JP2006074338A - Data concealing device, concealed data extraction device, data concealing method, concealed data extraction method, and program for them - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data concealing device capable of enhancing the security of concealed information embedded in an image. <P>SOLUTION: The data concealing device embeds a concealing image 720 to each least significant bit of a medium image 710 and references a key image to apply prediction coding to a medium image 710' to which the concealing image 720 is embedded. The key image is an image to each least significant bit of which a noise corresponding to a password is arranged, and each least significant bit only of the medium image 710' is scrambled. Thus, although a medium image 710" closely resembling the original medium image 710' is generated when coded data are decoded without referencing the key image, each least significant bit of the medium image 710" indicates the scrambled concealing image 720. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a data concealment device that embeds confidential information in a predetermined bit position of image data.

For example, Non-Patent Document 1 discloses a method for embedding an image B to be concealed with limited gradation and embedding the image B with limited gradation on the LSB (Least Significant Bit) side of the image A to be embedded. In this case, since the image A ′ in which the image B is embedded is common to the original image A and the MSB (Most Significant Bit) side, it is output as almost the same output image.
RZWang, CFLin, and JCLin, "Image hiding by optimal LSB substitution and genetic algorithm," Pattern Recognition, 34 (2001), pp.671-693, 2001.

  The present invention has been made from the above-described background, and an object thereof is to provide a data concealment device that conceals information at a predetermined bit position of image data.

[Data concealment device]
In order to achieve the above object, a data concealment device according to the present invention is a data concealment device for concealing concealment data, and embedding means for embedding concealment data in image data, and concealment data is transmitted by the embedding unit. Compared with the embedded image data and a predetermined key image, reference information generating means for generating reference information for designating a reference to the key image, code data of the reference information generated by the reference information means, Code generating means for generating a part of the code data of the image data in which the confidential data is embedded.

  Preferably, the embedding unit embeds the secret data in the image data by replacing a predetermined bit position of the image data with the secret data, and the reference information generating unit has a non-uniform value for the bit position. The constructed key image is compared with the image data in which the confidential data is embedded, and reference information for the key image is generated.

  Preferably, the reference information generation unit compares the gradation data of the target pixel included in the image data in which the confidential data is embedded with the gradation data of the pixel on the key image corresponding to the target pixel, The difference value of the gradation data is generated as the reference information, and the code generation unit generates the code data of the generated difference value as code data of image data in which confidential data is embedded.

  Preferably, the apparatus further includes key image generation means for generating a data string composed of non-uniform values based on a password, and arranging the generated data string at the bit position to generate the key image. The reference information generation means generates reference information for the key image using the key image generated by the key image generation means.

  Preferably, the image data in which the secret data is embedded is the key image, the embedding unit embeds the secret data in the key image, and the reference information generation unit includes the key image in which the secret data is embedded; Compared with a key image in which confidential data is not embedded, generates reference information for a key image in which confidential data is not embedded, and the code generation means converts the generated reference information code data into secret data Generated as part of the code data of the embedded key image.

  Further, the data concealment device according to the present invention is a data concealment device for concealing secret data, and target data generating means for generating encoding target data based on predetermined image data and secret data And reference information generating means for comparing the encoding target data generated by the target data generating means with the image data and generating reference information for designating a reference to the image data, and the reference information means Code generating means for generating the code data of the generated reference information as a part of the code data of the encoding target data;

[Confidential data extraction device]
The secret data extraction apparatus according to the present invention is a secret data extraction apparatus that extracts secret data from code data, and extracts reference data that extracts image data included in a key image based on the input code data. Means, extracted image generation means for generating a decoded image using the image data extracted by the reference data extraction means, and extraction of secret data from a predetermined bit position of the decoded image generated by the decoded image generation means And a bit extracting means.

[Data concealment method]
The data concealment method according to the present invention is a data concealment method for concealing concealment data, wherein the concealment data is embedded in the image data, the image data in which the concealment data is embedded, and a predetermined key image. In comparison, reference information designating reference to the key image is generated, and the code data of the generated reference information is generated as part of the code data of the image data in which the confidential data is embedded.

  The data concealment method according to the present invention is a data concealment method for concealing concealment data, which is generated by generating encoding target data based on predetermined image data and concealment data. Compare the encoding target data with the image data, generate reference information for designating a reference to the image data, and generate the reference data of the generated reference information as a part of the encoding data of the encoding target data Generate as

[Confidential data extraction method]
The secret data extraction method according to the present invention is a secret data extraction method for extracting secret data from code data, and extracts and extracts image data included in a key image based on the input code data. The decoded image is generated using the image data, and the secret data is extracted from the predetermined bit position of the generated decoded image.

[program]
Further, the program according to the present invention compares the step of embedding the secret data in the image data, the image data in which the secret data is embedded, and the default key image in the data concealment device that conceals the secret data. Generating the reference information for designating a reference to the key image, and generating the code data of the generated reference information as a part of the code data of the image data in which the secret data is embedded. To make it run.

  Further, the program according to the present invention includes a step of generating encoding target data based on predetermined image data and confidential data in a data concealment device for concealing confidential data, and the generated encoding target Comparing the data with the image data, generating reference information for designating a reference to the image data, and generating the code data of the generated reference information as part of the code data of the encoding target data And generating the data concealment device.

