JP5728106B1 - Digital watermark embedding device, digital watermark embedding program, digital watermark detection device, and digital watermark detection program - Google Patents

Abstract

Image quality deterioration at the time of embedding a digital watermark is suppressed. An electronic watermark embedding apparatus 100 performs a frequency transform (Fourier transform or the like) on two signal components (for example, a U component and a V component) in an original image (or video frame) in a frequency space. At specific coordinates, the coefficient is changed so that the first component vector (two-dimensional vector having real and imaginary parts of frequency components as elements) and the second component vector are orthogonal to each other. When embedding a bit having a value of 1 on the two-dimensional plane, the first component vector is rotated 90 degrees counterclockwise with respect to the second component vector, and when embedding a bit having a value of 0 is 90 clockwise. Set to rotate in degrees. [Selection] Figure 1

Description

  The present invention relates to a digital watermark embedding device, a digital watermark embedding program, a digital watermark detection device, and a digital watermark detection program.

There is a technique for embedding a digital watermark in an image (see, for example, Patent Document 1). In this technique, the following processing is performed.
When embedding a digital watermark, the frequency signal is changed so that a peak is generated at a specific coordinate of the amplitude component data subjected to Fourier transform.
・ When detecting a digital watermark, the peak point is obtained from the frequency data obtained by Fourier transform of the target image, the “deviation” between the peak component of the peak point and the coordinates is obtained from the coordinates, and the “deviation” of the image is obtained. to correct. A digital watermark is detected from the image after the “deviation” correction.

JP 2002-171395 A

The prior art has the following problems.
Double embedding is necessary: embedding a digital watermark to indicate peak points and embedding a digital watermark to indicate bit information. Therefore, the degree of image quality degradation that occurs when embedding a digital watermark is large.
-It is necessary to perform two-stage processing of image "deviation" correction and digital watermark detection. For this reason, the amount of processing when detecting a digital watermark is large. In addition, the detection accuracy decreases due to the image deterioration caused by the image “shift” correction process.

  An object of the present invention is to suppress, for example, image quality degradation when embedding a digital watermark. Another object of the present invention is to reduce the amount of processing when detecting a digital watermark, for example.

An electronic watermark embedding apparatus for embedding an electronic watermark in an input image according to an aspect of the present invention is provided:
The first component image signal is obtained by performing inverse frequency conversion on a first frequency signal that is composed of a plurality of first frequency components and whose vector direction is the first direction corresponding to the data amount of the digital watermark. A first converter to obtain;
The vector direction of the number of second frequency components configured by a plurality of second frequency components according to the data amount of the digital watermark is shifted by a predetermined angle to the side corresponding to the data value of the digital watermark with respect to the first direction. A second conversion unit that obtains a second component image signal by performing inverse frequency conversion on the second frequency signal in the second direction;
The first component image signal obtained by the first conversion unit, the second component image signal obtained by the second conversion unit, and the input image data including the first component and the second component. And a synthesis unit that synthesizes an image signal of a component included in addition to the component.

For example, the digital watermark embedding device is:
A signal component separation unit for separating the first component image signal and the second component image signal from the input image data;
The first conversion unit converts the image signal of the first component obtained by the signal component separation unit into a frequency signal, and according to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal Adjusting the vector direction of a number of frequency components to generate the first frequency signal, and inversely converting the first frequency signal to the image signal of the first component;
The second conversion unit converts the image signal of the second component obtained by the signal component separation unit into a frequency signal, and according to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal The second frequency signal is generated by adjusting vector directions of a number of frequency components, and the second frequency signal is inversely converted into the image signal of the second component.

Alternatively, the digital watermark embedding device may
A signal component separation unit that separates the image signal of the first component and the image signal of the second component from data of an image different from the input image;
The first conversion unit converts the image signal of the first component obtained by the signal component separation unit into a frequency signal, and according to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal Adjusting the vector direction of a number of frequency components to generate the first frequency signal, and inversely converting the first frequency signal to the image signal of the first component;
The second conversion unit converts the image signal of the second component obtained by the signal component separation unit into a frequency signal, and according to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal Adjusting the vector direction of a number of frequency components to generate the second frequency signal, and inversely converting the second frequency signal into the image signal of the second component,
The combining unit adds the first component image signal obtained by the first conversion unit, the second component image signal obtained by the second conversion unit, and the second image data to the first image data. An image signal of a component included in addition to the one component and the second component is synthesized, and the obtained image and the input image are synthesized.

