JP3774631B2 - Information embedding device, information restoring device and method, computer program, and storage medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a technique for embedding information as digital watermark in digital data, extracting the embedded information, and restoring original data.
[0002]
[Prior art]
In recent years, various information such as character data, image data, and voice data has been digitized due to the rapid development and spread of computers and their networks. Digital information is not deteriorated due to aging, etc., and can be stored in perfect condition forever, but can be easily copied, and copyright protection is a big problem. Therefore, security technology for copyright protection is rapidly gaining importance.
[0003]
One technique for protecting copyright is “digital watermarking”. Digital watermarking is a technology that tracks the unauthorized use of illegal copies by embedding the name of the copyright holder and the purchaser's ID in a form that cannot be perceived by humans in digital image data, audio data, or text data. is there. Since the digital watermark may be subjected to various attacks, resistance to the attack is necessary.
[0004]
Further, the digital watermark technique originally used for the purpose of security such as copyright protection is also applied to combining the main information and the sub information in an inseparable state. If the two pieces of information are simply combined, the main information and the sub information may be separated. Furthermore, the information amount after combining is often the sum of the information amount of the main information and the information amount of the sub information. On the other hand, when two pieces of information are combined using a digital watermark, main information and sub information are not likely to be separated, and the amount of information after combining is equal to the main information.
[0005]
However, sub-information embedded in main information as a digital watermark can be extracted once embedded, but it is difficult to remove. This is a problem when the main information before embedding the sub information is required.
[0006]
In order to solve this problem, a technique for completely restoring information embedded with a digital watermark to information before the digital watermark is embedded has been proposed. Hereinafter, this technique is referred to as reversible digital watermarking.
[0007]
The reversible digital watermark will be briefly described. The reversible digital watermark is embedded in the same way as a normal digital watermark. For example, the processing is performed based on the following expression.
[0008]
I '_{j, i}= I_{j, i}+ C_jXa_iX_i        (Formula 1)
Here, j is a positive number indicating the bit position of the region and the additional information Inf, i is a positive number indicating the pixel position, I ′_{j, i}Is an image with embedded watermark, I_{i, j}Is the original image, c_jIs Inf_j+1 when In is bit 1, Inf_jA constant that is -1 when is 0, a_iIs the weighting factor, x_iIs a pseudo-random number sequence in the range of -1 to +1. x_iIs called a carrier signal for embedding a digital watermark.
[0009]
Pseudo random number sequence x_iHowever, in this method, a random number pattern to be used is determined between a device that embeds a digital watermark and a device that extracts the embedded digital watermark and restores the original image. The simplest method is to use the same random number generation algorithm on the device that embeds the digital watermark and the device that extracts and reproduces the same, and to make the initial value when generating the random number the same.
[0010]
When restoring, the embedded digital watermark is first extracted. Then, using the extracted digital watermark, the original image is restored by the following formula.
[0011]
I ''_{j, i}= I '_{j, i}-C_jXa_iX_i        (Formula 2)
Where I ''_{i, j}Is the restored image data, I '_{i, j}Is the image data input to the image restoration device, c_jIs a constant that is +1 when the bit is 1 and -1 when the bit is 0 in the additional information Inf extracted by the digital watermark extraction unit, a_i, X_iIs the same as in Equation 1.
[0012]
The digital watermark embedding shown in Equation 1 and the original image restoration shown in Equation 2 will be specifically described.
[0013]
FIG. 7 shows a specific example of the digital watermark embedding shown in Equation 1. Each matrix represents a part of the image. Also, bit 1 as bit information, that is, c_j= Represents the case of +1.
[0014]
On the restoration side, the digital watermark is first extracted. This extraction algorithm will be briefly described. The value of each pixel of the 4 × 4 pixel block of the input image I ′ and the pseudo random number sequence x_i(As described above, the correlation with the embedding side is assumed to generate the same random number sequence). If the input image I ′ and the pseudo random number sequence x_iIs high (greater than a preset threshold value), it is determined that the embedded bit is 1. The input image I ′ and −x_iWhen the correlation with (the result of inverting the sign of each element of the pseudo random number) is high, it is determined that 0 is embedded. If the correlation with any of them is low, it is determined that the digital watermark information is not embedded. In the illustrated case, the 4 × 4 pixel block has been described. However, if one bit is embedded only in one block as illustrated, high accuracy cannot be expected. It is possible to improve the overall certainty. In the case of embedding m bits, embedding 1 bit in an n pixel block may be performed m times.