  Further, the program according to the present invention includes a step of extracting image data included in a key image based on input code data in a secret data extraction apparatus that extracts secret data from the code data, and the extracted image data The secret data extraction device is caused to execute a step of generating a decoded image and a step of extracting secret data from a predetermined bit position of the generated decoded image.

  According to the data concealment device of the present invention, the confidentiality of confidential information embedded in an image can be enhanced.

[Background and outline]
First, in order to help understanding of the present invention, its background and outline will be described.
FIG. 1 is a diagram for explaining embedding of confidential information in image data.
As illustrated in FIG. 1A, multi-valued image data has gradation data (pixel value or the like) at each pixel. Since the pixel value is expressed by a bit string, even if a lower bit such as LSB changes, the pixel value remains small. For example, when the pixel value of each pixel is expressed by 8 bits, even if the least significant bit changes, each pixel changes only one of 256 gradations.
A technique for embedding confidential information in lower bits of image data using such characteristics of image data is known. In the following description, image data in which confidential information is embedded is referred to as a medium image, and data in which bit patterns of confidential information (binary, images, etc.) are arranged on a plane is referred to as a confidential image.
For example, the image processing apparatus embeds the confidential image 720 in the medium image 710 by replacing the least significant bit of the medium image 710 (8 bits) illustrated in FIG. 1A with the bit pattern of the confidential image 720. Can do. In this case, the medium image 710 ′ in which the secret image 720 is embedded is obtained by changing the least significant bit of the original medium image 710, and the appearance of the image is hardly changed.
However, if the LSB of the medium image 710 is simply replaced with the secret image 720 as described above, the secret image 720 may be easily extracted.

Therefore, the image processing apparatus according to the present invention encodes the medium image 710 ′ by predictive encoding processing that refers to the key image. This key image is image data having a non-uniform bit pattern at the bit position in which the secret image 720 is embedded (in other words, in the bit position other than the bit position in which the secret information is embedded, it has a uniform bit pattern). is there. As a result, the lower bits of the medium image 710 ′ are scrambled and cannot be normally decrypted without the key image.
For example, when the image processing apparatus predictively encodes the medium image 710 ′ with reference to a key image including a pixel pattern having pixel values of 0 and 1, the image processing apparatus scrambles only the least significant bit of the medium image 710 ′. It will be over. In this case, when the code data is decoded without referring to the key image having the pixel pattern, the secret image 720 corresponding to the least significant bit of the medium image 710 ″ is illustrated in FIG. Thus, the scrambled state is obtained. The code data of the medium image 710 ″ is decoded with reference to a reference image limited to the least significant bit (for example, an image whose pixel values are all 0). Therefore, the component corresponding to the upper bits is normally decoded, and as illustrated in FIG. 1C, a medium image 710 ″ that approximates the medium image 710 ′ before encoding is obtained as a decoded image. .
On the other hand, when the encoded data is decoded with reference to the key image, a medium image 710 ′ is obtained as a decoded image, and the image processing apparatus uses a secret bit image from the predetermined bit position of the decoded image (medium image 710 ′). 720 can be extracted.

FIG. 2 is a diagram for explaining the outline of the predictive encoding process that refers to the key image. FIG. 2A illustrates a reference position referred to in the predictive encoding process, and FIG. The code data produced | generated by the prediction encoding process are illustrated typically.
When the medium image 710 ′ (FIG. 1B) to be encoded is encoded, the image processing apparatus generates prediction data with reference to the key image, and uses the generated prediction data to generate a prediction code Process. That is, the image processing apparatus encodes reference information for the key image as at least a part of code data of the medium image 710 ′. More specifically, as illustrated in FIG. 2A, the image processing apparatus performs gradation data (pixel value) of the pixel of interest X on the medium image 710 ′ (with the confidential image 720 embedded). Is encoded with reference to the reference pixel E set on the key image together with a plurality of reference pixels A to D set on the medium image 710 ′, When the pixel value of the reference pixel matches, the identification information (reference information) of the matched reference pixel is encoded, and when the pixel value of the target pixel X and the pixel value of the reference pixel do not match, the reference pixel A difference value (hereinafter referred to as a prediction error) between the pixel value of E and the pixel value of the target pixel X is encoded.

As illustrated in FIG. 2B, the code data generated in this way includes a code corresponding to each reference pixel and a prediction error (difference value between the target pixel X and the reference pixel E). For example, “symbol A” in the figure is a code generated when the pixel value of the target pixel X and the pixel value of the reference pixel A match (when the prediction of the reference position A is correct), and decoding is performed. In some cases, the data indicates the application of the pixel value of the reference pixel A.
Similarly, the symbol E is a symbol generated when the pixel value of the target pixel X matches the pixel value of the reference pixel E. At the time of decoding, the reference pixel E is referred to by referring to the reference pixel E on the key image. Instruct to apply the pixel value of E. Therefore, the pixel corresponding to the symbol E cannot be normally reproduced unless the reference pixel E on the key image can be referred to.
The “prediction error” in the figure is a code generated when the pixel value of the pixel of interest X and the pixel value of the reference pixel do not match, and at the time of decoding, the pixel value of the reference pixel E Used as a difference value from the pixel value of the target pixel X. Therefore, the pixel corresponding to the sign of the prediction error cannot be normally reproduced unless the reference pixel E on the key image can be referred to.

