JP2006217044A - Image processing method and apparatus for electronic watermarking - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing method and apparatus for electronic watermarking with higher confidentiality in comparison with that of prior arts and to provide a recording medium readable by a computer installed thereon with an image processing program. <P>SOLUTION: The image processing method is provided, which carries out electronic watermarking by using key image data to embed confidential information data to source image data. First, an image processing for extracting a characteristic point (S2) applies image processing to the key image data for the purpose of extracting a prescribed characteristic point to output the key image data after the image processing having a plurality of characteristic point coordinate values. Then an information embedding processing (S6) embeds the confidential information data to each coordinate of the source image data respectively corresponding to each characteristic point coordinate of the key image data after the image processing by using a prescribed embedding rule and outputs the image data after the embedding. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image processing method and apparatus for digital watermark, an image processing program including steps of the image processing method for digital watermark, and a computer-readable recording medium storing the image processing program. .

  In recent years, with the progress of informatization and networking in society, the Internet has become indispensable for daily life. Along with this, society's interest in information security, such as measures for preventing leakage of confidential information, is increasing (see Non-Patent Document 1).

  Currently, cryptography is generally used for communication while maintaining the confidentiality of information. However, by exchanging encrypted information, there is a risk that the communication of important information may be intercepted even if the content of the information being communicated is not decrypted. Therefore, it can be said that the secrecy of information is higher when the important information itself is kept secret. In view of this, there is a conceivable solution for concealing highly confidential information in digital content such as a still image, thereby eliminating the opportunity to discover confidential information and preventing leakage. A typical solution is digital watermarking (see Non-Patent Document 2).

  Digital watermarking is a technique for embedding information different from content into digital content such as still images, moving images, and music. At this time, it is common to embed the human body so that it cannot be perceptually perceived by the human visual / auditory biological characteristics.

  First, the existing technology and classification of digital watermark will be described below. There are various kinds of digital contents to which digital watermarks are applied as shown in FIG. Even when limited to still images, there are categories such as binary images (see Non-Patent Document 3), multi-valued images, color images, and the like (see Non-Patent Document 4). Hereinafter, the content of still image data before the digital watermark information is embedded is referred to as original image data.

As typical digital watermark embedding techniques used in still image data, the following three types can be cited (see Non-Patent Document 6).
(1) Pixel replacement method: When image data is expressed as 8 bits per pixel, the image can be divided into 8 bit planes. Each pixel includes luminance information indicating brightness. The least significant bit is a redundant bit plane that has little visual impact. The pixel replacement method is a method of embedding information in this redundant lower bit plane. Improvement methods such as a bit-plane diffusion method and a pixel space utilization method have been proposed in order to increase resistance to tampering.
(2) Frequency utilization method: A method of transforming image information into a frequency domain using a transform function such as DCT / Fourier transform and embedding digital watermark information in a redundant frequency domain. The main feature of this method is that it is strong against compression such as MPEG and JPEG (see Non-Patent Document 7).
(3) Density pattern method: A method used in a printer or the like, in which one pixel is divided into several micro dots, and gradation is expressed by pixels having only monochrome binary information. By dividing one pixel into micro dots, the redundancy is increased. The digital watermark information is embedded in the digital watermark embedding technique based on the density pattern method.

  Next, the usage forms of digital watermarks can be broadly classified into the following five types (see Non-Patent Document 8).

(A) Display of copyright and use conditions.
(B) Tracking unauthorized users.
(C) Recording usage restriction information.
(D) Detection of alteration of content.
(E) Concealment of confidential information.

  Depending on the reference, “digital watermark” for the purpose of concealing confidential information is classified as “steganography”. However, the word “steganography” is relatively low-profile and not well established. For this reason, in this specification, it is classified as “digital watermark”.

  In recent years, with the widespread use of digital contents such as still images, moving images, and music, embodiments for the purpose of protecting copyrights such as (a) to (d) above have been actively performed and matured (non-patented). Reference 9 and 10). On the other hand, with respect to the digital watermark for the purpose of concealing the confidential information (e), there are few published embodiment development examples, and there is much room for the embodiment regarding the definition and the amount of information that can be concealed (non-patent) Reference 11). Furthermore, the effect on the Internet has not been elucidated as much as encryption technology (see Non-Patent Document 12). Compared to the embodiment of the digital watermark for the purpose of copyright protection, many items to be implemented remain.

  There are significant differences in the technical conditions of digital watermarking as shown in the table of FIG. As can be seen from the table of FIG. 4, there is a significant difference in the information that must be protected for the two technologies. Since the copyright protection digital watermark embeds the author's name in order to clarify the copyright of the content, the content to be protected is the content. On the other hand, in the digital watermark for concealing confidential information, the content is used only to disguise it, so what is really meaningful is embedded information. In order to convey or hide this information to the other party, digital contents such as images are borrowed and disguised on the surface.