[0015]
Thus, when the digital watermark is extracted, a process for removing the watermark information is performed.
[0016]
FIG. 8 shows a specific example of removing the digital watermark embedded in the example shown in FIG. Here, a case where the embedded bit information is 1 is shown. As shown in FIGS. 7 and 8, it is usually possible to completely remove the embedded digital watermark and restore the original image.
[0017]
On the other hand, FIG. 9 also shows a specific example of the digital watermark embedding shown in Equation 1. In the example shown in FIG. 9, overflow has occurred in the image data in which the digital watermark is embedded (pixel values after the digital watermark is embedded are pixels 262 and 261). In this case, the rounding process as shown in FIG. 9 is usually performed. Although the digital watermark can be correctly extracted from the image that has been subjected to the rounding process, the original image cannot be completely restored at the place where the rounding process has been performed as shown in FIG.
[0018]
[Problems to be solved by the invention]
Basically, it is possible to restore the digital data before the digital watermark is embedded from the digital data in which the digital watermark is embedded by the above-described method. As described above, the digital data before the digital watermark is embedded cannot be completely restored. Here, overflow is, for example, I_{i, j}Is represented by an integer (8 bits) from 0 to 255, after processing of Equation 1, I ′_{i, j}Is a value that falls outside this range, and is often rounded to 0 when the value is smaller than 0 and to 255 when the value is larger than 255. When this rounding process occurs, the digital watermark cannot be completely removed by the digital watermark removal.
[0019]
For this reason, it has been difficult to completely restore the original image at the location where overflow occurs.
[0020]
The present invention has been made in view of the above-described conventional example, and intends to provide an information embedding device, an information restoration device and method, a computer program, and a storage medium that can completely reproduce original digital data. Is.
[0021]
[Means for Solving the Problems]
  In order to achieve the above object, an information embedding device of the present invention comprises the following arrangement. That is,
  Add / subtract values to each element of digital dataAnd by embedding 1-bit information using a plurality of elements constituting the digital data,An information embedding device for embedding additional information,
  An element with a value that exceeds the possible range of the element when the addition or subtraction is performedPosition ofDetecting means for detecting
  Detected by the additional information and the detection meansInformation on the position of an element whose value exceeds the possible range of the element, andGenerating means for generating as embedded information,
  When embedding in the digital data, an element exceeding the range that the element can take by the addition / subtractionpositionIs excluded from embedding, and is within the range of elementspositionAnd embedding means for embedding the real embedding information generated by the generating means as a digital watermark.
[0022]
  The apparatus on the side of restoring the digital data embedded by the embedding apparatus has the following configuration. That is,
  Add / subtract values to each element of digital dataAnd embedding 1-bit information using a plurality of elements constituting the digital data.An information restoration apparatus for inputting digital data not embedded for an element in which information is embedded and exceeding the range that the element can take by addition and subtraction, and restoring the original digital data,
  Digital watermark extraction means for extracting information embedded in the input digital data;
  Of the extracted information, theThe position of the element that is not to be embeddedA digital watermark removing unit that removes the embedded information for the element to be embedded based on the information and restores the original digital data;
[0023]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings.
[0024]
[Digital watermark embedding]
First, a digital watermark embedding apparatus according to this embodiment will be described. The digital watermark embedding device in the embodiment embeds additional information so that it can be completely restored to the image data before embedding the digital watermark.
[0025]
FIG. 2 is a diagram showing an internal configuration of the digital watermark embedding apparatus. The process flow of the digital watermark embedding apparatus will be described with reference to FIG.
[0026]
An image I is input to the digital watermark embedding apparatus. In order to simplify the description here, it is assumed that the image I is grayscale multivalued image data expressed with an accuracy of 8 bits per pixel. However, the present invention is not limited to this, and may be grayscale multivalued image data having a predetermined number of bits. Furthermore, when a color image composed of a plurality of components is input, one or a plurality of components may be selected as an input image. The input image I is input to the overflow area detection unit 201 and the digital watermark embedding unit 203.
