JP2000069273A - Coding method and decoding method for watermark - Google Patents

Abstract

PROBLEM TO BE SOLVED: To save a data amount of original image data to be stored in a distribution source by storing the original image data in the form of a compressed image with a smaller data amount. SOLUTION: Original image data as an original image are compressed based on, e.g. JPEG and converted into a VLC code or image compression data. Then the original image data are stored in a form of the VLC code. In order to imbed a watermark in the original image data, VLC the data at first are subjected to decoding and inverse quantization and the data are expanded into spectral data SPC through the processing above. Then the watermark, including code copyright information or the like, is imbedded in the spectral data SPC. Then the spectral data SPC the watermark are converted into a VLC code through quantization and VLC coding. Then the image data are distributed in the form of the VLC code, that is, in a form of compressed data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of encoding and decoding a watermark (digital watermark).
In particular, the present invention relates to an improvement for reducing a data amount (signal amount) of original image data to be stored.

[0002]

2. Description of the Related Art Watermark technology is a technique for embedding copyright and purchaser information in image data, which is data (signals) representing an image, in a form that is insensitive to the user, thereby preventing illegal copying. This technology prevents the next use. The distributor of the copyrighted image data stores the image data before embedding the watermark, that is, the original image data, and collects the image data suspected of illegal copying from the city when collecting it. By comparing with the original image data, the watermark is extracted, and further,
Decrypt to copyright information etc. The distribution source determines the purchaser based on the decrypted information.

FIGS. 10 and 11 are flow charts showing processing procedures of a conventional watermark encoding method and watermark decoding method, respectively. FIG.
2 is a timing chart showing the transition of the image format in these processes. These processes are executed by a distribution source (or a person authorized by the distribution source).

In the conventional watermark encoding method, original image data in the form of an uncompressed original image is
For example, it is compressed based on an image compression technique such as JPEG. Thereby, the image data is distributed in the form of the compressed image data. During this compression process, watermark embedding is performed. On the other hand, the above-described original image data is stored in the form of an original image.

The watermark encoding method will be described with reference to FIGS. 10 and 12. For example, original image data as an original image is converted into spectral data (spatial spectral data) by DCT transform which is a process of JPEG compression. The data is converted into SPC (step S51). Spectral data includes a number of spectral components as its constituent elements. Embedding of a watermark in which copyright information or the like is encoded is performed on the spectrum data SPC (step S52). Spectrum data SP
C is obtained by expanding the original image into a set of spectral components, that is, a set of spatial frequency components, and the DCT transform is a representative example of such expansion.

Subsequently, the spectrum data SPC including the watermark is converted into a VLC code (variable length code) through quantization and VLC coding (variable length coding) (step S53). Then, the image data is distributed to the market through a distributor or the like in the VLC code format, that is, the compressed image data format (step S5).
4). Also, the original image data remains in the original image format,
It is stored at the distribution source (step S55).

Next, a conventional watermark decoding method will be described with reference to FIGS. The image data collected in the market may be in the form of a VLC code or in the form of an original image. Regardless of which format is used, the image data is converted into spectral data SPC '(step S61).

For example, if the sampled image data is in the form of a VLC code, VLC decoding, which is the inverse operation of VLC coding,
Then, through inverse quantization which is an inverse operation of quantization, the data is converted into spectrum data SPC ′. If the collected image data is in the form of an original image, it is converted into spectrum data SPC ′ through DCT conversion.

The spectrum data SPC 'obtained through step S61 has a different value from the spectrum data SPC shown in FIG. 10 because it contains a watermark. By calculating the difference between the two, the watermark embedded in the spectrum data SPC 'is extracted (step S62). Next, decoding of the extracted watermark is performed (step S).
63) Accordingly, copyright information and the like are read.

[0010] In general, the collected image data is not the same as the image data immediately after distribution, but may have gone through various image processing. For this reason, there is a possibility that noise derived from such image processing is mixed in the collected image data. This noise is also shown in FIG.
Spectrum data SPC of FIG. 11 and spectrum data SP of FIG.
It causes a deviation from C ′.

In order to eliminate the adverse effect of noise in extracting the watermark, in the process of embedding the watermark in FIG. 10 (step S52), the watermark is repeatedly embedded in one image. In the process (step S62), a method of increasing the S / N ratio by calculating an average of repeated watermarks is employed.