According to one aspect of the present invention, an electronic watermark embedding program for embedding an electronic watermark in an input image is provided.
On the computer,
The first component image signal is obtained by performing inverse frequency conversion on a first frequency signal that is composed of a plurality of first frequency components and whose vector direction is the first direction corresponding to the data amount of the digital watermark. A first conversion process to obtain;
The vector direction of the number of second frequency components configured by a plurality of second frequency components according to the data amount of the digital watermark is shifted by a predetermined angle to the side corresponding to the data value of the digital watermark with respect to the first direction. A second conversion process for obtaining a second component image signal by performing frequency inverse conversion on the second frequency signal in the second direction;
The first component image signal obtained by the first conversion process, the second component image signal obtained by the second conversion process, and the input image data include the first component and the second component. And a synthesis process for synthesizing the image signal of the component included in addition to the component.

According to one aspect of the present invention, a digital watermark detection apparatus for detecting a digital watermark embedded in an input image is provided.
A signal component separation unit for separating a first component image signal and a second component image signal from the input image data;
A first frequency converting unit that obtains a first frequency signal by frequency converting the image signal of the first component obtained by the signal component separating unit;
A second frequency conversion unit that obtains a second frequency signal by frequency converting the image signal of the second component obtained by the signal component separation unit;
A plurality of first frequency components constituting the first frequency signal obtained by the first frequency converter, and a plurality of second frequency components constituting the second frequency signal obtained by the second frequency converter. A vector outer product calculation unit for calculating a vector outer product of a first frequency component and a second frequency component having the same coordinates in the frequency space, and
Based on the absolute value of the vector outer product calculated by the vector outer product calculation unit, the number of coordinates corresponding to the data amount of the digital watermark is selected from the coordinates in the frequency space, and the vector is selected for each selected coordinate. A determination unit that determines a data value of the digital watermark based on the sign of the vector outer product calculated by the outer product calculation unit.

According to one aspect of the present invention, a digital watermark detection program for detecting a digital watermark embedded in an input image is provided.
On the computer,
A signal component separation process for separating a first component image signal and a second component image signal from the input image data;
A first frequency conversion process for obtaining a first frequency signal by frequency-converting the image signal of the first component obtained by the signal component separation process;
A second frequency conversion process for obtaining a second frequency signal by frequency-converting the image signal of the second component obtained by the signal component separation process;
Of the plurality of first frequency components constituting the first frequency signal obtained by the first frequency conversion process, the number of first frequency components corresponding to the data amount of the digital watermark, and the second frequency conversion process Among the plurality of second frequency components constituting the obtained second frequency signal, the number of second frequency components corresponding to the data amount of the digital watermark is the same as the first frequency component having the same coordinates in the frequency space. A vector cross product calculation process for calculating a vector cross product with two frequency components;
Based on the vector outer product calculated in the vector outer product calculation process, a determination process for determining a data value of the digital watermark is executed.

  According to the present invention, it is possible to suppress deterioration in image quality when embedding a digital watermark. Furthermore, according to the present invention, it is possible to reduce the processing amount when detecting a digital watermark.

1 is a block diagram illustrating a configuration of a digital watermark embedding device according to Embodiment 1. FIG. FIG. 6 shows an example of a digital watermark embedding process according to the first embodiment. 1 is a block diagram illustrating a configuration of a digital watermark detection apparatus according to Embodiment 1. FIG. FIG. 6 shows an example of a digital watermark detection process according to the first embodiment. FIG. 4 is a block diagram illustrating a configuration of a digital watermark embedding device according to a second embodiment. 1 is a diagram illustrating an example of a hardware configuration of a digital watermark embedding device and a digital watermark detection device according to an embodiment of the present invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

Embodiment 1 FIG.
First, embedding of a digital watermark according to this embodiment will be described.

  FIG. 1 is a block diagram showing a configuration of digital watermark embedding apparatus 100 according to the present embodiment. FIG. 2 is a diagram illustrating an example of a digital watermark embedding process in which the digital watermark embedding apparatus 100 embeds a digital watermark in an input image.

  In FIG. 1, the digital watermark embedding apparatus 100 includes a signal component separation unit 101, a first conversion unit 110, a second conversion unit 120, and a synthesis unit 130.

  The signal component separation unit 101 receives, as input image data, still image data or moving image data (for example, each frame of video data) from the outside.

  The signal component separation unit 101 separates a U component image signal (pixel signal), a V component image signal (pixel signal), and a Y component image signal from input image data. The signal component separation unit 101 inputs a U component image signal to the first conversion unit 110, inputs a V component image signal to the second conversion unit 120, and inputs a Y component image signal to the synthesis unit 130. Here, the Y component is the luminance signal, the U component is the difference (color difference signal) between the luminance signal and the blue component, and the V component is the difference (color difference signal) between the luminance signal and the red component. In the present embodiment, the U component is an example of the first component, the V component is an example of the second component, and the Y component is a component included in the input image data in addition to the first component and the second component. It is an example. Note that this embodiment can be applied even if the U component and the V component are interchanged (that is, the V component is the first component and the U component is the second component). Further, the present embodiment can be applied even if a format other than the YUV format is used. For example, if the RGB format is used, any one of the R component, the G component, and the B component may be the first component, and any one of the other components may be the second component. When a format other than the YUV format is used, if there are two or more components other than the first component and the second component included in the input image data, the synthesizing unit 130 receives the image signals of these two or more components. Will be entered.

