S203 is only partially changed from the first embodiment, and therefore, only the changed portion will be described in this embodiment.

  FIG. 10 is a diagram illustrating an example of the arrangement state of the region information, and FIG. 11 is a diagram illustrating a detailed operation procedure of step S203 according to the second embodiment. First, in step S203e, region information is assigned to each text region of the document image 100. Next, in step S203f, the counter I is set to 0 (I = 0). In step S203g, one text area is selected from a plurality of text areas included in the document image 100, and set as the attention area. In step S203h, the area information assigned in step S203e is embedded in characters located on the outer periphery of the text area (that is, characters constituting the circumscribed rectangle of the text area). When embedding the area information, in this embodiment, as in the first embodiment, the area information is embedded by masking characters with a mask pattern prepared in advance.

  In step S203i, information set in the counter (hereinafter referred to as counter information) is embedded in the character located inside the character in which the area information is embedded in step S203h. Here, as will be described later, the counter information is incremented every time after being embedded, and thus functions as information indicating the order in which the watermark information is extracted. That is, when the same information is embedded as in the text area 1000 (area information is 0, counter information is 0), the text area 1000 is masked with the same mask pattern. On the other hand, when different information is embedded as in the text area 1001 (area information is 0, counter information is 2), the text area 1001 is masked with a different mask pattern. When there is no circumscribed rectangle of a character that does not constitute the circumscribed rectangle of the text area within the circumscribed rectangle of the character that constitutes the circumscribed rectangle of the text area among the characters included in the attention area, such as the text area 1003 The counter information is not embedded.

  In step S203j, it is determined whether or not the attention area is the last text area included in the document image 100. If it is determined in step S203j that the attention area is not the last text area, the counter is incremented in step S203k (I = I + 1), the process returns to step S203g, the next text area is selected, and the attention area and To do. On the other hand, if it is determined in step S203j that the attention area is the last text area, the series of processing in step S203 is terminated. Subsequent processing (step S204 and step S205) is the same as that of the first embodiment.

  As described above, according to the present embodiment, watermark information can be extracted more accurately according to embedded counter information.

  In the first embodiment, the watermark information 105 is embedded in each text region in which the region information is embedded by the region information embedding unit 104 according to an embedding algorithm different from that of the region information embedding unit 104. On the other hand, in the present embodiment, the watermark information 105 is embedded in each text region in which the region information is embedded by the region information embedding unit 104 according to an embedding algorithm different from that of the region information embedding unit 104 and the order information embedding unit (not shown). It will be.

<Third Embodiment>
In each of the above-described embodiments, each unit configuring the information embedding device 110 illustrated in FIG. 1 and the information extracting device 610 illustrated in FIG. 6 has been described as configured by hardware. However, some or all of the units constituting the information embedding device 110 and the information extraction device 610 may be executed by software, and the software may be executed by a computer to realize the functions of the units.

  FIG. 12 is a diagram illustrating a system configuration of the computer 1201. A computer 1201 is a general-purpose information processing apparatus such as a personal computer that is widely used. The computer 1201 can input an image read from an image input device 1217 such as a scanner or a copying machine, and can edit or store the image. .

  Further, an image obtained by the image input device 1217 may be printed from a printer 1216 or a copying machine. Various instructions from the user are performed by an input operation from a mouse (registered trademark) 1213 or a keyboard 1214. Inside the computer 1201, areas described later are connected by a bus 1207 so that various data can be exchanged. Note that depending on the apparatus to which the computer 1201 is applied, all of the components shown in FIG. 12 are not necessarily required, and may be omitted as appropriate.

  The CPU 1202 controls the entire computer 1201 using programs and data loaded in the main storage device 1203 and executes the above-described processes performed by the information embedding device 110 and the information extraction device 610 to which the computer 1201 is applied. .

  An example of the main storage device 1203 is a RAM. The main storage device 1203 has an area for temporarily storing data loaded from an HDD (hard disk drive) 1204, a CD-ROM drive 1209, a DVD-ROM drive 1210, and an FDD (floppy (registered trademark) disk drive) 1211. . The main storage device 1203 further has an area for temporarily storing image data acquired from the image input device 1217 via the I / F (interface) 1215. The main storage device 1203 further has a work area used when the CPU 1202 executes various processes. That is, the main storage device 1203 can provide various areas as appropriate.