Therefore, when the code data illustrated in FIG. 2B is decoded without referring to the key image used at the time of encoding, pixels corresponding to “code E” and “prediction error” are normally reproduced. Instead, a scrambled decoded image is generated. Here, since the key image of this example has only a pixel value of 0 or 1, the original pixel value of the pixel encoded as “Code E” is 0 or 1. Therefore, even when the key image used at the time of encoding cannot be referred to, the error between before encoding and after decoding is at most 1. The “prediction error” is a difference value between the pixel value (0 or 1) of the reference pixel E on the key image and the pixel value of the target pixel X, and even if the key image used at the time of encoding cannot be referred to. However, if decoding is performed with reference to a reference image limited to the least significant bit (for example, an image having all pixel values of 0), the error before and after encoding is at most one. Therefore, in the pixels corresponding to “code E” and “prediction error”, only the least significant bit is scrambled, and the component corresponding to the upper bit is normally decoded.
On the other hand, when decoding is performed with reference to the key image used at the time of encoding, the pixels corresponding to “code E” and “prediction error” are accurately reproduced from the most significant bit to the least significant bit. An image 710 ′ is reproduced.
As described above, the image processing apparatus according to the present embodiment corresponds to the secret image 720 by predicting and encoding the medium image 710 ′ in which the secret image 720 is embedded with reference to the key image in which the bit position is limited. The bit plane is scrambled to improve the security level for the secret image 720.

[Hardware configuration]
Next, the hardware configuration of the image processing apparatus 2 will be described.
FIG. 3 is a diagram illustrating the hardware configuration of the image processing apparatus 2 to which the data concealment method and the concealed data extraction method according to the present invention are applied, centering on the control apparatus 20.
As illustrated in FIG. 3, the image processing apparatus 2 includes a control device 21 including a CPU 212 and a memory 214, a communication device 22, a recording device 24 such as an HDD / CD device, an LCD display device or a CRT display device, and a keyboard. A user interface device (UI device) 25 including a touch panel and the like is included.
The image processing apparatus 2 is, for example, a general-purpose computer in which an encoding program 5 (data concealment program) and a decoding program 6 (confidential data extraction program) according to the present invention are installed as part of a printer driver, and a communication apparatus The image data is acquired via the printer 22 or the recording device 24, and the acquired image data is encoded or decoded and transmitted to the printer device 3.

[Encoding program]
FIG. 4 is a diagram illustrating a functional configuration of the first encoding program 5 that is executed by the control device 21 (FIG. 3) and implements the data concealment method according to the present invention.
As illustrated in FIG. 4, the encoding program 5 includes a bit substitution unit 500, an intra-picture prediction unit 510, a key image prediction unit 520, a prediction error calculation unit 530, a run counting unit 540, a selection unit 550, and a code generation unit 560. And a key image generation unit 570. Note that the combination of the intra-image prediction unit 510, the key image prediction unit 520, the prediction error calculation unit 530, the run counting unit 540, and the selection unit 550 is an example of the reference information generation unit according to the present invention.
In the encoding program 5, image data is input via the communication device 22 or the recording device 24. The input image data is rasterized in the previous stage of the encoding program 5.

  The bit replacement unit 500 (embedding unit) embeds confidential information to be concealed in the image data, and outputs the image data in which the confidential information is embedded to the intra-image prediction unit 510, the key image prediction unit 520, and the prediction error calculation unit 530. To do. Specifically, the bit replacement unit 500 replaces the default bit position of the default image data (medium image 710) with the secret image 720 configured with a bit string of secret information. It is desirable that the bit position where the confidential information is embedded is the lower bit. The bit replacement unit 500 in this example generates a medium image 710 ′ by replacing the bit plane of the least significant bit of the medium image 710 (FIG. 1) with the secret image 720, and generates the medium image 710 ′ in the image. The data is output to the prediction unit 510, the key image prediction unit 520, and the prediction error calculation unit 530. In this example, the confidential information is embedded in the least significant bit. However, the present invention is not limited to this. For example, the confidential information may be embedded in a plurality of bit positions (desirably, the lower bit positions are desirable). .

  The intra-image prediction unit 510 refers to a pixel value of a pixel (reference pixel) different from the target pixel in the image data in which the confidential information is embedded (that is, the medium image 710 ′), and uses this pixel value as a predicted value. The comparison result between the predicted value and the pixel value of the target pixel is output to the run counter 540. The intra-picture prediction unit 510 of this example compares the pixel values of the plurality of reference pixels A to D (FIG. 2) with the pixel value of the target pixel X (FIG. 2), and the pixel values match (that is, When the prediction is correct, a prediction unit ID (described later) for identifying itself is output to the run counting unit 540, and in other cases, the fact that there is no match is output to the run counting unit 540. Output. Note that the intra-picture prediction unit 510 may refer to one or more reference pixels, and may output a pixel value comparison result with reference to, for example, only the reference position A.

  The key image prediction unit 520 refers to the pixel value of another image (key image) different from the image data in which the confidential information is embedded (that is, the medium image 710 ′), and determines the pixel value of this key image as the predicted value. Then, the comparison result between the predicted value and the pixel value of the target pixel (pixel included in the medium image 710 ′) is output to the run counter 540. The key image prediction unit 520 of this example compares the pixel value of the reference pixel E (FIG. 2) included in the key image with the pixel value of the target pixel X (FIG. 2), and the pixel values match ( That is, when the prediction is correct, a prediction unit ID (described later) for identifying itself is output to the run counting unit 540, and in other cases, the fact that there is no match is notified to the run counting unit 540. Output.