  If the data amount of the digital watermark (for example, the number of bits of the watermark data) is determined in advance, this is not necessary. If not, information on the data amount of the digital watermark is input to the first conversion unit 110 from the outside. Is done. Further, as described above, the U-component image signal is input from the signal component separator 101 to the first converter 110.

  The first conversion unit 110 includes a plurality of first frequency components, and the vector direction of the number of first frequency components corresponding to the amount of digital watermark data (for example, the same number as the number of bits of watermark data) is the first direction. The first frequency signal is inversely converted to obtain a U component image signal. As will be described later, the first frequency signal is generated based on the U component image signal input from the signal component separation unit 101. That is, the first conversion unit 110 converts the U component image signal input from the signal component separation unit 101 into a U component image signal that has been subjected to vector adjustment in accordance with the data amount of the digital watermark.

  Digital watermark data (watermark data) is input to the second conversion unit 120 from the outside. Further, as described above, the V component image signal is input to the second conversion unit 120 from the signal component separation unit 101.

  The second conversion unit 120 includes a plurality of second frequency components, and the vector direction of the second frequency components of the number corresponding to the data amount of the digital watermark (for example, the same number as the number of bits of the watermark data) is the first direction described above. In a second direction shifted by a predetermined angle (for example, 90 degrees) to the side corresponding to the data value of the digital watermark (for example, the plus side if the bit value of the watermark data is 0, the minus side if 1). A certain second frequency signal is inversely transformed to obtain a V component image signal. As will be described later, the second frequency signal is generated based on the V component image signal input from the signal component separation unit 101. That is, the second conversion unit 120 converts the V component image signal input from the signal component separation unit 101 into a V component image signal that has been subjected to vector adjustment according to the data amount and data value of the digital watermark.

  The synthesizing unit 130, the U component image signal obtained by the first conversion unit 110, the V component image signal obtained by the second conversion unit 120, and the Y component image input from the signal component separation unit 101. The signal is combined and output. As a result, an output image in which a digital watermark is embedded in the input image is obtained.

  Below, the detail of the 1st conversion part 110 is demonstrated.

  The first conversion unit 110 includes a frequency conversion unit 102, a vector adjustment unit 104, and a frequency inverse conversion unit 106.

  The frequency conversion unit 102 converts the U component image signal obtained by the signal component separation unit 101 into a frequency signal. For example, as shown in FIG. 2, the frequency conversion unit 102 converts the U component image signal obtained by the signal component separation unit 101 into a two-dimensional Fourier coefficient by using the two-dimensional Fourier transform. The coefficient is output as a frequency signal. In this case, the frequency signal has two elements, a real part x and an imaginary part y, as frequency components for each coordinate in the frequency space. Each frequency component can be represented by a vector (or complex number x + yi) composed of these two elements.

The vector adjustment unit 104 includes a number of frequency components corresponding to the amount of digital watermark data (for example, the same number as the number of bits of watermark data) among a plurality of frequency components constituting the frequency signal obtained by the frequency conversion unit 102. The vector direction is adjusted to generate the first frequency signal described above. For example, as illustrated in FIG. 2, the vector adjustment unit 104 sets a vector direction in a predetermined direction (that is, the first direction) with respect to a frequency component vector located at a specific coordinate in the U component frequency space. Obtain the aligned vector U (x _u , y _u ). As many specific coordinates as the number of bits of the watermark data are set.

  The frequency inverse transform unit 106 inversely transforms the first frequency signal generated by the vector adjustment unit 104 into a U component image signal. For example, the frequency inverse transform unit 106 returns the U component frequency signal (that is, the first frequency signal) to the image signal (pixel signal) by two-dimensional Fourier transform opposite to the frequency transform unit 102.

  Below, the detail of the 2nd conversion part 120 is demonstrated.

  The second conversion unit 120 includes a frequency conversion unit 103, a vector adjustment unit 105, and a frequency inverse conversion unit 107.

  The frequency conversion unit 103 converts the V component image signal obtained by the signal component separation unit 101 into a frequency signal. For example, as shown in FIG. 2, the frequency conversion unit 103 uses the same two-dimensional Fourier transform as the frequency conversion unit 102 to convert the V component image signal obtained by the signal component separation unit 101 into a two-dimensional Fourier coefficient. The two-dimensional Fourier coefficient is converted and output as a frequency signal.