  Also, the requirements for the strength of the embedded watermark information are different. From the standpoint of copyright protection, copyright information must not be altered or deleted. Therefore, a condition that it is resistant to processing and tampering is required. On the other hand, in the case of concealing confidential information, it is strongly required to prevent leakage of embedded information. There must be a condition that no matter what happens, it must not be deciphered by a third party.

  In some cases, the digital watermark used to conceal confidential information is positioned as a “special encryption”, but here it is considered to be completely different from encryption. Cryptography is a technique (see Non-Patent Document 13) that mixes information data that a third party does not want to know and does not understand the content, but it cannot be hidden that there is dubious data there. That is, the secret content is protected, but the secret existence is emphasized. On the other hand, concealment of confidential information is a technique for making secret information secretly inserted into another casual information medium so that even the existence of a secret is unknown.

JP2003-250036A. JP 2001-313822 A. Japanese Patent Application Laid-Open No. 2004-179744. Japanese Patent Application Laid-Open No. 2004-109172. JP-A-11-317859. Ministry of Internal Affairs and Communications, “Survey Embodiment Report on Information Security Auditing in Local Governments”, http://www.soumu.go.jp/c-gyousei/daityo/pdf/031226_1_s6.pdf, December 2003. Kokoo Matsui, “Image Deep Cryptography”, Morikita Publishing, Tokyo, 1998. Niimi Michiko et al., "Large-capacity deep image encryption using area division by complexity", IEICE Transactions, D-2, Vol. J81-D-2, no. 6, pp. 1132-1140, June 1998. JPO, “Diagram of Researched Technology”, http://www.jpo.go.jp/shiryou/s_sonota/hyoujun_gijutsu/denshi_sukashi/jyukeizu.pdf, November 30, 2004. Kokoo Matsui, “Basics of Digital Watermarking”, Morikita Publishing, Tokyo, 1998. Japan Patent Office, “Digital watermark embedding technology”, http://www.jpo.go.jp/shiryou/s_sonota/hyoujun_gijutsu/denshi_sukashi/, November 30, 2004. Tsuji et al., “Application of Robust Digital Watermark to Fingerprint Image”, IEICE, vol. 102, no. 632 (IE2002 223-258), pp. 119-122, February 2003. Supervised by Naohisa Komatsu et al., “Digital Watermark Technology”, The Institute of Image Electronics Engineers (ed.), Tokyo Denki University Press, 2004. Onozuka, “Digital Watermarking and Content Protection”, Ohm Company Development Bureau (ed.), Ohm Company, Tokyo, 2001. Futada Takahashi, “Digital watermarking protects the multimedia era”, Nikkei Electronics, vol 2-24, no. 683, pp. 99-124, February 1997. ARTZ D, Digital "Steganography: Hiding Data within Data", IEEE Internet Computer, vol.5, no.3, pp.75-80, May 2001. Tsutomu Matsumoto et al., “Development of Embodiment of Secret Communication Using Steganography”, 18th IPA Technology Presentation, 1999. Tsutomu Matsumoto, “Digital Watermarking and Related Technologies”, IEICE Journal, vol. 120, no. 7pp. 430-433, July 2000. Yukiichi Sakai, “Basics and Applications of Digital Image Processing”, CQ Publisher, Tokyo, 2003.

  In recent years, digital watermarks are mainly used for protecting copyright information of digital contents. Therefore, the focus of the embodiment is how to apply the above-described embedding technique and embed a digital watermark that is resistant to processing and tampering without damaging the quality of digital content.

  However, as described above, digital watermarks are not limited to copyright protection but are also used to conceal confidential information. In that case, the condition required for embedding the digital watermark is more applicable to the condition that the third party cannot decipher anything, rather than the condition that it is resistant to processing and tampering. Therefore, an embedding method that meets this condition is required.

  A digital watermark used to conceal confidential information is also implemented. However, since most of them embed information according to a fixed rule, once the rule is deciphered, there is a high possibility that the embedded information is read out. Even if it is not, it is easy for the developer to decipher it. Furthermore, password authentication used when embedding or reading out digital watermark information is mostly passwords made up of alphanumeric characters.

  Therefore, it is necessary to implement how the digital watermark used for concealing confidential information is not deciphered by a third party.

  SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems and to provide an image processing method and apparatus for electronic watermarking with higher confidentiality compared to the prior art, and a computer-readable recording medium storing the image processing program. Is to provide.

  In the present invention, when embedding confidential information, in addition to the digital watermark information to be concealed and the digital content used for concealment, third digital content called a key file is used. We propose a method to determine the embedding parameters depending on the feature points of this key file.

An image processing method according to a first invention is an image processing method for digital watermarking in which confidential information data is embedded in original image data using key image data,
An image processing step of performing image processing for extracting predetermined feature points on the key image data and outputting image data for key after image processing having a plurality of feature point coordinate values;
Information for embedding the confidential information data in each coordinate of the original image data corresponding to each feature point coordinate of the key image data after the image processing using a predetermined embedding rule, and outputting the image data after embedding And an embedding step.