Alternatively, in the process of embedding the watermark (step S52), an error correction code is included in the watermark, and the process of taking out the error (step S6).
In 2), by detecting and correcting the decoding error,
A method of increasing the reliability of the watermark to be taken out has been adopted. In any method, in step S63, the presence or absence of reliability is determined, and when it is determined that there is reliability, copyright information and the like are output.

[0013]

However, according to the conventional method, since the original image data is stored in the form of the original image, a distribution source that normally needs to store many types of original image data is: There is a problem that the total data amount of the original image data to be stored becomes large. Large data volumes are inconvenient for storage and handling,
In addition, the required cost increases.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems in the conventional method, and provides a watermark encoding method and a decoding method capable of reducing the amount of original image data to be stored. The purpose is to:

[0015]

According to a first aspect of the present invention, there is provided a watermark encoding method for embedding a watermark in original image data, comprising the steps of: (a) converting the original image data into spectral data; Converting the uncompressed original image format to compressed image data format through image compression including: (b) storing the original image data in the compressed image data format; and (c) the quantum Converting the original image data from the format of the compressed image data to spectral data through image expansion including inverse quantization as an inverse operation of the conversion, (d) the spectrum obtained in the step (c) (E) overlaying the watermarked data in step (d) with the spectral data through the compressed image data through image compression including quantization. Includes a step of converting to another format, and a step of delivering the compressed image data obtained by converting by (f) said step (e). Moreover, the step (d) includes (d-1) a step of superimposing a random fluctuation of a sign on the spectral data obtained in the step (c), and (d-2) the step (d-1). A) obtaining a quantized coefficient by performing quantization on the spectral data on which the random fluctuations are superimposed, using a scaling coefficient having a larger value than the scaling coefficient used in the quantization; (d-3) configuring the spectral data, from among the spectral components giving a non-zero value as the quantization coefficient, selecting a watermark embedding target, (d-4) the step (d- Superimposing the watermark on the value before the fluctuation of the spectral component selected in 3) is superimposed,
It has.

According to a second aspect of the present invention, there is provided a watermark encoding method for embedding a watermark in original image data, comprising: (a) converting the original image data into image data including conversion into spectral data and quantization; Converting the uncompressed original image format to compressed image data format, (b) storing the original image data in the compressed image data format, and (c) performing the inverse operation of the quantization. Through image expansion including inverse quantization, the original image data is
Converting the compressed image data format to spectral data, (d) superimposing a watermark on the spectral data obtained in the step (c), and (e) watermarking in the step (d). The spectral data is superimposed, through the image compression including the quantization, a step of converting to the format of the compressed image data, (f) the compressed image data obtained by converting in the step (e) And delivering the data. Moreover, the step (d) comprises:
(d-1) the amplitude of the watermark to be superimposed,
Setting a value larger than 1/2 of the quantization step used in the quantization.

According to a third aspect of the invention, in the watermark encoding method of the second aspect, the step (d) is performed by (d-2) a predetermined constant L (≦ １ ／) which is set in advance. On the other hand, when the amplitude set in the step (d-1) is smaller than L times the value of the quantization table, the amplitude is changed to a predetermined constant K (L ≦ K ≦ 1
/ 2) a step of setting the value to twice the value.

According to a fourth aspect of the present invention, there is provided a watermark decoding method for extracting a watermark from collected image data, wherein (A) the step (b) of any of the first to third watermark encoding methods. Performing the same step as the step (c) on the original image data stored by the step of obtaining spectral data, and (B) converting the collected image data into spectral data And (C) calculating the difference between the two spectral data obtained in the steps (A) and (B), thereby extracting the watermark from the spectral data obtained in the step (B). And

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS 1. Schematic Procedure of Encoding and Decoding Methods In order to reduce the amount of original image data to be stored, in a preferred embodiment of the present invention, the format of an original image is Instead, the original image data is stored in the VLC code format, that is, in the format of the compressed image data. FIG. 1 and FIG. 2 are flowcharts showing the processing procedure of the watermark encoding method and the watermark decoding method of this embodiment, respectively. FIG. 3 is a timing chart showing the transition of the image format accompanying these processes. These processes are executed by a distribution source (or a person authorized by the distribution source). FIGS. 1 to 3 are each compared with FIGS. 10 to 12 showing the conventional method.