The vector adjustment unit 105 includes a number of frequency components corresponding to the amount of digital watermark data (for example, the same number as the number of bits of watermark data) among a plurality of frequency components constituting the frequency signal obtained by the frequency conversion unit 103. The vector direction is adjusted to generate the second frequency signal described above. For example, as shown in FIG. 2, the vector adjustment unit 105 sets the vector direction to a certain direction (that is, the second direction) for the frequency component vector located at the same coordinate as the specific coordinate in the V component frequency space. To obtain a vector V ₀ (x _v0 , y _v0 ) or V ₁ (x _v1 , y _v1 ). The specific coordinates are the same as the specific coordinates used in the vector adjustment unit 104. The alignment direction varies depending on each bit value of the input watermark data. For each specific coordinate, when the bit value corresponding to the specific coordinate is 0, V is in a direction rotated 90 degrees clockwise with respect to the vector direction of the frequency component of the U component aligned by the vector adjustment unit 104. The vector directions of the frequency components of the components are aligned. When the bit value corresponding to the specific coordinate is 1, the frequency component of the V component is rotated in the direction rotated 90 degrees counterclockwise with respect to the vector direction of the U component frequency component aligned by the vector adjustment unit 104. The vector direction is aligned.

  As a modified example of the present embodiment (hereinafter referred to as a first modified example), the vector adjustment unit 105 does not align vectors in a predetermined fixed direction, but when the bit value corresponding to a specific coordinate is 0 The vector direction of the frequency component of the V component is aligned in the direction rotated 90 degrees clockwise with respect to the vector direction of the frequency component of the U component of the same coordinate obtained from the frequency conversion unit 102, and the specific coordinate is When the corresponding bit value is 1, the vector of the V component frequency component is rotated 90 degrees counterclockwise with respect to the vector direction of the U component frequency component of the same coordinate obtained from the frequency converter 102. It can be configured to align the directions. In that case, even if the vector adjustment unit 104 does not perform vector adjustment of the frequency component of the U component, the same effect can be obtained. That is, in the first modification, the vector adjustment unit 104 can be omitted. In the first modification, the output of the frequency conversion unit 102 is passed not only to the vector adjustment unit 104 (or the frequency inverse conversion unit 106 when the vector adjustment unit 104 is omitted) but also to the vector adjustment unit 105. In the first modification, the vector adjustment unit 105 corrects the vector direction of the frequency component of the V component of the same coordinate by looking at the vector direction of the frequency component of the U component output from the frequency conversion unit 102. As a modification example (hereinafter referred to as a second modification example), the vector adjustment unit 104 corrects the vector direction of the U component of the same coordinate by looking at the vector direction of the V component frequency component output from the frequency conversion unit 103. Can be configured. In this case, the vector adjustment unit 105 may not perform the vector adjustment of the frequency component of the V component, or the vector adjustment unit 105 may be omitted. In the second modification, the output of the frequency conversion unit 103 is passed not only to the vector adjustment unit 105 (or the frequency inverse conversion unit 107 when the vector adjustment unit 105 is omitted) but also to the vector adjustment unit 104.

  The frequency inverse transform unit 107 inversely transforms the second frequency signal generated by the vector adjustment unit 105 into a V component image signal. For example, the frequency inverse transform unit 107 returns the V component frequency signal (that is, the second frequency signal) to the image signal (pixel signal) by two-dimensional Fourier transform reverse to the frequency transform unit 103.

  Below, the detail of the synthetic | combination part 130 is demonstrated.

  The combining unit 130 includes a signal component combining unit 108.

  The signal component synthesizing unit 108 is a U component image signal output from the frequency inverse transform unit 106, a V component image signal output from the frequency inverse transform unit 107, and a Y component input from the signal component separation unit 101. Are combined with the image signal to obtain an output image.

As described above, in this embodiment, the digital watermark embedding apparatus 100 embeds a digital watermark in an image by the following method.
A first component vector (in a specific coordinate in the frequency space, for a signal obtained by frequency transforming (for example, Fourier transform) two signal components (for example, U component and V component) in the original image (or video frame). The coefficient is changed so that the two-dimensional vector having the real component and the imaginary component of the frequency component are orthogonal to the second component vector (see, for example, FIG. 2).
When embedding a bit having a value of 1, the first component vector is rotated 90 degrees counterclockwise with respect to the second component vector on the two-dimensional plane, and when embedding a bit having a value of 0 is clockwise. Set to rotate 90 degrees. Note that the correspondence between the bit value (0 or 1) and the vector rotation direction (right or left) may be reversed. Further, the rotation angle of the vector need not be 90 degrees. However, the closer the rotation angle of the vector is to 90 degrees, the greater the outer product of the first component vector and the second component vector at the same coordinates. That is, the area of the parallelogram formed by the first component vector and the second component vector at the same coordinate (a rectangle when the rotation angle is 90 degrees) is increased. Therefore, when the digital watermark is detected, the probability that a peak occurs at the coordinates increases.