  The image processing method further includes a conversion step of converting the key information data into key pseudo image data and outputting the key image data as the key image data before the image processing step.

An image processing method according to a second aspect of the present invention is an image processing method for digital watermarking for reading confidential information data from image data after embedding confidential information data using key image data with respect to original image data. There,
An image processing step of performing image processing for extracting predetermined feature points on the key image data and outputting image data for key after image processing having a plurality of feature point coordinate values;
The confidential information data is read out from the coordinates of the embedded image data corresponding to the feature point coordinates of the key image data after the image processing using a predetermined reading rule, and the confidential information data after the reading is read out. And a reading step of outputting.

  The image processing method further includes a conversion step of converting the key information data into key pseudo image data and outputting the key image data as the key image data before the image processing step.

An image processing apparatus according to a third invention is an image processing apparatus for digital watermarking that embeds confidential information data in original image data using key image data,
Image processing means for performing image processing for extracting predetermined feature points on the key image data, and outputting image-processed key image data having a plurality of feature point coordinate values;
Information for embedding the confidential information data in each coordinate of the original image data corresponding to each feature point coordinate of the key image data after the image processing using a predetermined embedding rule, and outputting the image data after embedding And an embedding means.

  The image processing apparatus further includes conversion means for converting key information data into key pseudo image data and outputting the key image data as the key image data before processing by the image processing means.

An image processing apparatus according to a fourth aspect of the present invention is an image processing apparatus for digital watermarking for reading confidential information data from image data after embedding confidential information data using key image data with respect to original image data. There,
Image processing means for performing image processing for extracting predetermined feature points on the key image data, and outputting image-processed key image data having a plurality of feature point coordinate values;
The confidential information data is read out from the coordinates of the embedded image data corresponding to the feature point coordinates of the key image data after the image processing using a predetermined reading rule, and the confidential information data after the reading is read out. And a reading means for outputting.

  The image processing apparatus further includes conversion means for converting key information data into key pseudo image data and outputting the key image data as the key image data before processing by the image processing means.

  An image processing program according to a fifth invention includes the steps of the image processing methods described above.

  A computer-readable recording medium according to a sixth aspect of the present invention stores the image processing program.

  According to the image processing method and apparatus for digital watermark according to the present invention, the original image data in which the digital watermark information is embedded, the key image data, and the image processing for detecting feature points for the key image data (image The embedded digital watermark information can be read for the first time when all four elements of the effect processing) type and threshold value match. Therefore, even if the user is a developer, the digital watermark information embedded by the user cannot be decrypted. Compared with the existing technology in which embedding is performed according to a fixed rule, it is difficult for a third party to decipher, and it is possible to execute digital watermark embedding that is extremely confidential compared to the conventional technology.

  Hereinafter, embodiments according to the present invention will be described with reference to the drawings. In addition, in each following embodiment, the same code | symbol is attached | subjected about the same component.

  In the present embodiment, the still image content is selected because it is widely used all over the world and the digital watermark technology is most actively implemented (see, for example, Non-Patent Document 5). Among still image contents, in the present embodiment, an example of an image processing method and apparatus for image watermarking on a color image will be described below.

FIG. 1 is a block diagram showing a configuration of a digital watermark system according to an embodiment of the present invention, and FIG. 2 is an image processing apparatus 10-1, 10-2 (generically, denoted by reference numeral 10) of FIG. It is a block diagram which shows the structure of these. FIG. 5 is a flowchart showing a flow of information embedding processing for digital watermarking according to this embodiment. An image processing method of information embedding processing for digital watermark according to the present embodiment is an image processing method for digital watermark that embeds confidential information data in original image data using key image data. As shown in Figure 5,
(A) Step S2 of image processing for executing image processing for extracting predetermined feature points on the key image data and outputting post-image key image data having a plurality of feature point coordinate values When,
(B) The confidential information data is embedded in each coordinate of the original image data corresponding to each feature point coordinate of the key image data after the image processing using a predetermined embedding rule, and the embedded image data is embedded Step S6 of the information embedding process to be output is included.

  In the present embodiment, an example of a method for embedding a digital watermark will be described below for the purpose of how a still image cannot be decoded by a third party. It is assumed that a still image to be used is an uncompressed format and is embedded by applying a pixel replacement method.

  First, a method for determining an embedding position of digital watermark information using a key file, which is executed by using the image processing apparatus 10 of FIG. 2, which is a digital computer such as a personal computer, will be described with reference to FIGS. 6 is an image showing an example of the original image data A in FIG. 5, and FIG. 7 is an image showing an example of the key image data B in FIG. 8 is an image showing an example of the key image data B ′ after the image processing of FIG. 5. FIG. 9 is a diagram showing the image processing of FIG. 8 in the original image data A of FIG. 6 during the information embedding processing of FIG. It is an image when the later key image data B ′ is made to correspond.