  The HDD 1204 can store an OS (operating system) and various images (including document images). Furthermore, the HDD 1204 can also store programs and data that cause the CPU 1202 to execute the above-described processes performed by the information embedding device 110 and the information extraction device 610 to which the computer 1201 is applied. Programs and data stored in the HDD 1204 are appropriately loaded into the main storage device 1203 under the control of the CPU 1202 and are processed by the CPU 1202. Note that some of the information described as being stored in the main storage device 1203 may be stored in the HDD 1204.

  The video controller 1205 transmits display data such as image data and character data received from the main storage device 1203 and the HDD 1204 as a signal to the monitor 1206. The monitor 1206 includes a CRT, a liquid crystal screen, and the like, and displays images, characters, and the like based on signals received from the video controller 1205.

  The I / F 1208 connects the printer 1216 to the computer 1201. The computer 1201 transmits print data to the printer 1216 via the I / F 1208 and receives status information of the printer 1216 transmitted from the printer 1216.

  The CD-ROM drive 1209 reads programs and data stored in a CD-ROM as a storage medium, and transmits the read programs and data to the HDD 1204 and the main storage device 1203.

  The DVD-ROM drive 1210 reads a program and data stored in a DVD-ROM as a storage medium, and transmits the read program and data to the HDD 1204 and the main storage device 1203.

  The FDD 1211 reads programs and data stored in a floppy (registered trademark) disk as a storage medium, and sends the read programs and data to the HDD 1204 and the main storage device 1203.

  A mouse (registered trademark) 1213 and a keyboard 1214 function as devices for a user to input various operation instructions. A user of the computer 1201 can input various instructions to the CPU 1202 by operating a mouse (registered trademark) 1213 or a keyboard 1214. For example, when the user inputs a selection instruction, the mouse (registered trademark) 1213 or the keyboard 1214 is used.

  The I / F 1212 connects a keyboard 1214 and a mouse (registered trademark) 1213 to the bus. An operation instruction input from a mouse (registered trademark) 1213 or a keyboard 1214 is sent as a signal to the CPU 1202 via the I / F 1212.

  The I / F 1215 is for connecting an image input device 1217 that generates image data by reading a document, a film, or the like to the computer 1201. Image data generated by the image input device 1217 is sent to the HDD 1204 and the main storage device 1203 via the I / F 1215.

<Other embodiments>
Note that the present embodiment can be applied to a system including a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), and a device (for example, a copier, a multifunction peripheral, The present invention may be applied to a facsimile machine or the like.

  Further, the present invention may supply a computer-readable storage medium (or recording medium) that stores a computer program code of software that implements the functions of the above-described embodiments to a system or apparatus. Further, the present invention may be applied to the computer (or CPU or MPU) of the system or apparatus reading and executing the program code stored in the storage medium. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiment, and the storage medium recording the program code constitutes the present embodiment. In addition, an operating system (OS) or the like running on the computer performs part or all of the actual processing based on the instruction of the program code, and the functions of the above-described embodiments may be realized by the processing. Needless to say, it is included.

  Further, the program code read from the storage medium is written in a memory provided in a function expansion card inserted into the computer or a function expansion unit connected to the computer. Then, based on the instruction of the program code, the CPU or the like provided in the function expansion card or function expansion unit performs part or all of the actual processing, and the functions of the above-described embodiments are realized by the processing. It goes without saying that it is included in the invention.

  When the present embodiment is applied to the above-described computer-readable storage medium, the storage medium stores computer program codes corresponding to the above-described flowcharts and functional configurations.