  The above method eliminates the need for double embedding of embedding a digital watermark for indicating a peak point and embedding a digital watermark for indicating bit information, as in the prior art. That is, it is possible to avoid embedding a digital watermark for indicating a peak point necessary for image correction, which is independent of a digital watermark for indicating bit information. Therefore, according to the present embodiment, it is possible to suppress deterioration in image quality when embedding a digital watermark.

  Next, digital watermark detection according to this embodiment will be described.

  FIG. 3 is a block diagram showing a configuration of digital watermark detection apparatus 200 according to the present embodiment. FIG. 4 is a diagram illustrating an example of a digital watermark detection process in which the digital watermark detection apparatus 200 detects a digital watermark embedded in an input image.

  In FIG. 3, the digital watermark detection apparatus 200 includes a signal component separation unit 201, a first frequency conversion unit 202, a second frequency conversion unit 203, a vector outer product calculation unit 204, and a determination unit 210.

  Output signal data obtained by the digital watermark embedding apparatus 100 is input to the signal component separation unit 201 from the outside as input image data.

  The signal component separation unit 201 separates a U component image signal (pixel signal) and a V component image signal (pixel signal) from input image data. Unlike the signal component separation unit 101 of the digital watermark embedding apparatus 100, it is not necessary to separate the Y component image signal. The signal component separation unit 201 inputs the U component image signal to the first frequency conversion unit 202, and inputs the V component image signal to the second frequency conversion unit 203.

  The first frequency conversion unit 202 converts the frequency of the U component image signal obtained by the signal component separation unit 201 to obtain a first frequency signal.

  The second frequency conversion unit 203 converts the frequency of the V component image signal obtained by the signal component separation unit 201 to obtain a second frequency signal.

  The vector outer product calculation unit 204 includes a plurality of first frequency components constituting the first frequency signal obtained by the first frequency conversion unit 202 and a plurality of second frequency signals obtained by the second frequency conversion unit 203. For the second frequency component, a vector outer product of the first frequency component and the second frequency component having the same coordinates in the frequency space is calculated. For example, as shown in FIG. 4, the vector cross product calculation unit 204 uses the first frequency signal that is the output of the first frequency conversion unit 202 and the second frequency signal that is the output of the second frequency conversion unit 203. Calculate the outer product of vectors of frequency components with the same coordinates (the outer product of two-dimensional vectors mapped to three-dimensional vectors (the x and y components are left unchanged and the z component is set to 0)), and the z component is calculated as the coordinate The calculation result in (vector outer product data).

  Based on the absolute value of the vector outer product calculated by the vector outer product calculation unit 204, the determination unit 210 determines the number (for example, all coordinates) in the frequency space according to the amount of digital watermark data (for example, all coordinates). Coordinates (corresponding to the specific coordinates described above) are selected. The determination unit 210 determines the data value of the digital watermark (for example, the bit value of the watermark data) based on the sign of the vector outer product calculated by the vector outer product calculation unit 204 for each selected coordinate.

  Below, the detail of the determination part 210 is demonstrated.

  The determination unit 210 includes a peak extraction unit 205 and a bit determination unit 206.

  The peak extraction unit 205 extracts coordinates (peak coordinates corresponding to the number of embedded bits) at which the absolute value becomes a peak in the coordinate space of the vector outer product data that is the output of the vector outer product calculation unit 204. As the means, for example, the peak extraction unit 205 appropriately performs vector enlargement / reduction and rotation by a predetermined angle on a plurality of specific coordinates set by the digital watermark embedding apparatus 100. Then, the absolute value sum of the coordinate values is calculated, and the position where it becomes the maximum is calculated. This makes it possible to detect a digital watermark even if the input image is enlarged or reduced or rotated. It should be noted that even if the input image is translated, there is no influence, and thus vector translation processing is unnecessary.

  The bit determining unit 206 looks at the positive / negative coordinate values of the vector outer product data for each of the peak coordinates obtained by the peak extracting unit 205. It is determined that the value is 0. The bit determination unit 206 collects the bit determination results of all the peak coordinates and outputs them as digital watermark detection data.

As described above, in this embodiment, the digital watermark detection apparatus 200 detects a digital watermark from an image by the following method.
In a signal obtained by frequency-transforming (for example, Fourier transform) two signal components (for example, U component and V component) of the detection target image (or video frame), the first component vector and the second component of the same coordinates The cross product with the vector (the z component thereof, that is, the “cross product value”) is calculated to obtain vector cross product data (see, for example, FIG. 4).
The cross product value (the absolute value thereof) increases as the angle between the vectors becomes closer to 90 degrees. Also, the outer product value is positive if the second component vector is counterclockwise with respect to the first component vector, and negative if it is clockwise. Using this property, the coordinates at which the absolute value of the vector outer product data has a peak are extracted, and it is determined whether the bit value is 0 or 1 based on the sign of the outer product value in the absolute value peak coordinates.