[0027]
First, the overflow area detection unit 201 will be described. The overflow area detection unit 201 receives the image I, and detects whether or not an overflow occurs after the digital watermark embedding process by the subsequent digital watermark embedding unit 203 for all the pixels constituting the input image I. All the pixel positions where overflow occurs are extracted, and the coordinate information of the pixel group is collectively output.
[0028]
Hereinafter, the information on the position of the pixel group (which may be 0, 1 or more) where overflow occurs is referred to as overflow information R. The overflow information R is output to the subsequent encoding unit 202 and digital watermark embedding unit 203. The detailed operation of the overflow area detection unit will be described later.
[0029]
Next, the encoding unit 202 will be described. Overflow information R and additional information Inf are input to encoding section 202, overflow information R and additional information Inf are combined as one code string, and a combined code string w is output. For example, w is configured as shown in FIG. Note that w is a code string embedded as a digital watermark in the image I in the digital watermark watermark embedding unit 203 in the subsequent stage. For this reason, w may be compression-encoded to perform efficient digital watermark embedding. Further, when the additional information Inf is secret information, w may be encrypted. Furthermore, w may be error-correction-encoded so that the additional information Inf can be correctly extracted even when the additional information Inf is embedded as a digital watermark and then modified into different information by various attacks. . In any case, the pixel position indicated by the overflow information R is not to be embedded, and the pixel value of the original image is used as it is.
[0030]
Next, the digital watermark embedding unit 203 will be described. The digital watermark embedding unit 203 receives the image I, the overflow information R, and the code string w. The code string w is embedded in the image I as a digital watermark, and an image I ′ in which the digital watermark is embedded is output. The Here, the digital watermark is not embedded in the pixel indicated by the overflow information R. Therefore, the digital watermark embedding unit 203 does not overflow due to the digital watermark embedding. Detailed operation of the digital watermark embedding unit 203 will be described later.
[0031]
As described above, the additional information Inf is embedded in the image I as a digital watermark, and an image I ′ in which the digital watermark is embedded is generated.
[0032]
[Image restoration device]
Next, the image restoration apparatus in the present embodiment will be described with reference to FIG. In the image restoration apparatus, a process for completely restoring an image in which a digital watermark is embedded into image data before the digital watermark is embedded is performed.
[0033]
The image restoration apparatus receives an image I ′ embedded with a digital watermark by the above-described method. The input image I ′ is input to the subsequent digital watermark extraction unit 301 and digital watermark removal unit 303.
[0034]
First, the digital watermark extraction unit 301 will be described. The digital watermark extraction unit 301 receives the image I ′ embedded with the digital watermark, extracts the embedded code string w, and outputs the extracted code string w. In the digital watermark extraction unit 301, the digital watermark is extracted using all the target pixels in the input image I ′. That is, the digital watermark including the overflow information R (the digital watermark is not embedded at the position described in the overflow information R) that has not been embedded in the digital watermark embedding unit 203 described above. Target of extraction processing. Detailed operation of the digital watermark extraction unit 301 will be described later. The extracted code string w is output to the subsequent decoding unit 302.
[0035]
Next, the decoding unit 302 will be described. The decoding unit 302 receives the code string w extracted in the previous stage, is divided into overflow information R and additional information Inf that constitute the code string w, and outputs the divided overflow information R and additional information Inf. Here, when w is compression-encoded in the encoding unit 202 described above, decompression decoding processing is executed. Further, when w is encrypted in the encoding unit 202 described above, a decryption / decryption process for decrypting the code is executed. Further, when w is error correction encoded in the encoding unit 202 described above, error correction decoding is performed to correct an error when an error exists. The overflow information R and the additional information Inf are output to the subsequent digital watermark removal unit 303.
[0036]
Next, the digital watermark removal unit 303 will be described. The digital watermark removal unit 303 receives the image I ′ embedded with the digital watermark, the overflow information R, and the additional information Inf, and embeds the digital watermark from the image I ′ embedded with the digital watermark using the additional information Inf. The image before being restored is restored, and the restored image I is output. Here, the pixel indicated in the overflow information R is not subject to restoration processing. This is because the digital watermark is not embedded in the digital watermark embedding unit 203 described above for the pixel indicated in the overflow information R. The detailed operation of the digital watermark removal unit 303 will be described later.
