JP2003189085A - Electronic watermarking apparatus and method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a means for embedding/reading an invisible electronic watermark with high precision in Japanese document image. <P>SOLUTION: This electronic watermarking apparatus comprises a document inputting means, a region extracting means, a character recognizing means, a character extracting means, a document access control signal embedding means, a document outputting means, a document access control signal reading means, and a document control means. The document image is characterized by containing characters, tables and graphs. The document access control signal is characterized by being a signal for retrieving and managing the document. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital watermark device and a digital watermark method, and more particularly to an invisible digital watermark embedding and reading method for Japanese document images.

[0002]

2. Description of the Related Art In recent years, in a digital image forming apparatus such as a printer or a copying machine, the image quality has been remarkably improved, and it has become possible to easily obtain a printed matter of high image quality. In other words, anyone can obtain the required printed matter through image processing by a high-performance scanner, printer, copier, and computer. Therefore, problems such as illegal copying and falsification of documents have occurred, and in order to prevent or prevent them, the movement to embed the access control information in the printed matter itself has become active in recent years (digital watermark).

As such a function, the access control information is invisiblely embedded in the printed matter (invisible type), the bitmap pattern corresponding to the access control information is embedded in the margin of the document, and the document image is scrambled. Currently, encryption is common. Among them, the one that embeds the access control information invisibly is a general implementation method that embeds information by controlling the amount of space in the English character string, a type that rotates the character, and a scaling character. There have been proposed types such as a type that embeds information and a type that embeds information by transforming characters.

FIG. 8 shows a printed matter of a type that controls the amount of space in an English character string. Here, 801 is called a space. If the watermark bit to be embedded is 0, this space is p ← (1 + p) (p + s) / 2s
← (1-p) (p + s) / 2, and if the watermark bit is 1, p ← (1-p) (p + s) / 2s ← (1 + p)
(P + s) / 2.

FIG. 9 shows a printed matter in which characters are rotated. Here, (a) shows the state before rotation, and (b) shows the state after rotation. Reference numeral 901 indicates the rotation angle of the character.

FIG. 10 is an example of a type of printing in which characters are enlarged or reduced. 1001 indicates the original size. 1002
Indicates the size after scaling.

[0007]

However, in the case of invisiblely embedding the above-mentioned access control information, especially in the case of a text image, information is included in the text itself (usually a binary image) with little image redundancy. Since the characters are embedded, the characters and spaces feel uncomfortable, and the deterioration of the original quality is easily noticeable. Also, such images generally have poor read endurance.

The present invention has been made paying attention to the above points, and while suppressing the deterioration of the font, the accuracy and the amount of information embedding are maintained above a certain level, and the noise resistance is high even at the time of reading. An object of the present invention is to provide a digital watermarking device and a digital watermarking method that can realize the digital watermarking device.

[0009]

In order to solve the above problems, a digital watermarking device of the present invention has the following configuration. That is, a digital watermark device that embeds a digital watermark in a document and reads the document, and a document image input unit that inputs the document and converts the document into an electronic image, a region extraction unit that extracts a specific region of the input document image, and the extraction. Character recognition means for recognizing characters in a region, character extraction means for extracting characters from the character recognition result according to a predetermined rule, control signal embedding means for embedding a predetermined control signal in the extracted characters, and control signal embedded document Document image output means for outputting an image, control signal reading means for reading a predetermined control signal from the extracted character,
It has a document control means for controlling the document based on the read control signal.

Further, preferably, the document image in the digital watermark apparatus includes characters, tables and graphs.

Further, preferably, the predetermined control signal in the digital watermarking device is a signal for preventing or suppressing unauthorized use due to illegal reading of a document image, a signal for restricting access, and tampering prevention. And a signal for searching and managing documents. Characterized by including characters, tables and graphs.

Further, preferably, the electronic watermarking device is characterized in that extraction of a control information embedding target character is determined by its constituent element and appearance frequency.

Further, preferably, the digital watermarking device is characterized in that control information embedding and reading are carried out by using a relative relationship between constituent elements of a character.

According to the present invention, by skillfully utilizing the relative shape change of each radical of a complicated Chinese character having a radical structure, the deterioration of the font can be minimized while the information embedding accuracy is higher than a certain level. It is possible to secure a sufficient amount and realize a digital watermarking device having high noise resistance even at the time of reading. In addition, since there is no dependency on the font size in principle, it is an effective method even for a document with a small number of characters.