  In FIG. 1, an image processing apparatus 10-1 that executes information embedding processing for digital watermark and an image processing apparatus 10-2 that executes information reading processing for digital watermark are connected to a network such as the Internet 60, for example. The image processing apparatus 10-1 embeds the confidential information data C embedded in the original image data A, and then transmits the image data D after the information embedding to the image processing apparatus 10-2 via the Internet 60. To do. The image processing apparatus 10-2 receives the image data D after information embedding, and the image processing apparatus 10-2 reads the confidential information data C from the image data D after information embedding. In the example of FIG. 1, the image data after information embedding is transmitted via a network such as the Internet 60. However, the present invention is not limited to this, and the image data after information embedding is transferred to a recording medium such as an optical disk. You may make it convey after recording.

(1) In the digital watermark embedding method according to the prior art, only two digital contents of original image data A and embedded confidential information data C are prepared. On the other hand, the method according to the present embodiment uses a digital data file called key data (for example, key image data B) in addition to these digital contents. The key data used here is not special data. Almost all types of digital data files such as image data, audio data, and text data existing on a personal computer can be used.
(2) Determine the format of the key data file. If the key data file is in the image format, it is stored as it is in the temporary memory 22a in the RAM 22 of the image processing apparatus 10 as the key image data B. Otherwise, the key data is converted into binary data of a two-dimensional array and then stored as temporary pseudo image data on the temporary memory 22a. That is, the key information data may be converted into key pseudo image data and output as the key image data. In this embodiment, the two-dimensional original image data A and the two-dimensional key image data B are used. However, the present invention is not limited to this, and the multidimensional original image data A and the multidimensional key data are used. Image data B may be used.
(3) By performing effect image processing for extracting feature points on the two-dimensional key image data B stored in the temporary memory 22a of the image processing apparatus 10, the feature points of the key image data B are obtained. To detect. The feature point detection will be described later.
(4) On the original image data A in which confidential information data that is digital watermark information is embedded, the confidential information data is embedded in accordance with the coordinates of each feature point obtained in (3).

  Next, image processing for feature point detection will be described below. In this embodiment, three types of binarization processing, contour tracking processing, and dotted line processing are used as the method for detecting the feature points of the key image data B. The present invention is not limited to this, and other image processing such as edge processing may be used. Here, the binarization processing is processing for obtaining binary image data from multi-value image data, and generally threshold processing is used as a parameter. In the threshold processing of an image when the threshold value is t, binary image data is obtained by the following operation.

[Equation 1]
g (x, y)
= 1; when f (x, y) ≧ t = 0; when f (x, y) <t (1)

  The coordinates of the pixel value of 1 or 0 obtained at this time are used as feature points. In general, a method based on a density histogram is used as a method for obtaining a threshold value. However, considering that this threshold value itself can be used as a password, in this embodiment, the user can freely set it.

  In the contour tracking process, the contour line of an image can be found using an 8-neighbor contour line tracking method (see, for example, Non-Patent Document 14), and the coordinate value of the contour line can be used as a feature point. Furthermore, the dotted line process is a process for drawing the outline of an image with a dotted line. After obtaining the contour line with the difference operator, it can be obtained by combining expansion and reduction. The coordinate value of the contour dotted line obtained at this time can be used as a feature point.

  In the processing of FIG. 5, for example, image processing (S2) for extracting feature points is performed on the key image data B of FIG. 7 to obtain, for example, key image data B ′ after the image processing of FIG. .

  Next, the law of the image replacement method according to this embodiment will be described below. In the present embodiment, an existing pixel replacement method is applied to the algorithm of the digital watermark embedding unit. Also, the digital watermark information to be embedded is digital content in text format. In the pixel replacement method used this time, the digital watermark information to be embedded is converted into a binary display and embedded according to the digital watermark embedding rule. The embedding rules were defined as follows.

(1) The least significant bit is changed with digital watermark information.
(2) When the confidential information that is digital watermark information is bit 0, the pixel data is changed to an even number. If it is even from the beginning, it will not be changed.
(3) When the confidential information that is digital watermark information is bit 1, the pixel data is changed to an odd number. If it is odd from the beginning, it is not changed.
(4) However, when changing to odd or even, the increase rule is uniformly set, and the decrease rule is set only when the pixel value becomes the maximum value.

  As an example, FIG. 10 shows processing when 8 bits of confidential information that is digital watermark information “01100011” are embedded in the pixel values as shown in FIG. In accordance with the information embedding rule based on the pixel replacement method described above, the digital watermark information is embedded as follows.