  The prediction error calculation unit 530 predicts the pixel value of the target pixel by a predetermined prediction method, subtracts the prediction value from the actual pixel value of the target pixel, and performs the run counting unit 540 and the selection unit 550 as prediction error values. Output for. The prediction method of the prediction error calculation unit 530 only needs to correspond to the prediction method of the decoding program 6 (described later) for decoding the code data. In order to enhance the scramble effect, the prediction error calculation unit 530 preferably refers to a pixel on the key image and uses a difference value between the pixel value of this pixel and the pixel value of the target pixel as a prediction error. . Therefore, the prediction error calculation unit 530 of this example uses the pixel value at the same reference position (reference pixel E) as the key image prediction unit 520 as a prediction value, and this prediction value and the actual pixel value (pixel value of the target pixel X). Is calculated as a prediction error value.

  The run counting unit 540 counts the number of consecutive identical prediction unit IDs, and outputs the prediction unit ID and the continuous number thereof to the selection unit 550. The prediction unit ID and the number of consecutive parts are examples of reference information for the medium image 710 'or the key image. For example, when a prediction error value is input, the run counting unit 540 outputs the prediction unit ID counted by the internal counter and its continuous number, and then the input prediction error value is directly selected by the selection unit 550. Output for.

  The selection unit 550 selects the longest continuous prediction unit ID based on the prediction unit ID, the continuous number, and the prediction error value input from the run counting unit 540, and the prediction unit ID, the continuous number, and the prediction error value. Is output to the code generation unit 560 as prediction data.

  The code generation unit 560 encodes the prediction unit ID, the number of continuations, and the prediction error value input from the selection unit 550, and outputs them to the communication device 22 or the recording device 24. In addition, when partial access restriction is performed, the code generation unit 560 outputs layout information for designating an image area for which access restriction is to be performed, attached to the code data.

  The key image generation unit 570 generates a key image having a non-uniform bit pattern at the bit position where the confidential information is embedded based on the input password, and the generated key image is used as the key image prediction unit 520 and the prediction error. Output to the calculation unit 530. For example, the key image generation unit 570 generates a random number using the input password as an argument, and generates a key image by arranging a bit pattern corresponding to the generated random number in a bit plane in which confidential information is embedded. . The random number to be generated is not necessarily a complete random number, and may be a random number generated by a pseudo random number function, for example. In this case, since the data amount of the random number generated by the pseudo random number function can be compressed more than the complete random number, the data amount of the key image can be suppressed.

FIG. 5 is a diagram for explaining an encoding process performed by the encoding program 5. FIG. 5A illustrates the positions of pixels referred to by the intra-picture prediction unit 510 and the key image prediction unit 520. FIG. 5B illustrates the code associated with each reference pixel, and FIG. 5C illustrates the code data generated by the code generation unit 560.
As illustrated in FIG. 5A, the reference positions of the intra-image prediction unit 510 and the key image prediction unit 520 are set as relative positions with respect to the target pixel X. Specifically, the reference pixel A of the intra-picture prediction unit 510 is set upstream of the target pixel X in the main scanning direction, and the reference pixels B to D are main scanning lines above the target pixel X (upstream in the sub-scanning direction). Is set on. Also, the reference pixel E of the key image prediction unit 520 is set on another image (key image) different from the input image 710.

Further, as illustrated in FIG. 5B, the priority order is set for each of the reference pixels A to E, and when the prediction is correct with a plurality of reference pixels, the run counter 540 (FIG. 4) increases the number of consecutive prediction unit IDs according to the set priority. From the viewpoint of enhancing the scramble effect, it is desirable that the priority order of the reference pixels E referred to by the key image prediction unit 520 is higher than the reference pixels A to D referred to by the intra-image prediction unit 510.
Further, as illustrated in FIG. 5B, the code generation unit 560 associates the prediction unit (reference position) with the code, and the code corresponding to the reference position where the pixel of interest X matches the pixel value. Is output. Note that the code associated with each reference position is, for example, an entropy code set according to the hit rate of each reference position, and has a code length corresponding to the priority order.

  In addition, when the pixel values continuously match at the same reference position, the code generation unit 560 encodes the continuous number counted by the run counting unit 540. As a result, the code amount is reduced. In this way, as illustrated in FIG. 5C, the encoding program 5, when the pixel values match at any reference position, the code corresponding to the reference position and the pixel at the reference position If the pixel values do not match at any reference position, the difference between the pixel value of the predetermined reference position (reference pixel E) and the pixel value of the target pixel X ( A prediction error value) is encoded.

FIG. 6 is a flowchart for explaining the operation of the encoding process (S10) by the encoding program 5.
As shown in FIG. 6, in step 100 (S100), the encoding program 5 receives from the user a confidential image 720 (1 bit) to be concealed, a medium image 710 (8 bits) in which the confidential image 720 is embedded, and The input of a password for restricting access to the secret image is accepted.

  In step 110 (S110), the bit replacement unit 500 replaces the least significant bit of the medium image 710 with the secret image 720 to generate a medium image 710 ′ in which the secret image 720 is embedded, and the generated medium image 710 ′ is output to the intra-image prediction unit 510, the key image prediction unit 520, and the prediction error calculation unit 530.