The present invention can be summarized and organized as follows when summarized and summarized.

(1) In a digital watermark device that embeds a digital watermark in a document and reads the document, a document image input means for inputting the document and converting it into an electronic image, a region extracting means for extracting a specific region of the input document image, and Character recognition means for recognizing characters in the extraction region, character extraction means for extracting characters from the character recognition result according to a predetermined rule, control signal embedding means for embedding a predetermined control signal in the extracted characters, and control signal embedding A document image output means for outputting a document image,
An electronic watermarking apparatus comprising: a control signal reading unit that reads a predetermined control signal from the extracted character; and a document control unit that controls a document based on the read control signal.

(2) The digital watermark apparatus according to (1), wherein the document image includes characters, tables, and graphs.

(3) The predetermined control signal is a signal for preventing or suppressing unauthorized use due to illegal reading of a document image, a signal for restricting access, and a signal for preventing falsification. The electronic watermark device according to (1) above, which is a signal for searching and managing a document.

(4) The digital watermarking device according to (1), wherein the extraction of the control information embedding target character is determined based on its constituent elements and appearance frequency.

(5) The digital watermarking device according to (1), wherein the control information embedding and the reading are performed by using the relative relationship between the constituent elements of the character.

(6) In a digital watermark device that embeds and reads a digital watermark in a document, a document image input means for inputting the document and converting it into an electronic image, a region extracting means for extracting a specific region of the input document image, and Character recognition means for recognizing characters in the extraction region, character extraction means for extracting characters from the character recognition result according to a predetermined rule, control signal embedding means for embedding a predetermined control signal in the extracted characters, and control signal embedding A document image output means for outputting a document image,
An electronic watermarking method comprising: a control signal reading unit for reading a predetermined control signal from the extracted character; and a document control unit for controlling a document based on the read control signal.

(7) The electronic watermark method according to (6), wherein the document image includes characters, tables and graphs.

(8) The predetermined control signal is a signal for preventing or suppressing unauthorized use due to illegal reading of a document image, a signal for restricting access, and a signal for preventing falsification. The electronic watermark method according to (6) above, which is a signal for searching and managing a document.

(9) The electronic watermarking method according to (6), wherein the extraction of the control information embedding target character is determined by its constituent element and appearance frequency.

(10) The electronic watermarking method according to (6), wherein the control information embedding and the reading are performed by using the relative relationship of the character components.

[0026]

DETAILED DESCRIPTION OF THE INVENTION Embodiments of the present invention will be described in detail below with reference to the drawings.

(Embodiment 1) FIG. 1 is a block diagram showing a schematic configuration of a digital watermarking apparatus according to an embodiment of the present invention. (A) is a block diagram of a digital watermark embedding unit,
(B) is a block diagram of a digital watermark reading unit.

In (a), 2 is an image input means (input unit) such as a scanner, a camera, or a file reading device for inputting a target document in which a digital watermark is embedded, 4 is a processor for performing various processes, and 6 is a processor 4. A keyboard for inputting the command, 8 is a disk for storing embedded information or a read document image, 10 is a temporary data storage for various processes performed by the processor 4, or a document read by the image input means (input unit) 2. A memory for accumulating images, 12 is a display for indicating a command input to the processor 4 and a processing state, and 14 is an output unit (output unit) such as a printer for outputting a document image in which access control information is embedded.

In (b), 22 is an image input means (input section) such as a scanner, a camera, or a file reading device for inputting a document in which a digital watermark is embedded, 2
4 is a processor for performing various processes, 26 is a processor 24
A keyboard for inputting a command to the device, 28 is a disk for saving the read document image or searching the original file of the read document, 30 is a temporary data storage for various processing performed in the processor 24, or an image input means. (Input unit) A memory for accumulating the document image read by the input unit 22, a display 32 for inputting an instruction to the processor 24, and a display showing the processing state, and 34 and 36 are network interfaces for utilizing the read document access control information. , A printer.

Next, the operation will be described. When embedding a digital watermark, follow the instruction entered from the keyboard 6,
First, a digitized embedded document image is acquired from the image input means (input unit) 2 and developed in the memory 10.