(1) First, the maximum number of bytes X in which confidential information data that is digital watermark information can be embedded is obtained. The number X of feature points extracted from the key image data B by the feature point coordinate number X calculation process (S3) of FIG. 5 is counted, and one feature point is converted as one bit. When the confidential information is text information, since one byte is required to display one half-width character, the maximum half-width character can be embedded.
(2) For example, the confidential information data C, which is digital watermark text information, is counted by the confidential information data data byte count Y counting process (S1) in FIG. In this embodiment,
(3) After the confidential information data C is converted into binary data Cb by binary data conversion processing (S5), information of “number of bytes Y of confidential information data” is added by additional information addition processing (S5) in FIG. Additional information including 2 bytes and 1 byte of information of “completion recognition character” (for example, “e”) indicating the end of the embedded confidential information data is added, and the confidential information data (Y + 3 bytes) after the additional processing is added. Next, the information is embedded with a digital watermark, and in the feature point coordinate number X calculation process (S3) of Fig. 5, if the number of bytes Y + 3 of the confidential information data C 'after the addition process exceeds the maximum number of bytes X that can be embedded Returns an error.
(4) In the information embedding process (S6) in FIG. 5, the secret information data C ′ after the additional process including the binarized confidential information data C is sequentially applied to the feature point coordinates of the key image data B ′ after the image process. Is embedded in the coordinates of the original image data A for information embedding, and the image data D after information embedding is obtained and transmitted by transmission processing ((S7)).

  Next, processing for reading digital watermark information will be described below. When reading the digital watermark information, the reverse process of embedding is performed. The processing procedure is shown below.

(1) Image data in which digital watermark information is already embedded is received by reception processing, and the key image data B is designated and the file is read.
(2) Similar to the information embedding process for the digital watermark in FIG. 5, the key point is used in accordance with the type of feature point extraction effect process for the key image data and the threshold value which is a parameter used there. By performing image processing (S2) for feature point extraction on the image data B, key image data B ′ after image processing is obtained.
(3) The feature point coordinates of the key image data B ′ obtained in the above (1) and (2) are made to correspond to the coordinates of the image data D in which the confidential information data C ′, which is digital watermark information, is embedded. A read process is performed along the coordinate values according to a predetermined read rule to determine whether correct digital watermark information exists. If correct digital watermark information is embedded, the information is displayed.

  Note that the readout rule used in the present embodiment is a rule opposite to the embedding rule, and reads “Information 0” when the pixel value is an even number, and “Information 1” when the pixel value is an odd number. Is read.

  In the present embodiment, as described above, “completion recognition character” is used as a method for determining whether or not correct digital watermark information exists. When embedding digital watermark information in an image, a predetermined “completion recognition character” is added to the end of the embedded digital watermark information. When reading out the digital watermark, it is checked whether or not the “completion recognized character” exists correctly at the end of the digital watermark information (step S58 in FIG. 14). Only when the completion recognition character exists correctly (YES in step 46 in FIG. 15), the read watermark information is displayed and output.

  Further, the image processing apparatus 10 of FIG. 2 and the image processing for digital watermark executed thereby will be described in detail below. Here, the image processing apparatus 10 in FIG. 2 includes the processing function of the image processing apparatus 10-1 in FIG. 1 and the processing function of the image processing apparatus 10-2 in FIG. You may comprise separately for each processing function.

In FIG. 2, the image processing apparatus 10
(A) a CPU (central processing unit) 20 of a computer that calculates and controls the operation and processing of the image processing apparatus 10;
(B) a ROM (read only memory) 21 for storing a basic program such as an operation program and data necessary for executing the basic program;
(C) a RAM (random access memory) 22 including a temporary memory 22a that operates as a working memory of the CPU 20 and temporarily stores parameters and data necessary for image processing;
(D) an image memory 23 configured by, for example, a hard disk memory and storing image data to be transmitted / received, image data during image processing, and image data after image processing;
(E) a program memory 24 configured with, for example, a hard disk memory and read using the CD-ROM drive device 45 and storing the image processing programs of FIGS.
(F) a communication interface 61 that is connected to another image processing apparatus 10 via the Internet 60 and transmits / receives data to / from a communication interface 61 of the other image processing apparatus 10;
(G) A keyboard connected to a keyboard 41 for inputting predetermined data and instruction commands, receiving data and instruction commands input from the keyboard 41, performing predetermined interface processing such as signal conversion, and transmitting the data to the CPU 20 Interface 31;
(H) Connected to a mouse 42 for inputting instruction commands on the CRT display 43, receives data and instruction commands input from the mouse 42, performs predetermined signal conversion and other interface processing, and transmits them to the CPU 20. A mouse interface 32;
(I) Connected to a CRT display 43 that displays image data processed by the CPU 20, a setting instruction screen, and the like. The image data to be displayed is converted into an image signal for the CRT display 43 and output to the CRT display 43 for display. Display interface 33,
(J) A printer interface 34 that is connected to a printer 44 that prints image data processed by the CPU 20, predetermined processing results, and the like, performs predetermined signal conversion of print data to be printed, and outputs to the printer 44 for printing. When,
(K) Connected to a CD-ROM drive device 45 that reads program data of the image processing program from the CD-ROM 45a in which the image processing programs of FIGS. 11 to 14 are stored, and the program data of the read image processing program is predetermined. A drive device interface 35 for performing signal conversion of the data and transferring the program to the program memory 24,
These circuits 20-24, 31-34 and 61 are connected via the bus 30.

  FIG. 11 is a flowchart showing information embedding processing for digital watermark executed by the image processing apparatuses 10-1 and 10-2 (10) of FIG.