The watermark encoding method according to the embodiment will be described with reference to FIGS. 1 and 3. The original image data as the original image is image-compressed based on, for example, JPEG.
The VLC code is converted into compressed image data (step S1). The original image data is stored in the VLC code format (step S2). That is, the original image data is stored in an image format having a small data amount. As a result, the data amount of the original image data to be stored is reduced, so that the problem in the conventional watermark coding method is reduced or eliminated.

In order to embed a watermark in the original image data, first, the original image data having the format of the VLC code is subjected to VLC decoding and inverse quantization, and through these, the spectral data is converted to the SPC. Is image-expanded (step S3). Next, a watermark including the encoded copyright information is displayed on the spectrum data SPC.
Is embedded (step S4).

Subsequently, the spectral data SPC including the watermark is converted to VLC data through quantization and VLC coding.
It is converted into a code (step S5). And this VL
The image data is distributed to the market through a distributor or the like in the format of the C code, that is, the format of the compressed image data (step S6).

Next, a watermark decoding method corresponding to the method of FIG. 10 will be described with reference to FIGS. The image data collected in the city will be
It may be in the form of a C code, or it may be in the form of an original image. Regardless of which format is used, the image data is converted into spectral data SPC '(step S11).

Next, VLC decoding and inverse quantization are performed on the original image data in the VLC code format, and spectrum data SPC is obtained through them (step S).
12). This processing is performed in the same manner as the processing in step S3. The spectrum data SPC 'obtained through step S11 has a different value from the spectrum data SPC obtained in step S12 because it includes a watermark.

By calculating the difference between the two, the watermark embedded in the spectrum data SPC 'is extracted (step S13). Next, the extracted watermark is decrypted (step S14), whereby copyright information and the like are read. Steps S11, S13, and S14 are performed in the conventional steps S61, S62, and S63, respectively.
This is performed in the same manner as (FIG. 11). Therefore, a detailed description of these processes is omitted.

<2. Problems to be Improved in Coding Method>
According to the method described above, the problem that occurs during practical use that the amount of original image data to be stored is excessive,
Alleviated or eliminated. However, the process of embedding the watermark (step S4) is different from the conventional step S4.
If it is executed in the same manner as in FIG. 52 (FIG. 10), there is a new problem that the image quality is significantly deteriorated or the decoding accuracy rate is deteriorated. Here, this problem will be described.

The cause of the problem that the image quality deteriorates is as follows.
In the process of determining the position where the watermark is to be embedded in the image of one screen, the cause of the problem that the decoding accuracy rate decreases is in the process of determining the amplitude of the watermark to be embedded. . Both processes are included in the process of step S4.

When performing compression processing such as JPEG on image data, DC data is divided into blocks each of which is obtained by dividing one screen.
Processing such as T-transformation, quantization, and VLC code is performed.
The same applies to the reverse operation. Each block is composed of, for example, 8 × 8 pixels or 16 × 16 pixels.

As a position where the watermark is to be embedded, a specific block (generally a plurality) satisfying a certain condition is selected from among the plurality of blocks, and a specific spectral component in the selected block is selected. It is. The spectrum component is a component of the spectrum data, and the spectrum data is constituted by a set of the spectrum components. As described above, the “position where the watermark is embedded” includes both the spatial position on the screen and the position in the spectrum.

More specifically, for example, four AC components having the lowest spatial frequency (four AC low frequency components) are selected as specific spectral components. Of the various spectral components, the low-frequency component has the greatest effect on the image quality, so it is difficult to remove the watermark without deteriorating the image quality by image processing such as filtering. is there. That is, it is to increase the strength (removability) of the watermark.

As the specific block, for example, the above A
A block in which the amplitude of each of the four C low-frequency components is somewhat large and the high-frequency components are somewhat large is selected. The condition that the amplitude of the AC4 component is large is set to prevent the watermark from disappearing due to quantization after embedding the watermark and to make the watermark visually inconspicuous. The condition that the high frequency component is large to some extent is to make the watermark less noticeable.

These conditions per se regarding a specific block and a specific spectral component selected as a position where a watermark is to be embedded are well known in the art. These three
Among the conditions, the condition that the amplitude of the AC4 component is large is as follows.
More specifically, a larger scaling factor
When provisional quantization is performed using the SC, the determination is made under the condition that the obtained quantization coefficient Q does not become zero. “Large” means that the scaling coefficient SC is somewhat (for example, 1) larger than a scaling coefficient SC used for a quantization process performed to finally obtain a VLC code.