Further, embedded information (document access control information) is input from the keyboard 6 or the disk 8,
The processor 4 embeds the information in the document image developed on the memory 10. The document image in which the predetermined document access control information is embedded is output means (output unit) 14
Is output as a digital watermark embedded document.

On the other hand, when reading a digital watermark, in accordance with a command input from the keyboard 26, first, the digital watermark embedded document is digitized by the image input means (input section) 22 and expanded in the memory 30. Next, the processor 24 reads the embedded document access control information from the document image expanded on the memory 30, and performs a predetermined process based on the instruction. The predetermined processing is, for example, notification to the outside when an illegal reading is detected, search for the original document to the internal disk 28 or the outside, or printing out attribute information. Therefore, the network I / F 34 and the printer 36 are used.

Hereinafter, with reference to FIGS. 2, 3, 4, 5 and 6, the digital watermarking apparatus of this embodiment, particularly the processor 4,
The operation of various control processes executed by 24 will be described.

FIG. 2 is an overall flow chart when the processor 4 embeds a digital watermark in one document.

S200 is a text image input means (input section)
The document is read from the memory 2 and stored as electronic image data in the memory 10
This is a routine for performing control to transfer to. This routine also includes preprocessing such as orientation and skew correction of the read document.

In S202, the memory 10 in S200
This is a routine for performing area identification on the document image developed in the above and extracting all the character blocks in the image. This can be realized by applying, for example, the block selection technique described in JP-A-6-068301.

S204 is a routine for performing character recognition on all the character blocks extracted in S202.

S206 is a routine for extracting a target character in which the document access control information is to be embedded, from each character converted into the character code in S204.

S208 is a routine for inputting the document access control information to be embedded in the character extracted in S206. Here, the document access control information is, for example, copy restriction information, tampering prevention information, original document management information, or the like.

S210 is a routine for embedding the document access control information input in S208 in the characters extracted in S206.

S206 and S210 are routines which are the main purpose of the present invention, and will be described later in detail.

S212 is a routine for outputting the character image in which the document access control information is embedded in S210.

FIG. 3 is an overall flow chart when the processor 24 retrieves desired information from one digital watermark embedded document.

S300 is a text image input means (input section)
This is a routine for performing control to take in the electronic watermark embedded document from 22 and transfer it to the memory 30 as electronic image data. Similar to S200, this routine also includes preprocessing such as orientation and inclination correction of the read document.

In S302, the memory 30 in S300
This is a routine for performing area identification on the document image in which the digital watermark has been embedded in (3) and extracting all the character blocks in the document image.

S304 is a routine for performing character recognition on all the character blocks extracted in S302.

S306 is a routine for extracting only the characters in which the document access control information is embedded from the respective characters converted into the character code in S304.

S302, S304 and S306 are the same processing routines as S202, S204 and S206, respectively.

S308 is a routine for reading the document access control information from the characters extracted in S306.

S306 and S308 are S206 and S210.
This is a routine that is the main object of the present invention, and will be described later in detail.

S310 is a routine for performing a predetermined control in accordance with the document access control information read out in S308, and commands, for example, copy prohibition processing, document search processing, and the like.

FIG. 4 shows a target character extraction routine (S206, S) for embedding / reading a digital watermark, which is the main object of the present invention.
306) is a detailed flowchart.

S400 is a routine for controlling the transfer of the character code to the character extracting work memory based on the character recognition result.

S402 is a routine for determining whether or not all the character codes included in the document have been transferred to the character extracting work memory. If all the transfers have been completed, control is passed to S404, and if not, control is passed to S400.

S404 is a routine for counting, for each character, the target character in which the preset document access control information is to be embedded. Here, "set in advance" means setting in advance a kanji having a radical structure that is rather complicated, for example, a kanji having three or more component radicals. By making such a setting, it becomes possible to embed information above a certain level in a conspicuous and reliable manner. This will be described later in detail.

S406 is a routine for sorting the target characters counted in S404 by the count number.

In S408, if the count number exceeds a certain level,
That is, this is a routine for determining whether or not there are a certain number or more of embedding target characters that frequently appear in a document. this is,
In order to secure the embedding / reading accuracy, a certain number of target characters are taken, that is, the same information is repeatedly embedded in the same character. The greater the number above a certain number, the higher the accuracy.
As described in (3), it may be about two characters. By doing so, the accuracy of reading is greatly improved.