  In FIG. 11, first, the original image data A, the key image data B, and the confidential information data C stored in advance in the image processing apparatus 10 are designated using the keyboard 41 or the mouse 42, respectively, step S12. The image processing type and parameters for feature point extraction are input and specified. Next, in step S13, confidential information data byte count Y count processing (S1) is executed, in step S14 image processing for feature point extraction (S2) is executed, and in step S15 feature point coordinate count X count processing ( S3) is executed. In step S16, it is determined whether or not Y + 3> X. If NO, the number of feature points is smaller than the number of bytes of confidential information data, and the information embedding process cannot be executed. After that, the information embedding process for the digital watermark ends. On the other hand, if YES in step S16, binary data conversion processing (S4) is executed in step S18, additional information addition processing (S5) is executed in step S19, and then information embedding processing in FIG. 12 is executed in step S20. (S6) is executed. Further, after executing the transmission process in step S21, the information embedding process for the digital watermark is terminated.

  FIG. 12 is a flowchart showing information embedding processing (S6, S20) which is a subroutine of FIG.

  In FIG. 12, first, parameter n is initialized to 1 in step S31, the nth coordinate of key image data B ′ is read in step S32, and the nth coordinate of key image data B ′ is read in accordance with an embedding rule in step S33. The n-th data of the confidential information data C ′ is embedded in the coordinates of the original image data A corresponding to the coordinates, and the image data D after the information is embedded is stored in the temporary memory 22a. Next, in step S34, it is determined whether or not n ≧ Y + 3. If NO, the process proceeds to step S35, the parameter n is incremented by 1, and the process returns to step S32. If YES in step S34, the process returns to the original main routine.

  FIG. 13 is a flowchart showing information reading processing for digital watermark executed by the image processing apparatuses 10-1 and 10-2 (10) of FIG. It is assumed that the key image data B is stored in the image processing apparatus 10 in advance before executing this process.

  In FIG. 13, first, reception processing is executed in step S41, key image data B is designated in step S42, and the type and parameters of image processing for feature point extraction are designated in step S43. In step S44, image processing for feature point extraction (S2) is executed, in step S46, feature point number X counting processing (S3) is executed, and in step S46, information reading processing (S46) in FIG. 14 is executed. Next, in step S47, it is determined whether or not there is a “completion recognized character”. If NO, an error is displayed in step S48 and then the process proceeds to step S43 (or even if the information reading process for the digital watermark is terminated) On the other hand, if YES, the process proceeds to step S49. The original image data A is obtained by executing the restoration process of the original image data by performing the process of reversing the coordinates in which the confidential information data C ′ is embedded in step S49 in the reverse process of the information embedding process. Further, the confidential information data C read out in step S50 (the information obtained by deleting the additional information from the confidential information data C ′) and the restored original image data A are displayed and output, and information reading for the digital watermark is performed. The process ends.

  FIG. 14 is a flowchart showing information read processing (S46) which is a subroutine of FIG.

  In FIG. 14, first, the parameter n is initialized to 1, and the parameter Y is initialized to 1. Next, in step S52, the nth coordinate value of the key image data B ′ is read, and in step S53, from the coordinate value of the received image data D corresponding to the nth coordinate value of the key image data B ′. The nth data d ′ (n) of the information read data D ′ is read and stored in the temporary memory 22a. Next, in step S54, it is determined whether or not n = 2. When the determination is NO, the process proceeds to step S56. When the determination is YES, the process proceeds to step S55, where data (d '(1), d' (2)) The number indicated by is a parameter Y (the number of bytes of embedded confidential information data), and the process proceeds to step S56. In step S56, it is determined whether or not n ≧ Y + 3. If YES, the process proceeds to step S58. If NO, the process proceeds to step S57, the parameter n is incremented by 1, and the process returns to step S52. In step S58, it is determined whether or not the data d ′ (n) = “e” (completion recognition character) after the confidential information data C. If YES, the process proceeds to step S60. If NO, the process proceeds to step S59. move on. In step S59, it is determined that there is no “completion recognition character” and the process returns to the original main routine. On the other hand, after determining that there is a “completion recognition character” in step S60 and determining in step S61 the data (d ′ (3),..., D ′ (Y + 2)) as confidential information data C that is digital watermark information, Return to the original main routine.

  In the present embodiment, the program data of the image processing of FIGS. 11 to 14 is loaded into the program memory 24 when executed and stored in the CD-ROM 45a, but the present invention is not limited to this. You may store in various recording media, such as recording media of optical disks or magneto-optical discs, such as CD-R, CD-RW, DVD, and MO, or recording media of magnetic discs, such as a floppy disk. These recording media are computer-readable recording media. Further, the image processing program data of FIGS. 11 to 14 may be stored in the program memory 24 in advance and the image processing may be executed.

  The present inventor conducted the following experiment in order to confirm the effect of the proposed method. First, the specifications of the experimental image processing apparatus are shown below.