  The above method eliminates the need for two-stage processing such as image “shift” correction and digital watermark detection as in the prior art. In other words, it is not necessary to correct the “shift” of the image as a pre-processing for detecting the digital watermark, and the digital watermark can be detected directly from the frequency conversion data of the image. Therefore, according to the present embodiment, it is possible to reduce the amount of processing at the time of detecting a digital watermark (speeding up the processing). In addition, it is possible to prevent a decrease in detection accuracy due to image degradation caused by image “shift” correction processing.

Embodiment 2. FIG.
In the present embodiment, differences from the first embodiment will be mainly described.

  In the digital watermark embedding according to the first embodiment, the frequency vector is adjusted based on the data of the input image. In the digital watermark embedding according to the present embodiment, the frequency vector is adjusted based on the previously generated image. Thus, an electronic watermark signal independent of the input image is generated, and the signal is superimposed on the input image.

  Hereinafter, embedding of a digital watermark according to this embodiment will be described.

  FIG. 5 is a block diagram showing a configuration of digital watermark embedding apparatus 100a according to the present embodiment.

  In FIG. 5, a digital watermark embedding device 100a includes a signal component separation unit 101, a first conversion unit 110, and a second conversion unit 120 similar to those of the digital watermark embedding device 100 according to the first embodiment, a synthesis unit 130a, and an image. A generation unit 140 is provided.

  The image generation unit 140 generates image data different from the input image. This other image may be an arbitrary image. An example of a suitable image is an image in which all pixel values are uniform (that is, an image in which the U component is the same, the V component is the same, and the Y component is 0 in all the pixels).

  Data of an image different from the input image is input from the image generation unit 140 to the signal component separation unit 101.

  The signal component separation unit 101 separates a U component image signal (pixel signal), a V component image signal (pixel signal), and a Y component image signal from data of an image different from the input image. Similarly to the first embodiment, the signal component separation unit 101 inputs a U component image signal to the first conversion unit 110, inputs a V component image signal to the second conversion unit 120, and outputs a Y component image signal. Is input to the combining unit 130a.

  The operations of the first conversion unit 110 and the second conversion unit 120 are the same as those in the first embodiment.

  The synthesizing unit 130a includes a U component image signal obtained by the first conversion unit 110, a V component image signal obtained by the second conversion unit 120, and a Y component image input from the signal component separation unit 101. Synthesize the signal. Then, the combining unit 130a combines the obtained image and the input image and outputs the combined image. As a result, an output image in which a digital watermark is embedded in the input image is obtained.

  Below, the detail of the synthetic | combination part 130a is demonstrated.

  The combining unit 130 a includes a signal component combining unit 108 and an image superimposing unit 109.

  Similarly to the first embodiment, the signal component synthesis unit 108 performs U component image signal output from the frequency inverse transform unit 106, V component image signal output from the frequency inverse transform unit 107, and signal component separation. The Y component image signal input from the unit 101 is combined. As a result, an image in which a digital watermark is embedded in an image generated by the image generation unit 140 (that is, an image different from the input image) is obtained. The signal component synthesis unit 108 outputs the obtained image to the image superimposing unit 109.

  The image superimposing unit 109 superimposes the image data output from the signal component synthesis unit 108 on the input image data (that is, performs pixel value addition) to generate an output image.

  In the present embodiment, the same effect as in the first embodiment can be obtained. Further, for example, before an input image is input to the digital watermark embedding device 100a, a digital watermark is embedded in another image in advance, and when the input image is input, the image is immediately output by the image superimposing unit 109. It can also be said that an image is generated. Therefore, the processing at the time of embedding a digital watermark can be speeded up.

  The signal component synthesis unit 108 may amplify the synthesized image signal and then output the image to the image superimposing unit 109. By doing so, the detection accuracy of the digital watermark can be increased.

  In addition, the image superimposing unit 109 compares the image output from the signal component combining unit 108 and the input image for each pixel, and the difference between the U component, the V component, or the Y component between the corresponding pixels exceeds the threshold value. If the difference is present, the correction process may be performed so that the difference does not exceed the threshold value. By doing so, image degradation can be suppressed.

  Further, data of an image different from the input image may be input to the signal component separation unit 101 from the outside. In that case, the image generation unit 140 can be omitted.

  FIG. 6 is a diagram illustrating an example of a hardware configuration of the digital watermark embedding apparatuses 100 and 100a and the digital watermark detection apparatus 200 according to the embodiment of the present invention.

  In FIG. 6, digital watermark embedding apparatuses 100 and 100a and digital watermark detection apparatus 200 are computers, respectively, and include hardware such as an output device 910, an input device 920, a storage device 930, and a processing device 940. The hardware is used by each unit of the digital watermark embedding devices 100 and 100a and the digital watermark detection device 200 (which will be described as “unit” in the description of the embodiment of the present invention).