In step 120 (S120), the key image generation unit 570 generates a random number sequence using the input password as an argument, and generates the key image by arranging the generated random number sequence at the least significant bit position. The upper bits (bit positions other than the least significant bit) of the generated key image are all 0.
The key image generation unit 570 outputs the generated key image to the key image prediction unit 520 and the prediction error calculation unit 530.

In step 130 (S130), the encoding program 5 refers to the key image generated by the key image generation unit 570 and encodes the medium image 710 ′ in which the secret image 720 is embedded. More specifically, the intra-picture prediction unit 510 sets the pixel values of a plurality of reference pixels A to D (FIG. 5) set on the medium image 710 ′ in which the confidential image 720 is embedded, and the medium image 710 ′. Compared with the pixel value of the target pixel X above, if the pixel values match, the prediction unit ID for identifying itself is output to the run counting unit 540, otherwise it did not match To the run counter 540. Further, the key image prediction unit 520 compares the pixel value of the reference pixel E (FIG. 5) included in the key image with the pixel value of the target pixel X on the medium image 710 ′, and the pixel values match. In addition, a prediction unit ID (described later) for identifying itself is output to the run counting unit 540, and in other cases, the fact that they do not match is output to the run counting unit 540. The run counting unit 540 counts the continuous number of prediction unit IDs input from the intra-image prediction unit 510 or the key image prediction unit 520, and outputs the prediction unit ID and the continuous number thereof to the selection unit 550 as reference information. .
In addition, the prediction error calculation unit 530 uses the pixel value of the reference pixel E on the key image as a prediction value, and calculates the difference value between this prediction value and the pixel value of the target pixel X on the medium image 710 ′ as a prediction error value. Then, the calculated prediction error value is output to the selection unit 550 as reference information.

The selection unit 550 selects the longest continuous prediction unit ID based on the prediction unit ID and the continuous number input from the run counting unit 540 or the prediction error value input from the prediction error calculation unit 530. The prediction unit ID, the number of consecutive units, and the prediction error value are output to the code generation unit 560 as prediction data.
The code generation unit 560 encodes the prediction unit ID, the continuous number, or the prediction error value input from the selection unit 550.

  As described above, the encoding program 5 embeds the secret image 720 in the medium image 710 and encodes the medium image 710 ′ in which the secret image 720 is embedded with reference to the key image corresponding to the password.

[Decryption program]
FIG. 7 is a diagram illustrating a functional configuration of the decryption program 6 which is executed by the control device 21 (FIG. 3) and implements the secret data extraction method according to the present invention.
As illustrated in FIG. 8, the decoding program 6 includes a code decoding unit 610, a key image generation unit 620, an in-image extraction unit 630, an error processing unit 640, a key image extraction unit 650, a decoded image generation unit 660, and a bit extraction. Part 670.
In the decoding program 6, the code decoding unit 610 has a table for associating codes and prediction unit IDs (reference positions) with each other, similar to the example illustrated in FIG. 5B, based on the input code data. To specify the reference position. The code decoding unit 610 also decodes the number of consecutive prediction unit IDs and numerical values such as prediction errors based on the input code data.
The reference position, the continuation number, and the prediction error (that is, reference information) decoded in this way are input to the in-image extraction unit 630, the error processing unit 640, and the key image extraction unit 650.

The key image generation unit 620 generates a key image based on the input password by the same method as the key image generation unit 570 on the encoding side, and the generated key image is converted into an error processing unit 640 and a key image extraction unit. 650 is output. The key image generation unit 620 of this example generates a random number using the input password as an argument, and generates the key image by placing the generated random number in the least significant bit.
Note that the key image generation unit 620 of this example generates a key image (key image made up of uniform gradation data) in which the pixel values of all the pixels are 0 when no password is input, and the error processing unit 640. And output to the key image extraction unit 650.

  The in-image extraction unit 630 corresponds to a reference corresponding to a case where the prediction unit ID input from the code decoding unit 610 corresponds to any of the reference positions in the input image (that is, corresponding to the reference pixels A to D). With reference to the pixel at the position, the pixel value of that pixel is output to the decoded image generation unit 660 as decoded data. Further, when the continuous number is input together with the prediction unit ID, the in-image extraction unit 620 outputs the continuous number to the decoded image generation unit 660 in association with the pixel value corresponding to the prediction unit ID.

  When a prediction error is input from the code decoding unit 610, the error processing unit 640 outputs a pixel value corresponding to the input prediction error to the decoded image generation unit 670 as decoded data. The error processing unit 640 in this example adds the input prediction error and the pixel value of the pixel (reference pixel E) on the key image to obtain decoded data.

  When the prediction unit ID input from the encoding / decoding unit 610 corresponds to the reference position (reference pixel E) on the key image, the key image extraction unit 650 refers to the pixel of the key image and sets the pixel value of the pixel And the extracted pixel value and the input continuous number are output to the decoded image generation unit 660.