The scaling coefficient SC is one of the factors defining the quantization step. With image compression algorithms such as JPEG, it is common to all blocks,
By multiplying the value of the quantization table TB prepared by the algorithm itself by a scaling coefficient SC common to all the spectral components belonging to one block so that each spectral component has a distinct value, the quantization width A quantization step Δ corresponding to (roughness) is given. That is, Δ = SC × TB (Formula 1) The magnitude of the scaling coefficient SC is appropriately set to an appropriate value according to the data amount of the VLC code obtained by compression.

As described above, by performing quantization using a relatively large scaling coefficient SC, A
It is determined whether or not the value of the quantization coefficient Q of the four C low-pass components is zero. At this time, the spectral data SPC to be embedded with the watermark is not directly obtained from the original image as shown in FIG. 3, but is converted into a VLC code and then inversely converted. And was obtained through

That is, the spectral data SPC to be embedded with the watermark is obtained through image compression and image expansion along the path A indicated by the dotted line in FIG. The quantization process is an irreversible process, and even if the quantization and the subsequent inverse quantization are performed using the same quantization step Δ, the spectrum data SPC is the same as the spectrum data SPC before quantization. They do not play the same.
That is, the route A is irreversible.

For this reason, for example, the scaling coefficient SC =
1, when the spectral data SPC obtained by performing quantization and inverse quantization along the path A is quantized with the scaling coefficient SC = 2, the spectral component and its quantization coefficient Q Are drawn by dotted lines in the graph of FIG. In FIG. 4, the horizontal axis X represents the value of the spectrum component. In the graph of FIG. 4, the scaling coefficient SC is added to the spectrum data SPC directly obtained from the original image without passing through the path A.
The relationship between the spectral component X and the quantization coefficient Q when quantization is performed with = 1 and SC = 2 is also shown.

As shown in FIG. 4, the scaling coefficient SC =
Spectrum data SPC once quantized by 1
A deviation appears between the value of the quantization coefficient Q obtained by quantization with the scaling coefficient SC = 2 between the spectrum data SPC and the other spectrum data SPC. This means that a deviation appears at the position where the watermark is to be embedded between the two spectral data SPCs. With this deviation,
The problem is that the distortion of the image due to the watermark is visually noticeable.

The cause of the problem that the correct answer rate decreases will be explained with reference to FIG. Since the spectral data SPC to be embedded with the watermark has already undergone the processing of quantization and inverse quantization, each spectral component X is an integer of the quantization step Δ as shown by a white circle in FIG. It can only take twice the value. After the watermark is embedded, the quantization is performed using the quantization step Δ of the same size. Therefore, if the amplitude of the watermark is not appropriately selected, the watermark will be generated after the quantization. No mark will remain. This causes a problem that the decoding accuracy rate is deteriorated.

The broken line in FIG. 5 shows the relationship between the spectral component X that has not been quantized and the quantization coefficient Q. In the watermark encoding method according to the embodiment, as described below, a watermark embedding process (step S
4) is executed.

<3. Watermark Embedding Processing> FIG.
9 is a flowchart showing an internal flow of a process for embedding a watermark (step S4). In this processing, a luminance component Ysp of the spectrum data SPC, a scaling coefficient SC, a watermark WM, and a random number PN are supplied as processing targets. The luminance component Ysp is one of the three elements included in each spectral component constituting the spectral data SPC. As the random number PN, any value of +1 or -1 (in other words, a positive or negative sign)
Supplied randomly. Note that, in the following flowchart, a left-pointing arrow included in a mathematical expression drawn in each step indicates that the value on the right side is substituted for the variable on the left side.

When the process is started, in step S21, a minute numerical value having a random number PN as a code is assigned to a variable δ. For example, the variable δ is defined by δ = PN × 10 ⁻³ (Equation 2). Then, as a new variable Ytm, Ytm = Ysp + δ (Equation 3) is defined. That is, a fluctuation corresponding to the variable δ is added to the luminance component Ysp, and in order to make this a temporary luminance component,
The variable Ytm is introduced.