  The output device 910 is, for example, a display device such as an LCD (Liquid / Crystal / Display), a printer, or a communication module (communication circuit or the like). The output device 910 is used for outputting (transmitting) data, information, and signals by what is described as “unit” in the description of the embodiment of the present invention.

  The input device 920 is, for example, a keyboard, a mouse, a touch panel, or a communication module (communication circuit or the like). The input device 920 is used for inputting (receiving) data, information, and signals by what is described as a “unit” in the description of the embodiment of the present invention.

  The storage device 930 is, for example, a ROM (Read / Only / Memory), a RAM (Random / Access / Memory), a HDD (Hard / Disk / Drive), or an SSD (Solid / State / Drive). The storage device 930 stores a program 931 and a file 932. The program 931 includes a program for executing processing (function) described as “unit” in the description of the embodiment of the present invention. The file 932 includes data, information, signals (values), and the like that are calculated, processed, read, written, used, input, output, etc. by what is described as “parts” in the description of the embodiment of the present invention. It is.

  The processing device 940 is, for example, a CPU (Central Processing Unit). The processing device 940 is connected to other hardware devices via a bus or the like, and controls those hardware devices. The processing device 940 reads the program 931 from the storage device 930 and executes the program 931. The processing device 940 is used for performing calculation, processing, reading, writing, use, input, output, and the like by what is described as “unit” in the description of the embodiment of the present invention.

  In the description of the embodiment of the present invention, what is described as “unit” may be replaced with “circuit”, “device”, and “apparatus”. Further, what is described as “part” in the description of the embodiment of the present invention may be “part” replaced with “process”, “procedure”, and “process”. That is, what is described as a “unit” in the description of the embodiment of the present invention is realized by software alone, hardware alone, or a combination of software and hardware. The software is stored in the storage device 930 as the program 931. The program 931 causes the computer to function as what is described as “unit” in the description of the embodiment of the present invention. Alternatively, the program 931 causes the computer to execute the processing described as “unit” in the description of the embodiment of the present invention.

  As mentioned above, although embodiment of this invention was described, you may implement combining some of these embodiment. Alternatively, any one or some of these embodiments may be partially implemented. For example, only one of those described as “parts” in the description of these embodiments may be employed, or some arbitrary combinations may be employed. In addition, this invention is not limited to these embodiment, A various change is possible as needed.

  100 digital watermark embedding device, 100a digital watermark embedding device, 101 signal component separation unit, 102 frequency conversion unit, 103 frequency conversion unit, 104 vector adjustment unit, 105 vector adjustment unit, 106 frequency reverse conversion unit, 107 frequency reverse conversion unit, 108 signal component synthesis unit, 109 image superimposing unit, 110 first conversion unit, 120 second conversion unit, 130 synthesis unit, 130a synthesis unit, 140 image generation unit, 200 digital watermark detection apparatus, 201 signal component separation unit, 202 second 1 frequency conversion unit, 203 second frequency conversion unit, 204 vector outer product calculation unit, 205 peak extraction unit, 206 bit determination unit, 210 determination unit, 910 output device, 920 input device, 930 storage device, 931 program, 932 file, 940 Processing device.

Claims (6)