[0037]
As described above, the image I before the digital watermark is embedded is restored from the image I ′ embedded with the digital watermark.
[0038]
[Details of digital watermark embedding process]
Here, details of the digital watermark embedding process will be described with a specific example.
[0039]
Consider a case where n-bit information Inf is embedded in image I (note that information w is embedded when associated in the above embodiment). In this case, the image I is divided into n non-overlapping regions I._jDivide into (j = 1, 2,... n). And as a digital watermark embedding process,
I '_{j, i}= I_{j, i}+ C_jXa_iX_i      (Formula 3)
Is executed. Here, j is a positive number that specifies the region and the bit position of the additional information Inf, i is a positive number that indicates the pixel position, and I ′_{j, i}Is an image with an embedded watermark, c_jIs Inf_jIs a constant that is +1 when bit 1 and -1 when Infj is bit 0, a_iIs the weighting factor, x_iIs a pseudo-random number sequence in the range of -1 to +1. x_iIs called a carrier signal for embedding a digital watermark.
[0040]
Here, the maximum value (positive value) that the weighting factor can take is a_max, X_iThe maximum value (positive value) that can be taken by x_maxThen, the calculation shown in Equation 3 is
a_maxX_max≦ I ≦ 255-a_maxX_max      (Formula 4)
It is executed only when the conditions shown in
[0041]
However, in both the device on the embedding side and the device on the extracting side, a_i, X_iSince the values of the elements of are synchronized and known to each other, the above equation 4 is further developed,
a_iX_i≦ I ≦ 255-a_iX_i      (Formula 4 ')
It doesn't matter.
[0042]
If the condition shown in Expression 4 (or Expression 4 ') is not satisfied, the pixel position i is recorded as overflow information R. In a general natural image, the number of pixels that do not satisfy the condition shown in Expression 4 (pixels that are not to be embedded) is much smaller than the total number of pixels in the entire image. This indicates that the overflow information R is a relatively small amount of information. Therefore, there is no problem even if the overflow information R is combined as a code string w in the subsequent encoding unit 202 and then embedded as a digital watermark in the digital watermark embedding unit 203. As described above, it is possible to execute the digital watermark embedding process without overflowing by embedding the digital watermark according to Formula 1 under the condition shown in Formula 4 (or Formula 4 ').
[0043]
[Details of image restoration processing]
Next, details of the image restoration processing will be described with a specific example.
[0044]
A code string w is embedded as a digital watermark in the input image I ′. In order to extract the digital watermark w embedded using Equation 3, the PFA is calculated from the carrier signal x for embedding the digital watermark and the image I ′ input to the digital watermark extraction unit. It is determined whether or not the bit is embedded and, if embedded, whether the bit is 0 or 1.
[0045]
Here, PFA is the probability of determining that a digital watermark is embedded when the digital watermark is not embedded. In order to calculate this probability, a method called process test is used. Process testing is a technique known by experts in this field. For example, “A method for signature casting on digital images” by I.Pitas, I.C.I.I.P. Proceedings, pages 215 to 218, September 1996 is detailed as an example of using a process test for extracting a digital watermark. Test value q for each bit by process test_jIs calculated. The test value q_jFollows a standard normal distribution with a mean of zero and variance of 1 when digital watermark is not embedded in digital data, but with a non-zero mean and variance of 1 when digital watermark is embedded in digital data. Follow a normal distribution. Whether the digital watermark is embedded or not is determined by the calculated test value q_jIs determined by how far is 0.
[0046]
Further, the calculated test value q_jIt is possible to calculate information embedded as a digital watermark. If embedded using Equation 3, q_jIt is determined that the bit is 1 when is positive, while the bit is 0 when qj is negative.
[0047]
For the principle of extracting a digital watermark, refer to the contents described in the prior art.
[0048]
The digital watermark extraction process described above is executed including the pixels in which the digital watermark is not embedded because the above-described digital watermark embedding unit 203 does not satisfy the condition shown in Expression 4 (or Expression 4 '). Since the number of pixels that do not satisfy the condition shown in Equation 4 (or Equation 4 ′) is smaller than the total number of pixels shown in the entire image, the digital watermark is correctly extracted even if the digital watermark is extracted including this pixel. It is possible to extract.