Next, the processing of steps S22 to S26 is repeatedly executed for each block. Step S2
2, the variable Y corresponding to the luminance component Ysp including the fluctuation δ
tm is quantized, and further inversely quantized. That is, the quantization coefficient Q is calculated based on Q = round (Ytm / (TB × 2)) (Equation 4). Thereafter, the quantization coefficient Q is subjected to inverse quantization based on Qsp = Q × (TB × 2) (Equation 5) to obtain spectrum data Qsp as an inverse quantization coefficient.

Here, steps S1 and S5 in FIG.
Scaling factor SC used for quantization performed by
However, as an example, it is assumed that SC = 1. Therefore, in Equations (4) and (5), the scaling coefficient SC is “1”.
It is set to “2”, which is one step larger than that. Equation (4),
The two variables Q and Qsp obtained in (5) are used in the next step S
23.

The quantization coefficient Q obtained by the equation (4) is a value calculated based not on the luminance component Ysp itself but on the variable Ytm on which the fluctuation (noise) δ is superimposed. For this reason, as shown in FIG. 7, the relationship between the luminance component Ysp and the quantization coefficient Q is distributed in two ways according to the direction of fluctuation, that is, the sign of the variable δ. At the center of the variance, the relationship (indicated by a dashed line) between the luminance component Ysp obtained by directly converting the original image without quantization and the quantization coefficient Q is located.

Therefore, the quantization coefficient Q of the variable Ytm including random fluctuations (random noise) coincides on average with the quantization coefficient Q of the luminance component Ysp that has not been quantized. For this reason, the problem of the deterioration of the image quality described above is solved.

In step S23, the embedding position of the watermark is determined. Then, an identification number IND is provided for each determined position. The internal flow of step S23 is represented by the flowchart of FIG. In step S23, the quantization coefficient Q and the spectrum data Qs
p is processed and the identification number IND is output. That is, in step S31, the following three conditions are set as specific spectral components of a specific block to which an identification number IND is to be assigned: The spectral components belong to four AC low-frequency components (condition 1) The quantization coefficient Q is not zero (Condition 2) The sum of absolute values of the AC components of the spectrum data Qsp is equal to or more than a certain number (for example, 200) (Condition 3).

These conditions specifically express the above-mentioned conditions relating to the specific block and the specific spectral component. When the processing in step S31 ends, the selected identification number IND is output, and the processing in step S23 is completed. Thereafter, the process proceeds to step S24 in FIG.

In step S24, it is determined whether or not the identification number IND has reached a number sufficient to embed the watermark WM. More specifically, it is determined whether or not each bit forming the watermark includes two or more bits. If the result of the determination is "No", a message to that effect is output in step S25. Thereafter, the processing may be terminated or may be continued. On the other hand, if the result of the determination is “Yes”, step S
At 26, the amplitude of the watermark is determined. The determined value of the amplitude is assigned to a new variable WMs.

FIG. 9 is a flowchart showing the internal flow of step S26. In this process, the scaling coefficient SC is processed, and the amplitude WMs of the watermark is output. First, in step S41, the scaling coefficient SC is set to a value slightly smaller than 1 (for example, 1/1.
It is determined whether it is smaller than 1). The result of the judgment is
If "Yes", in step S42, the temporary variable W
The value of Ms is given by WMs = 0.5 (Equation 6). On the other hand, if the result of the determination is “No”,
In step S43, the value of the temporary variable WMs is given by WMs = (SC / 2) × 1.1 (Equation 7).

Next, in step S44, the final value of the variable WMs to be output is given by WMs = WMs × TB (Equation 8). The equation in Equation 8 differs from the mathematical equation concept, and its equal sign is synonymous with the left arrow used in the equations in the flowchart.

The variable WMs given by Expression 8 is treated as the amplitude of the watermark to be embedded in step S29 described later. That is, step S26
Is based on the fact that the spectrum data SPC can only take the discontinuous value indicated by the white circle in FIG. 5, and as the watermark amplitude, the quantization step Δ (= SC × TB)
This means that a value somewhat larger than 1/2 times is given. As a result, the fear that the watermark WM disappears due to the quantization is eliminated, and the problem that the decoding accuracy rate is deteriorated does not occur.

If the value of the temporary variable WMs calculated by the equation (7) is 0.5 or less, the distortion of the image resulting from the embedding of the watermark cannot be visually recognized from experience. It was found through simulation experiments.
Therefore, when the value of the temporary variable WMs is originally 0.5 or less, it is forcibly fixed to the value of 0.5 (steps S41 and S42).