  The decoded image generation unit 660 generates a decoded image based on the decoded data input from the in-image extraction unit 630, the decoded data input from the error processing unit 640, and the decoded data input from the key image extraction unit 650. Generate. More specifically, the decoded image generation unit 660 continuously arranges pixels having the input pixel value by the continuous number when the decoded data (pixel value and continuous number) is input from the in-image extraction unit 630. To do. In addition, when the decoded data (the sum value of the pixel value of the reference pixel E and the prediction error value) is input from the error processing unit 640, the decoded image generation unit 660 arranges the pixels having the sum value. In addition, when the decoded data (pixel value and continuous number) is input from the key image extracting unit 650, the decoded image generation unit 660 continuously arranges pixels having the input pixel value by the continuous number. The pixel group arranged in this way becomes a decoded image and is input to the bit extraction unit 670.

  The bit extraction unit 670 extracts confidential information from the decoded image input from the decoded image generation unit 660. More specifically, the bit extraction unit 670 extracts a predetermined bit position of the decoded image as a secret image. The bit extraction unit 670 of this example extracts the least significant bit of the decoded image as a secret image.

FIG. 8 is a flowchart for explaining the operation of the decryption process (S20) by the decryption program 6.
As shown in FIG. 8, in step 200 (S200), the decoding program 6 acquires code data to be decoded from the user.

  In step 210 (S210), the key image generation unit 620 requests the user to input a password. The decryption program 6 proceeds to the process of S220 when the password is input, and proceeds to the process of S230 when the password is not input.

  In step 220 (S220), the key image generation unit 620 generates a random number sequence using the input password as an argument, generates a key image in which the generated random number sequence is arranged in the least significant bit, and the generated key image Are output to the error processing unit 640 and the key image extraction unit 650. That is, the key image generation unit 620 generates the same key image as the image referred to at the time of encoding when a valid password (password input at the time of encoding) is input, and otherwise, at the time of encoding A key image different from the referenced image is generated.

  In step 230 (S230), the key image generation unit 620 generates a key image in which the pixel values of all the pixels are 0, and outputs the generated key image to the error processing unit 640 and the key image extraction unit 650. . That is, the key image generation unit 620 generates a key image that is different from the image referred to at the time of encoding.

In step 240 (S240), the code decoding unit 610 sequentially decodes the codes included in the input code data into reference information, and the decoded reference information (prediction unit ID, continuous number, or prediction error value). The image is output to the in-image extracting unit 630, the error processing unit 640, and the key image extracting unit 650.
The in-image extraction unit 630, the error processing unit 640, or the key image extraction unit 650 calculates a pixel value according to the reference information (prediction unit ID, continuous number, or prediction error value) input from the code decoding unit 610, The calculated pixel value is output to the decoded image generation unit 660.
Specifically, when the prediction unit ID input from the code decoding unit 610 corresponds to any of the reference pixels A to D, the in-image extraction unit 630 refers to the pixel at the corresponding reference position, and The pixel value of the pixel is extracted, and the extracted pixel value and the input continuous number are output to the decoded image generation unit 660 as decoded data. In addition, when the prediction error value is input from the code decoding unit 610, the error processing unit 640 and the input prediction error value and the pixel value of the reference pixel E on the key image input from the key image generation unit 620 Are combined into decoded data and output to the decoded image generation unit 660. Also, the key image extraction unit 650 refers to the reference pixel E on the key image input from the key image generation unit 620 when the prediction unit ID input from the code decoding unit 610 corresponds to the reference pixel E. The pixel value of this pixel is extracted, and the extracted pixel value and the input continuous number are output to the decoded image generation unit 660.
When the decoded data (pixel value and continuous number) is input from the in-image extracting unit 630, the decoded image generation unit 660 continuously arranges the pixels having the input pixel value by the continuous number, and the error processing unit 640. When the decoded data (the sum value of the pixel value of the reference pixel E and the prediction error value) is input from, the pixels of this sum value are arranged, and the key image extraction unit 650 receives the decoded data (the pixel value and the continuous number). Is input, the pixels having the input pixel values are continuously arranged in a continuous number. The pixel group arranged in this way becomes a decoded image.

In step 250 (S250), the bit extraction unit 670 extracts the least significant bit of the decoded image input from the decoded image generation unit 660 as a confidential image.
As described above, the decryption program 6 generates a key image based on the input password, reproduces the decoded image by referring to the generated key image in accordance with the encoded data, and from the reproduced decoded image. A secret image is extracted. Therefore, when an inappropriate password is input (that is, when a password other than the password input at the time of encoding is input) or when the password is not input, the key image used at the time of encoding is used. A key image different from the above is generated by the key image generation unit 620, and the least significant bit of the medium image 710 ′ is decoded in a scrambled state.

As described above, the image processing apparatus 2 according to the present embodiment encodes the medium image 710 ′ in which the confidential image 720 is embedded with reference to the key image, so that the confidential image embedded in the medium image 710 ′. Access to the secret image 720 can be restricted by scrambling the image 720. In particular, the image processing apparatus 2 of the present example encodes the medium image 710 ′ by applying a key image in which only the bit position in which the secret image 720 is embedded is non-uniform, thereby decoding without the key image. In addition, it is possible to generate a media image 710 ″ that is very similar to the media image 710 ′ before encoding while maintaining scrambling for the secret image 720.
Further, the image processing apparatus 2 can realize the scramble process for the secret image 720 and the encoding process for the medium image 710 ′ in which the secret image 720 is embedded by the same process.

[Modification 1]
Next, a modification of the first embodiment will be described.
The key image in the above embodiment is configured with pseudo random numbers at the bit positions where the confidential information is embedded, but is not limited to this, and the image has a non-uniform bit pattern at the bit positions where the confidential information is embedded. If it is. For example, the key image 730 may be the same image as the medium image 710.