[Table 1]
――――――――――――――――――――――――――――――――――――――
OS: Microsoft Windows XP Professional
CPU: Pentium 3800MHz
Memory: 512MB
――――――――――――――――――――――――――――――――――――――

  For embedding the digital watermark, 10 still images are prepared. Two images are selected from them, one as an original image, and the other as a key image. Using these two images, digital watermark information is embedded in advance. As the embedded watermark data, we used the full text excluding the original drawings used in the paper presentation. The number of embedded half-width characters was 17515 characters, and the time required for embedding was 1.05 seconds.

  In the readout experiment, the subject is provided with a tool used to read out the 10 images and digital watermark data described above. Explain to the subject that there are one image in which digital watermark information is embedded and one key image in 10 images. Ask the subject to find out the embedded image and key image, and try to read the digital watermark data with the tool. The experiment has no time limit and continues until the subject reads successfully or gives up.

  As a result of the above experiment, when eight subjects performed this experiment, all of them gave up halfway and could not read the embedded information. The average time for the subjects to give up was 10 minutes. The latest subject gave up 15 minutes.

  Based on the above experimental results, the round robin method was most frequently used when the subject searched for an image in which digital watermark information was embedded and a key image. However, although it is 10 images, the number of combinations such as effect types and threshold values is as follows:

[Equation 2]
₁₀ C ₁ × 1 × 255 + ₁₀ C ₂ × 2 × 1 = 111565 (2)

  The first term on the left side of Equation (2) is the number of combinations when the feature points of the key image are extracted by binarization, and is multiplied by 255 to set a threshold value. The second term on the left side of Equation (2) is the number of combinations when the remaining two types of effects are used.

  This number of combinations is not a number that humans can try manually. As you can see from the average time of the experiment, if you repeat the same work, you get bored in a short time. Some subjects used free software to identify images with embedded information from changes in binary image data. However, since the digital watermark information is embedded using the pixel replacement method, it cannot be specified even when compared with the image before embedding when it is converted to binary.

  Although this experiment limited to 10 images, the number of combinations is astronomical on a personal computer with many contents. In addition, although the key file is an image file this time, it is almost impossible to specify the digital content because it can be used almost any digital content that actually exists on a personal computer.

  As described above, in the method according to the present embodiment, when embedding digital watermark information in an image, the determination of the embedding position is not determined by a predetermined algorithm as in the existing technology, but the user can arbitrarily specify the embedding position. Use the key file you choose to decide. Therefore, it is possible to realize irregular and unpredictable embedding. By making the digital watermark embedding position completely dependent on the key file, the digital watermark information is not read out only with the image in which the digital watermark information is embedded.

  Also, if the type of feature point detection method of the key file and its threshold value do not match, the coordinates of the detected feature points will be shifted, so that the digital watermark information cannot be read correctly. However, when comparing an image in which digital watermark data is embedded with an image before being embedded, there is a difference in bit information. The amount of information that can be embedded in which it is effective to delete the original image in order to enhance confidentiality depends on the size of the original image.

  In the digital watermark technology, there is a contradictory property that the digital watermark information is not easily lost and the quality of the digital content is not deteriorated. This is because the more the flaws (noise) are reinforced to embed digital watermark information, the more the content deteriorates. However, this is required when protecting copyright information. From the standpoint of concealing confidential information, the loss of digital watermark information is not a problem. Therefore, the digital content can be embedded with almost no deterioration. Thus, it is considered that digital watermark embedding that is less likely to be detected by a third party can be realized.

  This time, after determining the embedding coordinates in the key file, the pixel replacement method was applied as an example, and the digital watermark information was embedded as text information. However, it is considered that other existing embedding methods can be applied to the embedding method after determining the coordinates by the key file. Further, the embedded digital watermark information can embed not only text information but also information such as images and sounds.

  When performing confidential communication on the network, the receiving side is instructed in advance to the key file as a secret key, and the content in which the digital watermark is embedded is distributed later. By doing so, it is possible to eliminate the opportunity for discovery of confidential information and to prevent leakage. Furthermore, even if intercepted, it is almost impossible to decrypt without a secret key file. Therefore, it is considered that confidential communication with higher confidentiality can be performed.

  Although inventions related to the present invention are disclosed in Patent Documents 1 to 5, “An image processing method for digital watermarking in which confidential information data is embedded in original image data using key image data”. An image processing step of executing image processing for extracting predetermined feature points on the key image data and outputting image data for key after image processing having a plurality of feature point coordinate values; Information embedding for embedding the confidential information data in each coordinate of the original image data corresponding to each feature point coordinate of the image data for key after image processing using a predetermined embedding rule and outputting the image data after embedding There is no disclosure or suggestion of an image processing method characterized by including a step.

  As described above in detail, according to the image processing method and apparatus for digital watermark according to the present invention, feature point detection for original image data, key image data, and key image data in which digital watermark information is embedded. The embedded digital watermark information can be read for the first time when all of the four elements of the type of image processing (image effect processing) and the threshold value match. Therefore, even if the user is a developer, the digital watermark information embedded by the user cannot be decrypted. Compared with the existing technology in which embedding is performed according to a fixed rule, it is difficult for a third party to decipher, and it is possible to execute digital watermark embedding that is extremely confidential compared to the conventional technology.