It is a block diagram which shows the functional structure of the information embedding apparatus 110 which concerns on one Embodiment of this invention. 5 is a flowchart showing an operation procedure of the information embedding device 110. It is a flowchart which shows the detailed operation | movement procedure of step S203. 6 is a diagram illustrating an example of processing in a dividing unit 102. FIG. 6 is a diagram illustrating an example of processing in a region information embedding unit 104. FIG. It is a block diagram which shows the functional structure of the information extraction apparatus 610 which concerns on one Embodiment of this invention. 5 is a flowchart illustrating an operation procedure of the information extraction device 610. It is a flowchart which shows the detailed operation | movement procedure of step S704. 6 is a diagram illustrating an example of processing in a region information embedding unit 104. FIG. It is a figure which shows an example of the arrangement | positioning state of area | region information. It is a figure which shows the detailed operation | movement procedure of step S203 which concerns on 2nd Embodiment. 2 is a diagram illustrating a system configuration of a computer 1201. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Document image 101 Image input part 102 Dividing part 103 circumscribed rectangle extraction part 104 Area information embedding part 105 Water mark information 106 Water mark information embedding part

Claims (9)

An image processing apparatus for embedding watermark information in a document image,
Input means for inputting the document image;
A dividing unit that divides the input document image into a text area composed of characters and a non-text area other than the text area;
Extracting means for extracting a circumscribed rectangle circumscribing characters included in the text area;
A first embedding of information indicating which text area each character belongs to, and embedding different area information in at least the adjacent text area in each character according to the circumscribed rectangle extracted by the extracting means Means,
A second embedding unit that embeds the watermark information in each text region in which the region information is embedded by the first embedding unit, according to an embedding algorithm different from the first embedding unit;
An image processing apparatus comprising: An image processing apparatus for extracting the watermark information from a document image in which watermark information is embedded,
Input means for inputting the document image;
A dividing unit that divides the input document image into a text area composed of characters and a non-text area other than the text area;
First extraction means for extracting region information indicating which text region each character belongs to for each circumscribed rectangle circumscribing the character included in the text region;
When different region information is extracted from the target text region, the target text region is subdivided so that only the same region information is included in the target text region, and the target text region Correction means for correcting so as to merge the text area from which the same area information is extracted with the text area of interest when the same area information is extracted from adjacent text areas;
Second extraction means for extracting the watermark information embedded for each text region;
An image processing apparatus comprising:   The first embedding means embeds the area information in the text area by masking each character included in each text area with a mask pattern corresponding to the area information, and before and after the circumscribed rectangle. The image processing apparatus according to claim 1, wherein the watermark information is embedded in the text area by changing an interval or an upper and lower interval between lines constituted by the circumscribed rectangle. A third embedding unit for embedding order information indicating an order of extracting the watermark information embedded for each of the text areas;
The first embedding means embeds the region information in characters located on the outer periphery in the text region,
The image processing apparatus according to claim 1, wherein the third embedding unit embeds the order information in characters other than characters located on an outer periphery in the text area. A third extracting means for extracting order information indicating an order of extracting the watermark information embedded in each text region;
The image processing apparatus according to claim 2, wherein the third extraction unit extracts the watermark information embedded for each of the text regions in accordance with the extracted order information. An image processing method for embedding watermark information in a document image,
An input step of inputting the document image;
A division step of dividing the input document image into a text area composed of characters and a non-text area other than the text area;
An extraction step of extracting a circumscribed rectangle circumscribing the characters included in the text region;
A first embedding of information indicating which text area each character belongs to, and embedding different area information in at least the adjacent text areas in each character according to the circumscribed rectangle extracted in the extraction step Process,
A second embedding step of embedding the watermark information in each text region in which the region information is embedded in the first embedding step, according to an embedding algorithm different from the first embedding step;
An image processing method comprising: An image processing method for extracting the watermark information from a document image in which watermark information is embedded,
An input step of inputting the document image;
A division step of dividing the input document image into a text area composed of characters and a non-text area other than the text area;
A first extraction step of extracting region information indicating which text region each character belongs to for each circumscribed rectangle circumscribing a character included in the text region;
When different region information is extracted from the target text region, the target text region is subdivided so that only the same region information is included in the target text region, and the target text region When the same area information is extracted from adjacent text areas, a correction step for correcting the merged text area from which the same text information is extracted as the target text area,
A second extraction step of extracting the watermark information embedded for each text region;
An image processing method comprising:   The computer program which makes a computer perform each process of Claim 6 or 7 when a computer reads and executes.   A computer-readable storage medium storing the computer program according to claim 8.

Images