FIG. 9 is a diagram for explaining data concealment processing to which the same key image 730 as the medium image 710 is applied.
As illustrated in FIG. 9, when the encoding program 5 refers to the same key image 730 as the medium image 710 and encodes the medium image 710 ′ in which the secret image 720 is embedded, the bit of the secret image 720 is 0. In the area where the prediction error occurs, the prediction in the reference pixel E (FIG. 5) is correct. In the area where the prediction error is encoded, the prediction error is small. In addition, when the medium image 710 is a photographic image or the like, it is highly possible that pixels that are close to each other have similar pixel values, and thus prediction in the reference pixels A to D may be correct. Therefore, when the medium image 710 ′ is encoded with reference to the same key image 730 as the medium image 710, the compression rate increases.

When encoding is performed with reference to the same key image 730 as the medium image 710, the code data has a higher ratio of codes (code E corresponding to the reference pixel E) instructing reference to the key image 730. The value of the prediction error greatly depends on the pixel value of the key image 730. Therefore, when the code data is decoded with reference to an image different from the key image 730 referred to at the time of encoding, a scrambled medium image 710 ″ is generated as a decoded image. Since “is scrambled even at the bit position where the confidential information is embedded, the confidential image 720 extracted from the medium image 710” is also scrambled.
On the other hand, when the code data is decoded with reference to the key image 730 referred to at the time of encoding, the original medium image 710 ′ is accurately reproduced as a decoded image, and the secret extracted from the medium image 710 ′. The image 720 is also reproduced in a viewable state.

FIG. 10 is a diagram illustrating a functional configuration of the second encoding program 52 in the present modification. It should be noted that among the components in this figure, the same reference numerals are assigned to the components that are substantially the same as those shown in FIG.
As illustrated in FIG. 10, the second encoding program 52 has a configuration in which the bit replacement unit 500 of the first encoding program 5 is replaced with the target image 505 and the key image generation unit 570 is deleted.
In the encoding program 52, the target image generation unit 505 generates encoding target data based on the confidential information to be concealed and the predetermined image data. The target image generation unit 505 of this example embeds the secret image 720 in the same medium image 710 (FIG. 9) as the key image 730, and generates a medium image 710 ′ in which the secret image 720 is embedded.

In this modification, the key image prediction unit 520 encodes the encoding target data (medium image 710 ′) generated by the target image generation unit 505 with reference to the same key image 730 as the medium image 710.
Also, the prediction error calculation unit 530 refers to the same key image 730 as the medium image 710, the pixel value on the key image 730, and the pixel value on the medium image 710 ′ generated by the target image generation unit 505. Is calculated as a prediction error.

  This medium image 710 ′ is obtained by replacing the least significant bit of the key image 730 with the secret image 720, and approximates the key image 730. For this reason, the target probability of prediction referring to the key image 730 is high, and the prediction error is reduced even if the target is not correct. Therefore, by shortening the code length of the code E (FIG. 5) corresponding to the reference pixel E and the code length for frequent prediction errors (+1, −1), the code amount of the medium image 710 ′ is reduced.

  As described above, the image processing apparatus 2 in the present modification refers to the medium image 710 in which the confidential information is embedded as a key image, and encodes the medium image 710 ′ in which the confidential information is embedded, thereby increasing the compression rate. Can do.

[Other variations]
The target image generation unit 505 may generate the encoding target data (medium image 710 ′) by comparing the bit pattern of the confidential information with the bit string of each pixel constituting the key image. Specifically, the target image generation unit 505 compares the bit pattern of the confidential information with the value of the bit position in which the confidential information is embedded in the bit string indicating the pixel value of the key image, and based on these combinations, The value of this bit position is determined.

Further, when the medium image 710 is an image having a uniform pixel value, there is a possibility that a code corresponding to the reference pixel is generated by the prediction by the reference pixels A to D. In this case, since the medium image 710 ′ is correctly decoded without referring to the key image, there is a possibility that the secret image is easily extracted.
Therefore, the encoding program 5 has a pixel value of the pixel of interest X and a pixel of the reference pixel E on the key image at least in a region where the pixel value is uniform, when the medium image 710 has a region composed of uniform pixel values. It is desirable to encode a difference value (ie, prediction error) from the value. Specifically, the selection unit 550 (FIG. 4) determines the prediction error (the pixel value of the target pixel X and the reference on the key image) calculated by the prediction error calculation unit 530 in a region where the pixel values of the medium image 710 are uniform. Difference value with respect to the pixel value of the pixel E) is selected and output to the code generation unit 560.