If the number of such target characters is a certain number or more, in order to guarantee the amount of embedded information, if it is possible to embed 3 bits of information per character as shown in FIG. As shown in), 11 characters are 30
It is possible to embed bit information, but to limit the number of characters in that way.

If it is determined that the number of target characters is not equal to or greater than a certain value, it is determined that the predetermined amount of information cannot be embedded, and control is passed to S414, and otherwise, control is passed to S410.

S410 is a routine for selecting the character having the highest appearance frequency in the document among the characters to be embedded and calculating the reference value for the embedding / reading operation (FIG. 6).
(Corresponds to (4)). This reference value will be described later.

S412 is a routine for performing embedding / reading operation of information other than the character for which the reference value is obtained, that is, the second or subsequent characters, for example, up to the eleventh embedding target character in the case of 30 bits, is obtained from the result of sorting in S406. is there. A specific method will be described later in detail with reference to FIGS. 5 and 6.

Step S414 is a routine for carrying out a predetermined process when it is judged in step S408 that the number of characters to be embedded is too small to embed. The predetermined process is, for example, issuing a warning.

FIG. 5 is a diagram for explaining the principle of the information embedding / reading method performed in S412, and FIG. 6 is an explanatory diagram for embedding a 30-bit information amount in an actual document.

In FIGS. 5 and 6, there are two small radicals in the upper part and one large radical in the lower part, such as a pattern of the letters "mold", in which the number of radicals is three. And explain the different patterns. The figure is exaggerated for the sake of explanation.

First, the extracted character image is decomposed into individual radicals and the reference value thereof is obtained. The reference value is the most important value when the document access information is embedded invisible in the present invention.

Specifically, as shown in FIG. 5A, it is represented by the ratio of the absolute height and width of the character image to the width and height of each radical separated from the width. To explain further,
Although it is called each radical here, only the upper two radicals,
The lower radicals are not defined as a percentage of their size. If this is defined for all radicals, the relative reference value will fluctuate, and further, the deformation of the letters will become large and deterioration will be noticeable. Is left and only the other two radicals (in this case) are used.

As a concrete information embedding method, consider embedding information of 3 bits per character by using the four reference values k, m, n and p defined previously. In FIG. 4 and S410, the reference values k, m, n, and p of the character having the highest reliability, that is, the frequency of appearance are calculated (see FIG. 6).
(Corresponds to (4)). On the other hand, a document access control to be embedded, for example, 30 bits is prepared, and S41 in FIG.
According to the rules shown in the right half of FIG. 5, k ', m', n ', p'in every 3 bits are sequentially obtained for the second and subsequent characters selected in 2, and the result is reflected. To change the aspect ratio of each radical of the target character (Fig. 6).
(Corresponds to (5)). Arbitrary information can be embedded by such a procedure.

At the time of reading, the character recognition is performed, the characters are extracted, sorted, and the first reference value is obtained in the same manner as at the time of embedding. Finally, the aspect ratio of each radical after the second rank is calculated. Is compared with a reference value to reproduce the allocated bit array. It is usually possible to perform this a plurality of times (here, at least twice or more) and judge by the majority decision, etc. to further improve the accuracy.

In FIG. 6, "001 100 111 0
00 010 011 111 101 111 00
It represents a procedure for actually embedding 30-bit information such as "0".

As described above, according to the present invention, after the area identification and the character recognition, the relative shape change of each radical of the complicated kanji character radicals is skillfully utilized to deteriorate the font. It is possible to realize a digital watermarking device with high noise immunity even at the time of reading, while securing the information embedding accuracy and amount above a certain level by sorting target characters according to the appearance frequency while minimizing the above. Further, since there is no dependency on the font size in principle, it is obvious that the method is effective even for a manuscript having a small number of characters.

(Embodiment 2) In the first embodiment, the radical structure adopted as the embedding target is a single pattern as shown in FIG. 5, but the present invention is not limited to this, and As shown in FIG. 7, a plurality of Chinese character patterns having radicals above a certain level may be set at the same time. In this case, the amount of embedded information is further increased.

[0072]

As described above, according to the present invention, by performing the region identification and the character recognition, and skillfully utilizing the relative shape change of each radical of the complicated kanji of the radical structure, While minimizing the deterioration of the
It is possible to secure a sufficient amount and realize a digital watermarking device having high noise resistance even at the time of reading. Also,
Since there is no dependency on font size in principle, it is clear that this method is effective even for a manuscript with a small number of characters.