In a digital watermark embedding apparatus that embeds a digital watermark in an input image,
A signal component separation unit for separating a first component image signal and a second component image signal from the input image data;
The image signal of the first component obtained by the signal component separation unit is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal adjust the direction, the vector direction of the frequency components of the number corresponding to the data amount of the digital watermark generates the first frequency signal which is a first direction, the said first frequency signal by frequency inversion first a first converter you converted back to 1-component image signal,
The second component image signal obtained by the signal component separation unit is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal. adjust the direction, the second direction vector direction of the number of frequency component corresponding to the data amount of the digital watermark shifted by a predetermined angle on the side corresponding to the data value of the electronic watermark to the first direction a second conversion unit is second to generate a frequency signal, you inverse transformation to the image signal of the second component the second frequency signal by frequency inverse conversion,
The first component image signal obtained by the first conversion unit, the second component image signal obtained by the second conversion unit, and the input image data including the first component and the second component. An electronic watermark embedding apparatus comprising: a combining unit that combines image signals of components included in addition to the components. In a digital watermark embedding apparatus that embeds a digital watermark in an input image,
A signal component separation unit that separates a first component image signal and a second component image signal from data of an image different from the input image;
The image signal of the first component obtained by the signal component separation unit is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal adjust the direction, the vector direction of the frequency components of the number corresponding to the data amount of the digital watermark generates the first frequency signal which is a first direction, the said first frequency signal by frequency inversion first a first converter you converted back to 1-component image signal,
The second component image signal obtained by the signal component separation unit is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal. adjust the direction, the second direction vector direction of the number of frequency component corresponding to the data amount of the digital watermark shifted by a predetermined angle on the side corresponding to the data value of the electronic watermark to the first direction a second conversion unit is second to generate a frequency signal, you inverse transformation to the image signal of the second component the second frequency signal by frequency inverse conversion,
The image signal of the first component obtained by the first conversion unit, the image signal of the second component obtained by the second conversion unit, and the first component and the first image data of the other image An electronic watermark embedding apparatus comprising: a synthesis unit that synthesizes an image signal of a component included in addition to two components and synthesizes the obtained image and the input image . In a digital watermark embedding program for embedding a digital watermark in an input image,
On the computer,
A signal component separation process for separating a first component image signal and a second component image signal from the input image data;
The image signal of the first component obtained by the signal component separation processing is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal adjust the direction, the vector direction of the frequency components of the number corresponding to the data amount of the digital watermark generates the first frequency signal which is a first direction, the said first frequency signal by frequency inversion first a first conversion processing you converted back to 1-component image signal,
The image signal of the second component obtained by the signal component separation process is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal adjust the direction, the second direction vector direction of the number of frequency component corresponding to the data amount of the digital watermark shifted by a predetermined angle on the side corresponding to the data value of the electronic watermark to the first direction and there the second to generate a frequency signal, said second frequency signal by inverse frequency transform you reverse conversion to an image signal of the second component second conversion process,
The first component image signal obtained by the first conversion process, the second component image signal obtained by the second conversion process, and the input image data include the first component and the second component. An electronic watermark embedding program for executing a synthesis process for synthesizing an image signal of a component included in addition to a component. In a digital watermark embedding program for embedding a digital watermark in an input image,
On the computer,
A signal component separation process for separating a first component image signal and a second component image signal from data of an image different from the input image;
The image signal of the first component obtained by the signal component separation processing is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal adjust the direction, the vector direction of the frequency components of the number corresponding to the data amount of the digital watermark generates the first frequency signal which is a first direction, the said first frequency signal by frequency inversion first a first conversion processing you converted back to 1-component image signal,
The image signal of the second component obtained by the signal component separation process is converted into a frequency signal, and a vector of frequency components corresponding to the data amount of the digital watermark among a plurality of frequency components constituting the frequency signal adjust the direction, the second direction vector direction of the number of frequency component corresponding to the data amount of the digital watermark shifted by a predetermined angle on the side corresponding to the data value of the electronic watermark to the first direction and there the second to generate a frequency signal, said second frequency signal by inverse frequency transform you reverse conversion to an image signal of the second component second conversion process,
The first component image signal obtained in the first conversion process, the second component image signal obtained in the second conversion process, and the data of the other image include the first component and the first component. An electronic watermark embedding program , comprising: combining an image signal of a component included in addition to two components; and executing a combining process of combining the obtained image and the input image . In a digital watermark detection apparatus for detecting a digital watermark embedded in an input image,
A signal component separation unit for separating a first component image signal and a second component image signal from the input image data;
A first frequency converting unit that obtains a first frequency signal by frequency converting the image signal of the first component obtained by the signal component separating unit;
A second frequency conversion unit that obtains a second frequency signal by frequency converting the image signal of the second component obtained by the signal component separation unit;
A plurality of first frequency components constituting the first frequency signal obtained by the first frequency converter, and a plurality of second frequency components constituting the second frequency signal obtained by the second frequency converter. A vector outer product calculation unit for calculating a vector outer product of a first frequency component and a second frequency component having the same coordinates in the frequency space, and
Based on the absolute value of the vector outer product calculated by the vector outer product calculation unit, the number of coordinates corresponding to the data amount of the digital watermark is selected from the coordinates in the frequency space, and the vector is selected for each selected coordinate. An electronic watermark detection apparatus comprising: a determination unit that determines a data value of the electronic watermark based on the sign of the vector outer product calculated by the outer product calculation unit. In a digital watermark detection program for detecting a digital watermark embedded in an input image,
On the computer,
A signal component separation process for separating a first component image signal and a second component image signal from the input image data;
A first frequency conversion process for obtaining a first frequency signal by frequency-converting the image signal of the first component obtained by the signal component separation process;
A second frequency conversion process for obtaining a second frequency signal by frequency-converting the image signal of the second component obtained by the signal component separation process;
Of the plurality of first frequency components constituting the first frequency signal obtained by the first frequency conversion process, the number of first frequency components corresponding to the data amount of the digital watermark, and the second frequency conversion process Among the plurality of second frequency components constituting the obtained second frequency signal, the number of second frequency components corresponding to the data amount of the digital watermark is the same as the first frequency component having the same coordinates in the frequency space. A vector cross product calculation process for calculating a vector cross product with two frequency components;
A digital watermark detection program for executing a determination process for determining a data value of the digital watermark based on the vector cross product calculated by the vector cross product calculation process.

Images