[0049]
The digital watermark extracted as described above is a code string w, which is composed of overflow information R and additional information Inf, and is divided into each. Thereafter, the following processing is executed in the digital watermark removing unit.
[0050]
If i ⊂ R then I '' = I '_{j, i}
else I '' = I '_{j, i}-C_jXa_iX_i        (Formula 5)
Where I ''_{i, j}Is the restored image data, I '_{i, j}Is the image data input to the image restoration device, c_jIs a constant that is +1 when the bit is 1 and -1 when the bit is 0 in the additional information Inf extracted by the digital watermark extraction unit, a_i, X_iIs the same as in Equation 3. That is, when i is not included in the overflow information R, the process opposite to the expression 3 is executed, and when i is included in the overflow area R, the process is not executed.
[0051]
When the code string w is correctly extracted by the above processing, the restored image I ″ can be made equal to the original image I.
[0052]
[Description of specific equipment]
As described above, it is easy for those skilled in the art that the apparatus for embedding the digital watermark and the apparatus for extracting the embedded electronic data and restoring the original image in the embodiment can be realized by a program that executes the above-described processing. I can think of it.
[0053]
Therefore, a specific configuration of the apparatus is shown in FIG. 1, and thereafter, processing (program) for embedding a digital watermark and restoration processing (program) will be described.
[0054]
In FIG. 1, a host computer 101 is, for example, a generally popular personal computer.
[0055]
Inside the host computer 101, blocks to be described later are connected by a bus 107, and various data can be transferred.
[0056]
In the figure, reference numeral 102 denotes a monitor (display device), and reference numeral 103 denotes a CPU capable of controlling the operation of each internal block or executing a program stored therein. A ROM 104 stores a BIOS and a boot program. Reference numeral 105 denotes a RAM used for storing a program processed by the CPU 103 and storing image data to be processed. Reference numeral 106 denotes a hard disk (HD) capable of storing in advance a program (OS, application, etc.) and image data to be transferred to the RAM 105, etc., and storing processed image data.
[0057]
Reference numeral 108 denotes a CD drive capable of reading or writing data stored in a CD (CD-R) which is one of external storage media. Reference numeral 109 denotes an FD drive capable of reading from and writing to the FD as in the case of 108. Reference numeral 110 denotes a DVD drive that can read from and write to a DVD as in the case of 108. When a program for image editing or a printer driver is stored on a CD, FD, DVD or the like, these programs are installed on the HD 106 and transferred to the RAM 105 as necessary. . Reference numeral 113 denotes an interface (I / F) connected to the keyboard 111 or the mouse 112 in order to accept input instructions. A communication interface 114 connects to a network such as the Internet.
[0058]
The output destination of the image incorporating the digital watermark may be a storage medium such as HD or FD, or a file server on the network. On the contrary, the same applies to the case of inputting information embedded with a digital watermark.
[0059]
When the apparatus functions as a digital watermark embedding apparatus, processing is performed according to the procedure shown in FIG. Note that a program according to the procedure described below is installed in the HD 106. In addition, although the information Inf to be embedded is input from a keyboard or the like, information stored in advance in the HD 106, the ROM 104, the RAM 105, or the like may be used.
[0060]
First, in step S1101, image data in which a digital watermark is embedded is input. The input source is not limited to a storage medium, network download, image scanner, or the like. In step S1102, all overflow pixel positions are detected from the input image data, and the detected information is stored in the RAM 105 as overflow information R. Next, the pre-input additional information Inf and the overflow information R are synthesized and encoded (lossless encoding) to create information w to be actually embedded, and temporarily stored in the RAM 105.
[0061]
Next, the input image data is focused on for each pixel, and it is determined whether or not the focused pixel position is included in the overflow information R. If it is not included in the overflow information R, one bit is extracted from the information w, and a digital watermark embedding process corresponding to the bit state is executed.
[0062]
On the other hand, if the pixel is included in the overflow information R, the digital watermark embedding process is not executed.
[0063]
This process is executed for all bit information and all pixels (steps S1106 and 1107). After all bit information and all pixels are executed, the image data in which the digital watermark is embedded is output to the RAM 105 or HD 106 (step S1108).