[Brief description of drawings]

FIG. 1A is a block diagram showing a schematic configuration of a digital watermark embedding unit of a digital watermark apparatus according to an embodiment of the present invention.

FIG. 1B is a block diagram showing a schematic configuration of a digital watermark reading unit of a digital watermark device according to an embodiment of the present invention.

FIG. 2 is an overall flow chart when the processor 4 embeds a digital watermark in one document

FIG. 3 is an overall flow chart when the processor 24 extracts desired information from one digital watermark embedded document.

FIG. 4 is a target character extraction routine for embedding / reading a digital watermark, which is the main object of the present invention (S206, S306).
Detailed flow chart of

FIG. 5 is an explanatory diagram of the principle of the information embedding / reading method performed in S412.

6 (1), (2), (3), (4), (5) Explanatory diagram for embedding a 30-bit information amount in an actual document.

FIG. 7 is an explanatory diagram of the second embodiment (other embedding /
Read radical pattern)

FIG. 8 is an explanatory diagram of a conventional technique.

9A and 9B are explanatory views of a conventional technique.

FIG. 10 is an explanatory diagram of a conventional technique.

[Explanation of symbols]

2 Image input means (input section) 4 processors 6 keyboard 8 discs 10 memory 12 display 14 Output means (output section) 22 Image input means (input section) 24 processors 26 keyboard 28 discs 30 memory 32 display 34 Network Interface 36 Printer

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Hiroaki Ikeda             3-30-2 Shimomaruko, Ota-ku, Tokyo             Non non corporation F term (reference) 5B029 CC27                 5B057 AA11 BA02 CB19 CE08 CE09                       CG07 CH18 DA08 DA15 DA17                       DC19 DC31 DC36                 5B064 AA07                 5C076 AA14 BA06

Claims (10)

[Claims] 1. A digital watermark device for embedding a digital watermark in a document and reading the document, a document image inputting unit for inputting the document and converting it into an electronic image, a region extracting unit for extracting a specific region of the input document image, and the extracting. Character recognition means for recognizing characters in a region, character extraction means for extracting characters from the character recognition result according to a predetermined rule, control signal embedding means for embedding a predetermined control signal in the extracted characters, and control signal embedded document An electronic device comprising: a document image output unit for outputting an image; a control signal reading unit for reading a predetermined control signal from the extracted character; and a document control unit for controlling a document based on the read control signal. Watermark device. 2. The digital watermarking device according to claim 1, wherein the document image includes characters, a table, and a graph. 3. The predetermined control signal is a signal for preventing or suppressing unauthorized use due to illegal reading of a document image, a signal for restricting access, a signal for preventing falsification, and a document. The digital watermarking device according to claim 1, wherein the digital watermarking device is a signal for searching and managing. 4. The digital watermarking device according to claim 1, wherein the extraction of the control information embedding target character is determined by its constituent element and appearance frequency. 5. The digital watermarking device according to claim 1, wherein the control information embedding and the reading are performed by using the relative relationship of the constituent elements of the character. 6. A digital watermark device for embedding a digital watermark in a document and reading the document, a document image inputting unit for inputting the document and converting it into an electronic image, a region extracting unit for extracting a specific region of the input document image, and the extracting. Character recognition means for recognizing characters in a region, character extraction means for extracting characters from the character recognition result according to a predetermined rule, control signal embedding means for embedding a predetermined control signal in the extracted characters, and control signal embedded document An electronic device comprising: a document image output unit for outputting an image; a control signal reading unit for reading a predetermined control signal from the extracted character; and a document control unit for controlling a document based on the read control signal. Watermark method. 7. The digital watermarking method according to claim 6, wherein the document image includes characters, a table, and a graph. 8. The predetermined control signal is a signal for preventing or suppressing unauthorized use due to illegal reading of a document image, a signal for restricting access, a signal for preventing falsification, and a document. 7. The digital watermarking method according to claim 6, wherein the digital watermarking method is a signal for searching and managing. 9. The digital watermarking method according to claim 6, wherein the extraction of the control information embedding target character is determined by its constituent element and appearance frequency. 10. The digital watermarking method according to claim 6, wherein the control information embedding and the reading are performed by using the relative relationship of the constituent elements of the character.