It is a figure explaining embedding of the secret information with respect to image data. It is a figure explaining the outline of the prediction encoding process which refers to a key image, (A) illustrates the reference position referred in a prediction encoding process, (B) is produced | generated by a prediction encoding process. The code data is schematically illustrated. It is a figure which illustrates the hardware constitutions of the image processing apparatus 2 to which the data concealment method and secret data extraction method concerning this invention are applied centering on the control apparatus 20. FIG. It is a figure which illustrates the functional structure of the 1st encoding program 5 which is performed by the control apparatus 21 (FIG. 3), and implement | achieves the data concealment method concerning this invention. It is a figure explaining the encoding process performed by the encoding program 5, (A) illustrates the position of the pixel referred by the intra image prediction part 510 and the key image prediction part 520, (B) is respectively The code | symbol matched with the reference pixel of this is illustrated, (C) is a figure which illustrates the code | cord | chord data produced | generated by the code | symbol production | generation part 560. FIG. It is a flowchart explaining operation | movement of the encoding process (S10) by the encoding program 5. FIG. It is a figure which illustrates the functional structure of the decoding program 6 which is performed by the control apparatus 21 (FIG. 3), and implement | achieves the secret data extraction method concerning this invention. It is a flowchart explaining operation | movement of the decoding process (S20) by the decoding program 6. FIG. It is a figure explaining the data concealment process to which the same key image 730 as the medium image 710 is applied. 3 is a diagram illustrating a functional configuration of a second encoding program 52. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 2 ... Image processing apparatus 5,52 ... Encoding program 500 ... Bit substitution part 505 ... Target image generation part 510 ... Intra-picture prediction part 520 ... Key image prediction part 530 ... -Prediction error calculation part 540 ... Run counting part 550 ... Selection part 560 ... Code generation part 570 ... Key image generation part 6 ... Decoding program 610 ... Code decoding part 620 ... Key image generation unit 630 ... In-image extraction unit 640 ... Error processing unit 650 ... Key image extraction unit 660 ... Decoded image generation unit 670 ... Bit extraction unit

Claims (13)

A data concealment device for concealing confidential data,
An embedding means for embedding confidential data in image data;
Reference information generating means for comparing the image data in which the secret data is embedded by the embedding means and a default key image, and generating reference information for designating a reference to the key image;
A data concealment apparatus comprising: code generation means for generating code data of reference information generated by the reference information means as part of code data of image data in which confidential data is embedded. The embedding unit embeds the confidential data in the image data by replacing the predetermined bit position of the image data with the confidential data,
The reference information generation unit generates reference information for the key image by comparing the key image having a non-uniform bit position with image data in which confidential data is embedded. The data concealment device described. The reference information generation unit compares the gradation data of the target pixel included in the image data in which the confidential data is embedded with the gradation data of the pixel on the key image corresponding to the target pixel, and compares the gradation data. A difference value of data is generated as the reference information,
The data concealment device according to claim 1, wherein the code generation unit generates code data of the generated difference value as code data of image data in which confidential data is embedded. Based on a password, further comprising key image generation means for generating a data string composed of non-uniform values, and arranging the generated data string at the bit position to generate the key image;
The data concealment device according to claim 2, wherein the reference information generation unit generates reference information for the key image using the key image generated by the key image generation unit. The image data in which the confidential data is embedded is the key image,
The embedding means embeds secret data in the key image,
The reference information generating means compares the key image in which the confidential data is embedded with the key image in which the confidential data is not embedded, and generates reference information for the key image in which the confidential data is not embedded,
The data concealment device according to claim 1, wherein the code generation unit generates the code data of the generated reference information as a part of code data of a key image in which confidential data is embedded. A data concealment device for concealing confidential data,
Target data generation means for generating encoding target data based on the predetermined image data and the confidential data;
Reference information generating means for comparing the encoding target data generated by the target data generating means and the image data, and generating reference information for specifying a reference to the image data;
A data concealment apparatus comprising: code generation means for generating code data of reference information generated by the reference information means as a part of code data of the encoding target data. A secret data extraction device that extracts secret data from code data,
Reference data extraction means for extracting image data included in the key image based on the input code data;
Decoded image generation means for generating a decoded image using the image data extracted by the reference data extraction means;
A secret data extraction apparatus comprising: bit extraction means for extracting secret data from a predetermined bit position of a decoded image generated by the decoded image generation means. A data concealment method for concealing confidential data,
Embed confidential data in the image data,
Compare the image data in which the confidential data is embedded with the default key image, and generate reference information for specifying a reference to the key image.
A data concealment method for generating code data of generated reference information as part of code data of image data in which confidential data is embedded. A data concealment method for concealing confidential data,
Based on the predetermined image data and the confidential data, the encoding target data is generated,
Compare the generated encoding target data with the image data, and generate reference information for specifying a reference to the image data,
A data concealment method for generating code data of generated reference information as part of code data of the encoding target data. A secret data extraction method for extracting secret data from code data,
Based on the input code data, the image data included in the key image is extracted,
Using the extracted image data, generate a decoded image,
A secret data extraction method for extracting secret data from a predetermined bit position of a generated decoded image. In a data concealment device that conceals confidential data,
Embedding confidential data in the image data;
Comparing the image data in which the confidential data is embedded with a predetermined key image, and generating reference information for specifying a reference to the key image;
A program for causing the data concealment apparatus to execute the step of generating the encoded data of the reference information as a part of the encoded data of the image data in which the confidential data is embedded. In a data concealment device that conceals confidential data,
Generating encoding target data based on the predetermined image data and the confidential data;
Comparing the generated encoding target data with the image data, and generating reference information for specifying a reference to the image data;
A program for causing the data concealment apparatus to execute the step of generating the code data of the generated reference information as a part of the code data of the encoding target data. In the secret data extraction device that extracts the secret data from the code data,
Extracting image data included in the key image based on the input code data;
Generating a decoded image using the extracted image data; and
A program that causes the secret data extraction device to execute a step of extracting secret data from a predetermined bit position of the generated decoded image.

Images