[0064]
In addition, before embedding all pixels, when embedding of all bits of information w is completed, dummy information may be embedded or information w may be embedded repeatedly.
[0065]
As a result, the embedded image information I ′ is generated on the RAM. This image I ′ is finally stored as a file on the HD or the like, but the subsequent processing may be transferred via a network or stored in a removable medium.
[0066]
Next, processing for extracting a digital watermark and restoring an original image will be described.
[0067]
First, image data I ′ in which a digital watermark is embedded is input to the RAM 105 (step S1201). The input source may be any as described above. Thereafter, in step S1202, digital watermark information is extracted from the embedded image I '. The principle of digital watermark extraction is as described above.
[0068]
Next, the information extracted in step S1203 is decoded and stored in the RAM 105 as information w (= overflow information R + embedded information Inf), and the extraction process is completed. The embedded information Inf may be displayed on the display screen.
[0069]
In step S1204, the image data is focused on for each pixel, and it is determined whether or not the focused pixel position is included in the overflow information R.
[0070]
If it is not included in the overflow information R, digital watermark removal processing is executed in step S1205.
[0071]
On the other hand, when an overflowing pixel is included, the digital watermark removal process is not executed for the pixel, and the pixel value of the input image data is used as it is.
[0072]
This process is executed for all pixels (step S1206). Thus, after being executed for all pixels, the original image is reproduced by removing the digital watermark. This result (original image data) is output to the RAM 105, the HD 106, etc., and displayed depending on the case.
[0073]
<Second Embodiment>
In the above embodiment (first embodiment), the case has been described where the number of pixels in which an overflow phenomenon occurs is relatively small with respect to the number of pixels of the entire image. That is, the case where the information amount of the overflow information R is relatively small and is a relatively desirable information amount for embedding as a digital watermark has been described. This is the case of natural images such as photographs.
[0074]
On the other hand, depending on the image data, the amount of overflow information R may be large. For example, when embedding a digital watermark on a luminance value, the overflow information R becomes large in the case of an overall bright image or an overall dark image.
[0075]
When the information amount of the overflow information R becomes too large, the information amount of the code string w including the overflow information R also increases, and the code string w cannot be embedded as a digital watermark. Therefore, the information amount of the overflow information R is desirably as small as possible. Here, an example of processing for reducing the overflow information R will be described as a second embodiment.
[0076]
FIG. 5 is a diagram showing an internal configuration of the digital watermark embedding apparatus according to the second embodiment. The difference from the apparatus shown in FIG. 2 is that the correction unit 501 is added and the processing in the encoding unit 503. Since other operations are the same as those in FIG.
[0077]
Image data I is input to the correction unit 501, correction processing is performed to reduce the amount of overflow information R, and correction indicating the corrected image data I ″ ′ and what correction is performed Information C is output.
[0078]
The correction processing is processing for reducing the information amount of the overflow information R. For example, when the entire image is bright, processing for reducing all pixel values by a fixed amount (d) is executed. In this case, as the correction information C, information “the pixel value of the entire image is reduced by a certain amount (d)” is output.
[0079]
Further, in the encoding unit 503, in addition to the overflow information R and the additional information Inf, the correction information C is also information to be encoded. The encoded w has correction information C, overflow information R, and additional information Inf as shown in FIG.
[0080]
A digital watermark is embedded by the above processing. Next, the image restoration apparatus in this embodiment will be described.
[0081]
FIG. 6 is a diagram showing an internal configuration of the image restoration apparatus in the second embodiment. The difference from the apparatus shown in FIG. 3 is the addition of the reverse correction unit 604 and the processing in the decoding unit 602. Since other operations are the same as those in FIG. 6 and FIG.
[0082]
The decoding unit 602 receives the code string w extracted by the preceding digital watermark extraction unit 601, decodes it into additional information Inf, overflow information R, and correction information C, and outputs them.
[0083]
Next, the reverse correction unit 604 receives the image data I ′ ″ from which the digital watermark has been removed by the previous digital watermark removal unit and the correction information C decoded by the decoding unit 602. The process opposite to the process described above is executed, and the corrected image data I is output. For example, when information indicating that “the pixel value of the entire image has been reduced by a certain amount (d)” is input as the correction information C, the correction unit 604 performs a process of adding a certain amount (d) of the pixel value of the entire image. Is done.
[0084]
Through the above processing, the image data I ′ input to the image restoration apparatus can be completely restored to the image data I before the digital watermark is embedded.
[0085]
In addition, although it is the determination method of the said correction information C, if an example is shown, it will be as follows.
[0086]
The luminance distribution of the embedded image is examined. If the frequency of high luminance is higher than that on the low luminance side, for example, correction is performed so that the frequency shifts to the low luminance side. On the contrary, if the frequency of low luminance is higher than that on the high luminance side, correction is performed so that the low luminance side is shifted to the low luminance side.
[0087]
In some cases, the following may be used. In order to make the explanation easy to understand, consider a case where the paper on which the text is printed is read by an image scanner (8 bits per pixel) and the read image is an image I.
[0088]
Normally, when a document (binary image) in which text is described is read and each pixel is expressed as a luminance value, the pixel distribution is concentrated at the highest luminance and the lowest luminance as shown in FIG. There are almost no pixels.
[0089]
In such a situation, if a certain value is uniformly added or subtracted for all the pixels, it will overflow on the high luminance or low luminance side, which is not preferable. Therefore, the following is performed.
[0090]
First, in the case shown in the drawing, the range that can be taken by the low luminance is distributed in luminance values of 0 to a. Therefore, Expression 4 (or Expression 4) in the range in which the frequency of intermediate luminance is “0”. Search for places that satisfy ') and replace them with their brightness values. For example, if a luminance value of 50 to 50 + a can be found, the luminance value of 0 to a is converted to a value of 50 to 50 + a (0 is converted to 50, 1 is converted to 51...). The same conversion is performed for the range where high luminance can be obtained.
[0091]
As a result, particularly in the case of a binary image such as a sentence, as the correction information C, what value the luminance 0 to a has been converted to, and what value the luminance b to 255 has been converted to. If it has information indicating whether it has been converted, it can be completely restored. In the above description, the image is expressed by the luminance component. However, the present invention is not limited to this. Of course, the image may be expressed by other components such as density.
[0092]
As described above, according to the first and second embodiments, overflowing pixels (elements of digital data) are excluded from embedding, and information that is not overflowed is embedded with a digital watermark so that the original image Data can be completely restored.
[0093]
As described above, the present invention can be realized by a program executed by a computer. Accordingly, the present invention naturally includes a computer program. Since the present invention can be configured by supplying a program to a computer, means for supplying the program code to the computer, specifically, a storage medium storing the program code is also included in the scope of the present invention.
[0094]
As a storage medium for storing such a program code, for example, a floppy disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.
[0095]
Further, the computer controls various devices according to only the supplied program code, so that not only the functions of the above embodiments are realized, but also the OS (operating system) on which the program code is running on the computer. Such program code is also included in the scope of the present invention when the above embodiment is realized in cooperation with a system) or other application software.
[0096]
Further, after the supplied program code is stored in the memory of the function expansion board of the computer or the function expansion unit connected to the computer, the program code is stored in the function expansion board or function storage unit based on the instruction of the program code. A case where the CPU or the like provided performs part or all of the actual processing and the above-described embodiment is realized by the processing is also included in the scope of the present invention.
[0097]
【The invention's effect】
As described above, according to the present invention, it is possible to embed a digital watermark and to reproduce the original digital data that was the object to be embedded.
[Brief description of the drawings]
FIG. 1 is a block diagram of an information processing apparatus to which an embodiment is applied.
FIG. 2 is a block diagram of a digital watermark embedding apparatus according to the first embodiment.
FIG. 3 is a block diagram of an image restoration apparatus according to the first embodiment.
FIG. 4 is a diagram illustrating an example of a code string w in the first embodiment.
FIG. 5 is a block diagram of a digital watermark embedding apparatus according to a second embodiment.
FIG. 6 is a block diagram of an image restoration apparatus according to a second embodiment.
FIG. 7 is a diagram illustrating details of processing for embedding a digital watermark.
FIG. 8 is a diagram showing the contents of processing for restoring original data.
FIG. 9 is a diagram illustrating an example of overflow when embedding a digital watermark.
FIG. 10 is a diagram illustrating an example in which original data cannot be restored when an overflow occurs.
FIG. 11 is a flowchart showing a digital watermark embedding procedure in the embodiment.
FIG. 12 is a flowchart illustrating a procedure for restoring original data according to the embodiment.
FIG. 13 is a diagram illustrating an example of a code string w according to the second embodiment.
FIG. 14 is a diagram illustrating an example of a pixel distribution for explaining a correction process in the second embodiment.

Claims (11)

An information embedding device for embedding additional information by adding / subtracting values to / from each element constituting digital data and embedding 1-bit information using a plurality of elements constituting the digital data ,
Detecting means for detecting a position of an element having a value exceeding a possible range of the element when the addition or subtraction is performed;
Generating means for generating, as actual embedded information, the additional information and information on the position of an element having a value exceeding the possible range of the element detected by the detecting means;
When embedding in the digital data, element positions exceeding the range that the element can take by the addition / subtraction are excluded from embedding, and the actual embedding information generated by the generating means for the element position within the range that the element can take is electronic An information embedding device comprising: an embedding unit embedded as a watermark. The digital data is image data,
The information embedding apparatus according to claim 1, wherein the detection unit detects a pixel position exceeding a possible range of pixel values when the addition and subtraction are performed.   Furthermore, it has an encoding means for compressing and encoding at least one of the additional information and the information detected by the detecting means, and the generating means generates actual embedded information based on the encoding result by the encoding means The information embedding device according to claim 1, wherein the information embedding device is an information embedding device.   Furthermore, it has an encryption means for encrypting at least one of the additional information and the information detected by the detection means, and the generation means generates actual embedded information based on an encryption result by the encoding means. The information embedding apparatus according to claim 1, wherein the information embedding apparatus is an information embedding apparatus.   And an encoding unit that performs error correction encoding on at least one of the additional information and the information detected by the detection unit, and the generation unit converts the actual embedded information based on a result of encryption by the encoding unit. The information embedding apparatus according to claim 1, wherein the information embedding apparatus generates the information embedding apparatus. The digital data further comprises correction means for correcting the digital data in order to reduce the number of elements having values exceeding the possible range of the elements when the addition and subtraction are performed.
The information embedding apparatus according to claim 1, wherein the generation unit generates actual embedding information by adding information indicating details of correction performed by the correction unit. Information is embedded by adding / subtracting a value to / from each element constituting digital data and embedding 1-bit information using a plurality of elements constituting the digital data , and the element can be taken by addition / subtraction An information restoration device that inputs unembedded digital data for elements that exceed the range and restores the original digital data,
Digital watermark extraction means for extracting information embedded in the input digital data;
Among the extracted information, digital watermark removal that removes the embedded information for the element to be embedded and restores the original digital data based on the position information of the element that is not to be embedded An information restoration apparatus comprising: means. An information embedding device for embedding additional information by adding / subtracting values to / from each element constituting digital data and embedding 1-bit information using a plurality of elements constituting the digital data ,
A detection step of detecting a position of an element having a value exceeding a possible range of the element when the addition or subtraction is performed;
A generation step of generating, as actual embedding information , information on the position of an element having a value that exceeds a range that can be taken by the element detected in the detection step;
When embedding in the digital data, element positions exceeding the range that the element can take by the addition and subtraction are excluded from embedding, and the actual embedding information generated in the generation step for the element positions within the range that the element can take is electronic An information processing apparatus control method comprising: an embedding step of embedding as a watermark. Information is embedded by adding / subtracting a value to / from each element constituting digital data and embedding 1-bit information using a plurality of elements constituting the digital data , and the element can be taken by addition / subtraction An information restoration device that inputs unembedded digital data for elements that exceed the range and restores the original digital data,
A digital watermark extraction step of extracting information embedded in the input digital data;
Among the extracted information, digital watermark removal that removes the embedded information for the element to be embedded and restores the original digital data based on the position information of the element that is not to be embedded And a method for controlling the information restoration apparatus. A computer program for causing a computer to realize the function of the information embedding device according to any one of claims 1 to 6 or the function of the information restoration device according to claim 7. A storage medium storing the computer program according to claim 9.

Images