It is a block diagram which shows the structure of the digital watermark system which concerns on one Embodiment of this invention. It is a block diagram which shows the structure of the image processing apparatuses 10-1 and 10-2 (10) of FIG. It is a figure which shows the classification | category of the digital watermark based on a prior art. It is a table | surface which shows the difference between concealment of the confidential information based on a prior art, and copyright protection. It is a flowchart which shows the flow of the information embedding process for the digital watermark based on this embodiment. It is an image which shows an example of the original image data A of FIG. 6 is an image showing an example of key image data B in FIG. 5. FIG. 6 is an image showing an example of key image data B ′ after image processing in FIG. 5. FIG. 8 is an image when the key image data B ′ after the image processing of FIG. 8 is associated with the original image data A of FIG. 6 during the information embedding processing of FIG. It is a table | surface which shows the example of embedding using the embedding rule which concerns on this embodiment. It is a flowchart which shows the information embedding process for the digital watermark performed by the image processing apparatuses 10-1 and 10-2 (10) of FIG. 12 is a flowchart showing information embedding processing (S6, S20) which is a subroutine of FIG. It is a flowchart which shows the information read-out process for the digital watermark performed by the image processing apparatus 10-1, 10-2 (10) of FIG. It is a flowchart which shows the information read-out process (S46) which is a subroutine of FIG.

Explanation of symbols

A ... Original image data,
B ... Image data for key,
B′—Key image data after image processing,
C: Confidential information data,
D: Image data after information embedding,
S2, S14, S44 ... image processing for feature point extraction,
S6, 20 ... Information embedding process,
S46 ... Information reading process,
10 Image processing device,
20 ... CPU,
21 ... ROM,
22 ... RAM,
22a: temporary memory,
23. Image memory,
24 ... Program memory,
30 ... Bus
31 ... Keyboard interface,
32 ... Mouse interface,
33 ... Display interface,
34 ... Printer interface,
35 ... Drive device interface,
41 ... Keyboard,
42 ... mouse,
43 ... CRT display
44 ... Printer,
45 ... CD-ROM drive device,
45a ... CD-ROM,
51. Communication cable,
61: Communication interface.

Claims (10)

An image processing method for digital watermarking for embedding confidential information data using key image data for original image data,
An image processing step of performing image processing for extracting predetermined feature points on the key image data and outputting image data for key after image processing having a plurality of feature point coordinate values;
Information for embedding the confidential information data in each coordinate of the original image data corresponding to each feature point coordinate of the key image data after the image processing using a predetermined embedding rule, and outputting the image data after embedding And an embedding step.   2. The image processing method according to claim 1, further comprising a conversion step of converting the key information data into the pseudo image data for the key and outputting the key image data as the key image data before the image processing step. An image processing method for electronic watermarking for reading confidential information data from image data after embedding confidential information data using key image data with respect to original image data,
An image processing step of performing image processing for extracting predetermined feature points on the key image data and outputting image data for key after image processing having a plurality of feature point coordinate values;
The confidential information data is read out from the coordinates of the embedded image data corresponding to the feature point coordinates of the key image data after the image processing using a predetermined reading rule, and the confidential information data after the reading is read out. An image processing method comprising: a reading step of outputting.   4. The image processing method according to claim 3, further comprising a conversion step of converting the key information data into key pseudo image data and outputting the key image data as the key image data before the image processing step. An image processing apparatus for digital watermarking that embeds confidential information data using key image data with respect to original image data,
Image processing means for performing image processing for extracting predetermined feature points on the key image data, and outputting image-processed key image data having a plurality of feature point coordinate values;
Information for embedding the confidential information data in each coordinate of the original image data corresponding to each feature point coordinate of the key image data after the image processing using a predetermined embedding rule, and outputting the image data after embedding An image processing apparatus comprising an embedding unit.   6. The image processing apparatus according to claim 5, further comprising conversion means for converting key information data into key pseudo image data and outputting the key image data as the key image data before processing by the image processing means. . An image processing apparatus for digital watermark for reading confidential information data from image data after embedding confidential information data using key image data with respect to original image data,
Image processing means for performing image processing for extracting predetermined feature points on the key image data, and outputting image-processed key image data having a plurality of feature point coordinate values;
The confidential information data is read out from the coordinates of the embedded image data corresponding to the feature point coordinates of the key image data after the image processing using a predetermined reading rule, and the confidential information data after the reading is read out. An image processing apparatus comprising: a reading means for outputting.   8. The image processing apparatus according to claim 7, further comprising conversion means for converting key information data into pseudo image data for key and outputting the key image data as the key image data before processing by the image processing means. .   An image processing program comprising the steps of the image processing method according to claim 1. A computer-readable recording medium storing the image processing program according to claim 9.

Images