It is useful to increase the watermark amplitude as much as possible without visually observing the distortion of the image, in order to increase the strength of the watermark. The value of the temporary variable WMs is 0.5,
The amplitude WMs finally output is equivalent to WMs = 0.5 × TB (Equation 9).

In general, a constant L given by 0 <L ≦ 0.5 (Equation 10) is set, and the value of the temporary variable WMs calculated by Equation 7 is expressed as WMs <L. When (Equation 11) is satisfied, the value of the constant L may be forcibly fixed. The fact that the value of the temporary variable WMs is L is equivalent to that the finally output amplitude WMs is WMs = L × TB (Equation 12). Among various values of the constant L (≦ 0.5), L = 0.5 is the most desirable form in that the strength of the watermark is maximized.

More generally, when the condition of Expression 11 is satisfied, the temporary variable WMs may be set to a constant K given by L ≦ K ≦ 0.5 (Expression 13). The fact that the value of the temporary variable WMs is K is equivalent to the finally output amplitude WMs being WMs = K × TB (Equation 14).

When the processing in step S44 is completed, the calculated amplitude WMs is output, and the processing in step S26 is completed. Thereafter, the process proceeds to step S27 in FIG.

In step S27, the binary data forming the watermark WM is converted from a format represented by "0" and "1" to a format represented by "-1" and "1". You. Further, the watermark WM is copied for each identification number IND. Next, in step S28, the watermark WM is multiplied by the random number PN. In the following step S29, the watermark WM is multiplied by the amplitude WMs calculated in step S26. As a result, a final watermark WM to be embedded is obtained.

Next, the process of step S30 is repeatedly executed for each identification number IND. In step S30, a watermark is added to the luminance component Ysp for each identification number IND.
The WM is superimposed to obtain a luminance component Ysp 'after embedding the watermark. When watermarks are embedded in all the identification numbers IND, step S4 ends, and the process proceeds to step S5 (FIG. 1).
Since the processes of steps S24 and S27 to S30 are well known in the conventional watermark encoding method (FIG. 10), detailed description thereof will be omitted.

[0059]

According to the first aspect of the present invention, the original image data is stored in the form of a compressed image having a small data amount. The amount of original image data to be stored is reduced. That is, the burden on the distribution source for storing the original image data is reduced. Moreover, a random fluctuation is added to the spectrum data to search for a spectral component that can be a watermark embedding target, and then quantization is performed in a quantization step larger than the steps (a) and (e). Then, a spectral component in which the value of the obtained quantization coefficient does not become zero is found.

Therefore, the deterioration of the image quality due to the deviation appearing between the two spectral data obtained in the irreversible processes (a) and (c) is suppressed. Also, since watermark embedding is performed on the spectral components before the fluctuation is superimposed, the influence of the fluctuation does not remain in the distributed image data.

In the method of the second invention, the original image data is
Since the image data is stored in the form of a compressed image having a small data amount, the data amount of the original image data to be stored can be reduced at a distribution source that normally needs to store many types of original image data. That is, the burden on the distribution source for storing the original image data is reduced. In addition, the amplitude of the watermark to be superimposed is set to a value larger than half the quantization step used in the quantization in steps (a) and (e), so that the embedded watermark is The inconvenience of disappearing after the conversion can be avoided.
This also avoids the problem that the decoding accuracy rate decreases.

In the method according to the third invention, the amplitude of the watermark to be superimposed is equal to the constant L of the value of the quantization table.
The value is set so as not to fall below (≦）) times. As a result, the strength of the watermark is increased within a range where the image distortion is not visually recognized.

In the method of the fourth invention, it is determined whether or not the collected image data is an illegal copy of the image data distributed after being encoded by the method of the first to third inventions. be able to.

[Brief description of the drawings]

FIG. 1 is a flowchart of a watermark encoding method according to an embodiment.

FIG. 2 is a flowchart of a watermark decoding method according to the embodiment;

FIG. 3 is a timing chart of the procedure in FIGS. 1 and 2;

FIG. 4 is a graph illustrating a first factor of a decrease in a decoding accuracy rate.

FIG. 5 is a graph illustrating a second cause of a decrease in the decoding accuracy rate.

FIG. 6 is a flowchart showing an internal flow of step S4 in FIG. 1;

FIG. 7 is a graph illustrating the reason why the first factor is eliminated.

FIG. 8 is a graph showing an internal flow of step S23 in FIG.

FIG. 9 is a graph showing an internal flow of step S26 in FIG.

FIG. 10 is a flowchart of a conventional watermark encoding method.

FIG. 11 is a flowchart of a conventional watermark decoding method.

FIG. 12 is a timing chart of the procedure of FIGS. 10 and 11;

[Explanation of symbols]

 δ Fluctuation Δ Quantization step Q Quantization coefficient SPC, SPC 'Spectrum data SC Scaling coefficient TB Quantization table WM Watermark

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI theme coat ゛ (reference) // H04N 7/08 H04N 7/08 Z 7/081 F term (reference) 5B057 AA20 CA16 CB16 CC03 CE08 CE09 CG05 CG07 CH01 DA17 DB02 5C059 KK43 MA00 MA23 MC11 MC26 ME01 SS12 UA02 UA38 5C063 AA01 AB03 AC02 CA23 CA29 CA36 DB09 5C076 AA14 AA40 BA09 5C078 BA21 BA57 BA62 CA14 DA01 DA02 DB11

Claims (4)

[Claims] 1. A watermark encoding method for embedding a watermark in original image data, comprising the steps of: (a) converting the original image data into an uncompressed original image format through image compression including conversion to spectral data and quantization; From the step of converting to a format of compressed image data, and (b) storing the original image data in the format of the compressed image data,
(c) converting the original image data from the format of the compressed image data to spectral data through image decompression including inverse quantization as an inverse operation of the quantization, (d)
A step of superimposing a watermark on the spectral data obtained in the step (c), and (e) the spectral data in which the watermark is superimposed in the step (d), through image compression including the quantization, Converting the compressed image data into a format of compressed image data; and (f) distributing the compressed image data obtained by the conversion in the step (e), wherein the step (d) comprises: (d- 1) a step of superimposing a random fluctuation with a sign on the spectrum data obtained in the step (c), and (d-2) the spectrum data on which the random fluctuation is superimposed in the step (d-1). For, a step of obtaining a quantization coefficient by performing quantization by using a scaling coefficient having a larger value than the scaling coefficient used in the quantization, (d-3) configuring the spectral data, As the quantization coefficient From the spectral components to provide a non-zero value,
Selecting the embedding target of the watermark, and (d-4) superimposing the watermark on a value before the fluctuation of the spectral component selected in the step (d-3) is superimposed. And a watermark encoding method comprising: 2. A watermark encoding method for embedding a watermark in original image data, comprising: (a) converting the original image data into an uncompressed original image format through image compression including conversion into spectral data and quantization; From the step of converting to a format of compressed image data, and (b) storing the original image data in the format of the compressed image data,
(c) converting the original image data from the format of the compressed image data to spectral data through image decompression including inverse quantization as an inverse operation of the quantization, (d)
A step of superimposing a watermark on the spectral data obtained in the step (c), and (e) the spectral data in which the watermark is superimposed in the step (d), through image compression including the quantization, Converting the compressed image data into a format of compressed image data; and (f) distributing the compressed image data obtained by the conversion in the step (e), wherein the step (d) comprises: (d- 1) The amplitude of the watermark to be superimposed is
Setting a value larger than の of the quantization step used in the quantization. 3. The watermark encoding method according to claim 2, wherein said step (d) comprises: (d-2) a predetermined constant L (≦ １ ／) set in advance.
In contrast, when the amplitude set in the step (d-1) is smaller than L times the value of the quantization table,
The amplitude is calculated as a predetermined constant K (L ≦ L) of the value of the quantization table.
A watermark encoding method, further comprising the step of setting the value to a value of (K ≦ １ ／) times. 4. A watermark decoding method for extracting a watermark from collected image data, comprising: (A)
By performing the same step as the step (c) on the original image data stored in the step (b) of the watermark encoding method according to any one of claims 1 to 3, spectral data can be obtained. Obtaining, and (B) converting the collected image data to spectral data,
(C) by calculating the difference between the two spectral data obtained in the steps (A) and (B), the step of extracting the watermark from the spectral data obtained in the step (B), Watermark decoding method comprising: