JP2003169309A - Content feature amount extraction device and program therefor, content verification program and method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a content feature amount extraction device and a program therefor, which is used for content verification data and extracts feature amount capable of specifying the content, and to provide a content verification data generating device and a program therefor, which is capable of surely verifying the content and authenticating a content distributor. <P>SOLUTION: The content feature extraction device 1 is so formed as to include a contracted image generating section 10 provided with a two-dimensional block separating means 11, a two-dimensional block average value calculating means 13 and an image synthesizing means 15, a two-dimensional DCT (discrete cosine transformation) section 20 (frequency component converting means), and a DCT coefficient selecting section 30 (feature amount output means). <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to content authentication management technology, and more particularly, to a content feature amount extraction apparatus for extracting a feature amount of a content, a program therefor, and content authentication for performing content distributor authentication and content authentication. The present invention relates to a data generation device, its program, and a content authentication method.

[0002]

2. Description of the Related Art In recent years, moving image contents (hereinafter referred to as contents) have been digitized, and in particular, all contents distributed on the Internet are digital data. With the development of personal computers,
The number of cases where this digitized content is illegally distributed is increasing. Especially if the content is paid,
The user who receives the distribution of the content needs to confirm whether the content is a defective product. In order to make this confirmation, the user must compare the delivered content with the original content owned by the original distributor of the content to perform authentication.

Conventionally, as a method of performing this authentication, for example, the encrypted content obtained by encrypting the content is distributed to the user, and the user supplies a key for decrypting the encrypted content from the original distributor of the content. It is possible to decrypt the encrypted content with the key only if the encrypted content is delivered by a legitimate content distributor, and you can impersonate the legitimate content distributor. When the content is distributed by another content distributor, the authentication is performed by utilizing the fact that the content cannot be decrypted.

As another method, a part of the content is cut out, and S of the data of the part is cut out.
There is a method in which a hash value is calculated by a hash function such as HA (Secure Hash Algorithm), the hash value is distributed with the content, and the user checks the hash value by a dedicated playback device or dedicated playback software for the content.

There is also a method of embedding information unique to the original distributor of the content in the content as a digital watermark, and performing content distributor authentication and content authentication based on the embedded information. In this method, the content original distributor embeds the content original distributor's ID or unique information determined by the content original distributor as a digital watermark in the content to be distributed, and distributes the content in which the digital watermark is embedded to the user. After that, the user who has delivered the content in which the digital watermark is embedded acquires (downloads) the software for detecting the digital watermark from the original distributor of the content through the network, and the software detects the digital watermark and detects the original content. Content distributor authentication and content authentication have been performed by extracting information unique to the distributor.

As described above, in order to perform content distributor authentication and content authentication, an encryption method, a hash function method, or a digital watermark embedding method is generally used.

[0007]

However, according to the above-described conventional technique of performing encryption, the decrypted content cannot be authenticated, and the once decrypted content is easy to be authenticated. There was a problem that tampering would be possible. Further, in the method using the hash function which is the conventional technique, the user needs a dedicated device or software for checking the hash value distributed with the content, and the setting of the device or the installation work of the software is required. For example, the user is overloaded.

Further, in the conventional technique of embedding a digital watermark, there is a possibility that the digital watermark disappears by processing the content, for example, compressing the content when the content is distributed. is there.
This problem can be solved by increasing the embedding strength of the digital watermark (increasing the amount of change in the pixel value), but instead, the problem of degrading the quality of the content occurs.

The present invention has been made in order to solve these problems, and provides a content feature amount extraction apparatus and a program thereof which are used as content authentication data and extract a feature amount capable of identifying a content. An object of the present invention is to provide a content authentication data generation device and its program, and a content authentication method, which can provide the content and the content distributor surely.

[0010]

The present invention was devised to achieve the above-mentioned object. First, the content feature amount extraction apparatus according to claim 1 is a moving image input in chronological order. A content feature amount extraction device for extracting a feature amount of a moving image content based on an image frame of the image content, and a reduced image generation unit for generating a reduced image of the image frame from the image frame, and the reduction. A frequency component conversion unit that converts the reduced image generated by the image generation unit into a frequency component, and a feature that outputs part or all of the data converted by the frequency component conversion unit as the feature amount of the moving image content. And a quantity output means.

According to this structure, the content feature amount extraction device generates the feature amount based on the image frames of the moving image content input in chronological order. Therefore, the feature amount is the time of the moving image content. It is a feature quantity that captures the physical characteristics. Further, since the content feature amount extraction device generates the frequency component of the reduced image as the feature amount from the reduced image generated from the image frame, the feature amount is
It is a feature amount that captures the planar feature of the image frame. As a result, the content feature amount extraction device can generate a feature amount that captures temporal and planar features of the moving image content, and this feature amount is used as authentication data for authenticating the content, for example. can do.

According to a second aspect of the present invention, there is provided the content feature amount extracting apparatus according to the first aspect, wherein the reduced image generating means divides the image frame into a plurality of pixel data. The reduced image is generated by extracting the average value of the respective pixel values as the representative pixel value.

According to this structure, the content characteristic amount extraction device generates a reduced image by using the average value of the pixel values of each divided area obtained by dividing the image frame as the representative pixel value. Since the average value of the pixels is a value that does not change the most due to compression of MPEG2 or the like, the feature amount of the moving image content generated by this content feature amount extraction device changes even if the moving image content is compressed or expanded. It becomes a feature amount that does not
It can be used as authentication data for authenticating the content.

Further, in the content feature quantity extraction device according to claim 3, in the content feature quantity extraction device according to claim 1, the reduced image generation means extracts each image frame from the plurality of input image frames. By generating a divided area divided for each of a plurality of pixel data and extracting the average value of each pixel value of a divided area group formed of a plurality of divided areas existing at the same position of each image frame as a representative pixel value. The configuration is such that a reduced image is generated.

According to this structure, the content characteristic amount extraction device calculates the average value of each pixel value of the divided area group which is present at the same position as the divided area into which the image frame is divided and which is composed of the divided areas of the plurality of image frames. A reduced image is generated as a representative pixel value. As a result, the content feature amount extraction device can reduce the amount of calculation for generating one reduced image from a plurality of image frames and generating the frequency component by the frequency component conversion means. Further, since the average value of the pixels calculated for generating the reduced image is a value that does not change the most due to compression such as MPEG2, the feature amount of the moving image content generated by this content feature amount extraction device is It becomes a feature amount that does not change even if the moving image content is compressed / decompressed, and can be used as authentication data for authenticating the content.

Furthermore, the content feature quantity extraction device according to claim 4 is any one of claim 1 to claim 3.
In the content feature amount extraction device described in the item 1, the frequency component conversion unit is configured to convert the reduced image generated by the reduced image generation unit into a frequency component by two-dimensional orthogonal transformation.

According to this structure, the content feature quantity extraction device uses the frequency component conversion means to convert the reduced image into two.
By the dimensional orthogonal transformation, it is transformed into a frequency component which becomes a feature amount of the image. For example, the content feature amount extraction device converts the reduced image into frequency components using discrete cosine transform. As a result, the content feature amount extraction device can generate a feature amount that captures the planar feature of the moving image content.

The content feature quantity extraction device according to claim 5 is the content feature quantity extraction device according to any one of claims 1 to 3, wherein the frequency component conversion means is a reduced image generation means. A plurality of reduced images generated from the above are converted into frequency components by three-dimensional orthogonal transformation.

With such a configuration, the content feature quantity extraction device performs three-dimensional orthogonal transformation on the plurality of reduced images by the frequency component conversion means, so that the plurality of reduced images also include temporal changes. It is converted into the frequency component which becomes the feature amount of the image. Here, the three-dimensional orthogonal transformation means performing a two-dimensional orthogonal transformation on the reduced image and further performing an orthogonal transformation for a plurality of reduced images for each element of the frequency component. As a result, the content feature amount extraction device can generate a feature amount that captures the planar and temporal features of the moving image content.

Further, the content characteristic amount extraction program according to the sixth aspect of the present invention is a program for extracting a characteristic amount of a moving image content based on an image frame of the moving image content input in chronological order. A reduced image generation unit that generates a reduced image of the image frame from the image frame, a frequency component conversion unit that converts the reduced image generated by the reduced image generation unit into frequency components, and a frequency component conversion unit that converts the reduced image. A part or all of the data is made to function as a feature amount output means for outputting as the feature amount of the moving image content.

According to this structure, the content feature amount extraction program generates the feature amount based on the image frames of the moving image content input in chronological order. Therefore, the feature amount is the time of the moving image content. It is a feature quantity that captures the physical characteristics. In addition, since the content feature amount extraction program generates the frequency component of the reduced image as the feature amount from the reduced image generated from the image frame,
The feature amount is a feature amount that captures the planar feature of the image frame. As a result, the content feature amount extraction program can generate a feature amount that captures the temporal and planar features of the moving image content, and this feature amount is used as authentication data for authenticating the content, for example. can do.

The content feature quantity extraction device according to claim 7 is a content feature quantity extraction device that extracts a feature quantity of a moving image content based on image frames of the moving image content input in chronological order. A partial area extracting unit that extracts one or more partial areas that are rectangular areas that are part of the image frame from the image frame, and a partial area that is extracted by the partial area extracting unit.
A configuration is provided that includes a frequency component conversion unit that converts the frequency component, and a feature amount output unit that outputs a part or all of the data converted by the frequency component conversion unit as a feature amount of the moving image content.

According to this structure, the content characteristic amount extraction device generates the characteristic amount based on the image frames of the moving image content input in chronological order. Therefore, the characteristic amount is the time of the moving image content. It is a feature quantity that captures the physical characteristics. Further, since the content feature amount extraction device extracts one or more partial regions, which are rectangular regions of a part of the image frame, from the image frame to generate the feature amount,
Depending on the number of partial areas, it is determined whether to increase the number of partial areas to generate a more accurate feature amount, or to reduce the number of partial areas to reduce the calculation amount and generate the feature amount at high speed. be able to.

Further, the content feature amount extraction program according to claim 8 is characterized in that a computer is provided for extracting the feature amount of the moving image content based on the image frames of the moving image content input in time series. Partial area extracting means for extracting at least one partial area which is a rectangular area of a part of the image frame from the image frame; frequency component converting means for converting the partial area extracted by the partial area extracting means into frequency components; A part or all of the data converted by the frequency component conversion means is made to function as a feature amount output means for outputting the feature amount of the moving image content.

According to this structure, the content feature amount extraction program generates the feature amount based on the image frames of the moving image content input in chronological order. Therefore, the feature amount is the time of the moving image content. It is a feature quantity that captures the physical characteristics. In addition, the content feature amount extraction program extracts one or more partial regions, which are rectangular regions of a part of the image frame, from the image frame to generate the feature amount. Therefore, the number of the partial regions depends on the number of the partial regions. To generate more accurate features, or
By reducing the number of partial regions, it is possible to reduce the amount of calculation and determine whether to generate the feature amount at high speed.

According to a ninth aspect of the present invention, in the content authentication data generation device, the moving image content is based on the moving image content and a distributor identifier that identifies an original distributor who is a distributor of the moving image content. Is a content authentication data generation device for generating authentication data for authenticating whether or not the video image content is distributed from an original distributor of the video image content, wherein the content of the video image content is based on the image frame of the video image content. The feature amount extracting means for extracting the feature amount, the distributor amount synthesizing means for synthesizing the feature amount of the moving image content extracted by the feature amount extracting means, and the distributor identifier, and the distributor identifier synthesizing means. A confirmation code generating means for generating content authentication data from the combined data is provided.

According to this structure, the content authentication data generation device uses the confirmation code from the combined data obtained by combining the feature amount of the moving image content and the distributor identifier for specifying the original distributor who distributes the moving image content. To create. For example, a hash value is generated from the combined data by a hash function. As a result, the content authentication data generation device generates authentication data for authenticating not only the moving image content itself but also the original distributor of the moving image content.

Furthermore, the content authentication data generation program according to the tenth aspect of the present invention is based on the moving image content and the distributor identifier for specifying the original distributor who is the distributor of the moving image content. A computer for generating authentication data for authenticating whether or not the video content was delivered by the original distributor of the video content, and extracting a feature amount of the video content based on the image frame of the video content. From the data synthesized by the distributor identifier synthesizing unit for synthesizing the feature amount extracting unit, the feature amount of the moving image content extracted by the feature amount extracting unit, and the distributor identifier, It is configured to function as a confirmation code generating means for generating content authentication data.

According to this structure, the content authentication data generation program uses the confirmation code from the synthetic data obtained by synthesizing the characteristic amount of the moving image content and the distributor identifier for specifying the original distributor who distributes the moving image content. To create. For example, a hash value is generated from the combined data by a hash function. As a result, the content authentication data generation program generates authentication data for authenticating not only the moving image content itself but also the original distributor of the moving image content.

Furthermore, in the content authentication method according to the eleventh aspect, the moving image content is based on the moving image content and a distributor identifier that identifies an original distributor who is a distributor of the moving image content. Authentication data generating means for generating authentication data for authenticating whether or not the moving image content is distributed by the original distributor, and original authentication generated by the original distributor in the same procedure as the authentication data generating means. What is claimed is: 1. A content authentication method for authenticating moving image content in a certificate authority having a storage unit for pre-registering data, the identification for identifying the moving image content and the original distributor of the moving image content. The information based on the identification information transmitted in the authentication information transmitting step, and the distribution based on the identification information transmitted in the authentication information transmitting step. A distributor identifier acquiring step of acquiring an identifier; an authentication data generating step of generating authentication data by the certificate authority based on the moving image content and the distributor identifier; and a step of generating the authentication data by the certificate authority. An authentication data comparison step of comparing the authentication data generated in step 1 with the original authentication data registered in advance, and an authentication result reply step of returning the comparison result of this comparison step as the content authentication result. And

According to this method, the content authentication method first transmits the moving image content and the identification information for identifying the original distributor of the moving image content to the certificate authority.
Then, the certificate authority acquires the distributor identifier based on the identification information, and generates authentication data based on the moving image content and the distributor identifier. Then, the authentication data generated here is compared with the original authentication data generated by the original distributor registered in advance in the certificate authority of the content, and the comparison result is returned as the authentication result. As a result, the moving image content user to whom the moving image content is distributed authenticates the moving image content and the original distributor of the moving image content by the certification authority.

[0032]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the drawings. (First Embodiment: Configuration of Content Feature Amount Extracting Device) FIG. 1 is a block diagram showing an overall configuration of a content feature amount extracting device according to a first embodiment of the present invention. As shown in FIG. 1, the content feature amount extraction device 1
Includes a reduced image generation unit 10 (reduced image generation unit), a two-dimensional DCT unit 20 (frequency component conversion unit), and a DCT coefficient selection unit 30 (feature amount output unit). The content feature amount extraction device 1 is a device that extracts a feature amount of a content based on image frames of moving image content (hereinafter referred to as content) input in time series and various parameters input from the outside. .

The reduced image generating section 10 includes a two-dimensional block dividing means 11, a two-dimensional block average value calculating means 13,
The image synthesizing means 15 is provided, and a reduced image is generated from the image frame of the content input in time series.

The two-dimensional block dividing means 11 has a block size (n
The image frame is divided in units of (x, ny). Note that nx is the number of pixels in the horizontal direction, and ny is the number of pixels in the vertical direction. Here, the data of each divided block (block data) is output to the two-dimensional block average value calculating means 13.

The two-dimensional block average value calculating means 13
The average value of the pixel values of each block data input from the dimensional block dividing means 11 is calculated as a representative value of the block. The average value data (average value data) calculated here is output to the image synthesizing unit 15.

The image synthesizing means 15 collects the average value data input from the two-dimensional block average value calculating means 13,
One image is generated. This image is a reduced image having a size of 1 / nx in the horizontal direction and 1 / ny in the vertical method with respect to the image frame input to the reduced image generation unit 10. Here, the generated reduced image is a two-dimensional DC.
It is output to the T section 20.

The two-dimensional DCT unit 20 is a reduced image generation unit 1
The two-dimensional DCT (Discrete Cosine Tra) is applied to the reduced image input from the image synthesizing unit 15 of 0.
nsform: discrete cosine transform) to convert the image data of the reduced image into DCT coefficients (frequency components) and output the DCT coefficients to the DCT coefficient selector 30.

The DCT coefficient selection unit 30 selects the number of frames (nf), D, which is a part of the parameters input from the outside, from the DCT coefficients input from the two-dimensional DCT unit 20.
CT coefficient number (m), DCT coefficient position (ACi (i =
The DCT coefficient to be output is selected based on 0, ..., M-1)). Here, the DCT coefficient of the reduced image converted by the two-dimensional DCT unit 20 is set to the number of frames (n
f) The DCT corresponding to the DCT coefficient position (ACi (i = 0, ..., m-1)) is collected from the DCT coefficients.
The coefficients are selected by the number of DCT coefficients (m). The selected DCT coefficient is output as the feature amount of the content.

With the above configuration, the content feature quantity extraction device 1 extracts the feature quantity of the content based on the image frames of the content input in time series and the predetermined parameters input from the outside. You can It should be noted that the content feature quantity extraction device 1 can realize each unit in a computer as each function program, or can combine each function program and operate as a content feature quantity extraction program.

In the present embodiment, the feature quantity of the content is generated using the DCT, but instead of the DCT, a DST (Discrete Line Transform) is used.
rm), KLT (Karhunen Loeve Tra)
nsform), DFT (Discrete Four)
It can also be realized by using an analysis method such as ier Transform).

(First Embodiment: Operation of Content Feature Quantity Extracting Device) Next, the operation of the content feature quantity extracting device 1 according to the first embodiment will be described with reference to FIGS. 1 and 8. To do. FIG. 8 is a flow chart for explaining the operation of the content feature quantity extraction device 1.

In the content feature quantity extraction device 1, the reduced image generation unit 10 inputs the content in image frame units (step S1), and the image frame is a block size (nx × ny) which is a part of the parameter. )
Block division is performed (step S2). Then, the average value of the pixel values of each block divided into blocks is calculated (step S3), and an image (reduced image) in which the average value is combined by the number of blocks is generated (step S4). next,
The content feature amount extraction device 1 causes the two-dimensional DCT unit 20 to perform processing on the reduced image generated in step S4.
Two-dimensional DCT is performed to calculate the DCT coefficient (step S5).

Then, the content feature quantity extraction device 1
The DCT coefficient selection unit 30 determines whether or not the DCT coefficients have been collected for the number of frames (nf) which is a part of the parameter (step S6). If the collection is not completed (No), go to step S1. Return and continue processing. On the other hand, when the collection of the DCT coefficients for the number of frames (nf) is completed (Yes), the number of DCT coefficients (m), which is a part of the parameters, and the DCT coefficient position are calculated from the DCT coefficients for the number of frames (nf). (ACi (i = 0, ..., m
-1)) is selected and output as a feature amount (step S7), and the operation ends.

By the above operation, the content feature quantity extraction device 1 extracts the feature quantity of the content based on the image frames of the content input in chronological order and various parameters input from the outside. You can In addition, since the feature quantity generated by the content feature quantity extraction device 1 is generated from data obtained by averaging the image frame in block units before converting the image frame into frequency components, the content is encoded by MPEG. A feature amount that does not change even when compressed / decompressed by, for example, can effectively function as a feature amount of the content.

(Second Embodiment: Structure of Content Feature Quantity Extracting Device) FIG. 2 is a block diagram showing the entire structure of a content feature quantity extracting device according to a second embodiment of the present invention. As shown in FIG. 2, the content feature amount extraction apparatus 1A changes the two-dimensional DCT unit 20 of the content feature amount extraction apparatus 1 shown in FIG. 1 into a three-dimensional DCT unit 21, and sets the DCT coefficient selection unit 30 as an input parameter. Since the DCT coefficient selecting unit 30A is changed to the configuration described above, the configurations other than the three-dimensional DCT unit 21 and the DCT coefficient selecting unit 30A are denoted by the same reference numerals as those shown in FIG. Omit it.

The three-dimensional DCT unit 21 is a reduced image generation unit 1
The reduced image generated by the image synthesizing unit 15 of 0 is the number of frames (n
f) is input, two-dimensional DCT is performed on each reduced image, DCT coefficients are generated, and further, DCT coefficients are collected for the number of frames (nf) for each coefficient. Is performed to generate a three-dimensional DCT coefficient and output it to the DCT coefficient selection unit 30A.
In addition, for the DCT coefficient generated here, D
It is also possible to generate a four-dimensional DCT coefficient by using CT and use this DCT coefficient. Furthermore, by sequentially performing DCT on this DCT coefficient,
It is also possible to generate a multidimensional DCT coefficient and use the DCT coefficient.

The DCT coefficient selection section 30A is a three-dimensional DCT.
From the DCT coefficients input from the unit 21, the number of DCT coefficients (m), which is a part of the parameters input from the outside, the DCT coefficient position (ACi (i = 0, ..., M-
The DCT coefficient to be output is selected based on 1)). Here, from the DCT coefficient of the reduced image converted by the three-dimensional DCT unit 21, the DCT coefficient position (ACi (i
= 0, ..., M-1)) corresponding to the number of DCT coefficients (m) are selected and output as the feature amount of the content.

Thus, in this embodiment (FIG. 2),
Since the DCT coefficient is calculated for the reduced image of only the number of frames designated based on the parameters, when the number of frames is determined in advance to operate, the feature amount can be extracted efficiently and at high speed. it can.

The content feature quantity extraction device 1A is
It is also possible to realize each means in the computer as each function program, and to combine each function program to operate as a content feature amount extraction program. Also, instead of DCT, DST, KL
It can also be realized by using an analysis method such as T or DFT.

(Second Embodiment: Operation of Content Feature Quantity Extracting Device) Next, with reference to FIGS. 2 and 9, the operation of the content feature quantity extracting device 1A according to the second embodiment will be described. To do. FIG. 9 shows the content feature amount extraction device 1
6 is a flowchart for explaining the operation of A.

In the content feature amount extraction device 1A, the reduced image generation unit 10 inputs the content in image frame units (step S10), and the image frame is a block size (nx × n) which is a part of the parameter.
y) is divided into blocks (step S11). Then, the average value of the pixel values of each block divided into blocks is calculated (step S12), and an image (reduced image) is generated by combining the average value by the number of blocks (step S1).
3).

Next, the content feature quantity extraction device 1A
The three-dimensional DCT unit 21 first performs two-dimensional DCT on the reduced image generated in step S13.
The CT coefficient is calculated (step S14). Here, it is determined whether or not the DCT coefficients for the number of frames (nf), which is a part of the parameters, are collected (step S15). If the collection is not completed (No), the process returns to step S10 to continue the processing. . On the other hand, D for the number of frames (nf)
If the CT coefficient collection is completed (Yes), the process proceeds to step 16. Then, for each DCT coefficient calculated in step S14 and collected for the number of frames (nf), DCT is performed for each coefficient (step S16).

The result of the two-dimensional DCT in step S14 is further referred to as three-dimensional DCT in which DCT is performed in step S16, and is executed by the three-dimensional DCT unit 21. Then, the content feature amount extraction device 1A
The CT coefficient selecting unit 30A selects, from each DCT coefficient, one corresponding to the number of DCT coefficients (m), which is a part of the parameters, and the DCT coefficient position (ACi (i = 0, ..., M-1)). , Is output as a feature amount (step S1)
7).

By the above operation, the content feature amount extraction apparatus 1A can extract the feature amount of the content based on the image frames of the content input in chronological order and various parameters input from the outside. it can. In addition, since the feature amount generated by the content feature amount extraction device 1A is generated from the data obtained by averaging the image frame in block units before converting the image frame into frequency components, the content is encoded by MPEG. A feature amount that does not change even when compressed / decompressed by, for example, can effectively function as a feature amount of the content.

(Third Embodiment: Structure of Content Feature Quantity Extracting Device) FIG. 3 is a block diagram showing the overall structure of a content feature quantity extracting device according to a third embodiment of the present invention. As shown in FIG. 3, the content feature amount extraction device 1B includes a reduced image generation unit 10A (reduced image generation unit), a three-dimensional DCT unit 21A (frequency component conversion unit), and a DCT coefficient selection unit 30A (feature amount output). And)). This content feature amount extraction device 1B
Is a device for extracting a feature amount of a moving image content (hereinafter referred to as a content) input in time series, and various parameters input from the outside based on various parameters input from the outside.

The reduced image generator 10A comprises a three-dimensional block dividing means 12 and a three-dimensional block average value calculating means 14.
And the image synthesizing means 15 to generate a three-dimensional reduced image from image frames of contents input in time series.

The three-dimensional block dividing means 12 has a block size (n
Based on (x, ny) and the number of frames (nf), the input content is divided into (nx × ny × nf) units of three-dimensional blocks (divided region groups). Note that n
x is the number of pixels in the horizontal direction, ny is the number of pixels in the vertical direction,
nf is the number of image frames of the content. Here, the data of each divided block (block data) is output to the three-dimensional block average value calculating means 14.

The three-dimensional blocks are the first to nth blocks.
It is generated from frames divided in units of the number of frames (nf), such as the f-th frame, the (nf + 1) th to (2xnf) th frame, and the (2xnf + 1) th to (3xnf) th frame. Is the number of frames such as the 1st to nfth frames, the 2nd to (nf + 1) th frames, and the 3rd to (nf + 2) th frames (n
f) The form may be generated from a frame having overlapping units.

The three-dimensional block average value calculating means 14 uses 3
The average value of pixel values of each block data input from the dimensional block dividing means 12 is calculated as a representative value of the block. The average value data (average value data) calculated here is output to the image synthesizing unit 15.

The image synthesizing means 15 collects the average value data input from the three-dimensional block average value calculating means 14,
A three-dimensional reduced image is generated. The image synthesizing unit 15 continuously generates the reduced images having the size of 1 / nx in the horizontal direction and 1 / ny in the vertical method for the image frames input to the reduced image generation unit 10A. The three-dimensional DCT unit 21A sets the DCT block size (nk, nl, nm) that is a part of the input parameter for the three-dimensional reduced image generated by the image combining unit 15 of the reduced image generating unit 10A. By performing three-dimensional DCT based on this, it is converted into a DCT coefficient (frequency component) and output to the DCT coefficient selecting unit 30A. Note that nk is
The number of pixels in the horizontal direction, nl is the number of pixels in the vertical direction, nm
Is the number of pixels in the depth.

The DCT coefficient selecting section 30A is a three-dimensional DCT.
From the DCT coefficients input from the unit 21A, the number of DCT coefficients (m) that is a part of the parameters input from the outside and the DCT coefficient position (ACi (i = 0, ..., m-
The DCT coefficient to be output is selected based on 1)). Here, from the DCT coefficient of the reduced image converted by the three-dimensional DCT unit 21A, the DCT coefficient position (ACi
The DCT coefficient corresponding to (i = 0, ..., m-1)) is set to DC
The number of T coefficients (m) is selected and output as the feature amount of the content.

With the above configuration, the content feature quantity extraction device 1B extracts the feature quantity of the content based on the image frames of the content input in time series and the predetermined parameters input from the outside. You can It should be noted that the content feature amount extraction apparatus 1B can realize each unit in a computer as each function program, or can combine each function program and operate as a content feature amount extraction program. Also, instead of DCT, DST, KLT, DF
It can also be realized by using an analysis method such as T.

(Third Embodiment: Operation of Content Feature Quantity Extracting Device) Next, with reference to FIGS. 3 and 10, the operation of the content feature quantity extracting device 1B according to the third embodiment will be described. To do. FIG. 10 is a flowchart for explaining the operation of the content feature amount extraction device 1B.

In the content feature amount extraction device 1B, whether the reduced image generation unit 10A inputs the content in image frame units (step S20), and whether the image frames for the number of frames (nf), which is a part of the parameter, have been input. It is determined (step S21), the number of frames (n
If f) has not been input (No), the process returns to step S20. On the other hand, when the number of frames (nf) is input, the input image frame is a three-dimensional image designated by the block size (nx, ny) and the number of frames (nf) that are part of the parameters. Divide into block size (nx × ny × nf) units (step S
22). Then, the average value of the pixel values of each block divided into block size units is calculated (step S23),
An image (reduced image) is generated by synthesizing the average value for the number of blocks (step S24).

Next, the content feature quantity extraction device 1B
The three-dimensional DCT unit 21A first sets the size (n) specified for the reduced image generated in step S24.
Two-dimensional DCT of (k × nl) is performed to calculate the DCT coefficient (step S25). Then, the calculated DCT
Collect the coefficients for the specified nm reduced images, and
DCT is further performed for each T coefficient (step S26).
It is to be noted that, in step S26, the DCT coefficient existing at the same DCT coefficient position of each reduced image is further subjected to DCT for each reduced image, which is called three-dimensional DCT.
It is executed by the CT unit 21A.

Then, the content feature quantity extraction device 1B
Is calculated by the DCT coefficient selection unit 30A from each DCT coefficient, the number of DCT coefficients (m) that is a part of the parameter, D
The one corresponding to the CT coefficient position (ACi (i = 0, ..., M-1)) is selected and output as a feature amount (step S27).

By the above operation, the content feature amount extraction apparatus 1B extracts the feature amount of the content based on the image frames of the content input in chronological order and various parameters input from the outside. You can In addition, since the feature amount generated by the content feature amount extraction device 1B is generated from the data obtained by averaging the image frame in block units before converting the image frame into frequency components, the content is encoded by MPEG.
It becomes a feature amount that does not change even when compressed / decompressed with
It can effectively function as a feature amount of the content.

(Fourth Embodiment: Configuration of Content Feature Amount Extracting Device) FIG. 4 is a block diagram showing the overall configuration of a content feature amount extracting device according to a fourth embodiment of the present invention. As shown in FIG. 4, the content feature amount extraction device 1C includes a partial area extraction unit 40 (partial area extraction unit), a DCT unit 22 (frequency component conversion unit), and a DC.
It is configured to include a T coefficient selection unit 30A (feature amount output means). The content feature amount extraction device 1C is a moving image content (hereinafter referred to as content) that is input in time series.
This is a device for extracting the feature amount of the content based on the image frame and various parameters input from the outside.

The partial area extraction unit 40 is provided with a block dividing means 41 and a block selecting means 42, and divides the content into two-dimensional or three-dimensional block areas from the image frames of the content input in time series. The data of the designated block area is output. The block dividing unit 41 determines the input content as a two-dimensional or three-dimensional block size (nx) based on the block size (nx, ny) and the number of frames (nf) that are part of the parameters input from the outside. Xnyxnf)
It is divided into units. Note that nx is the number of pixels in the horizontal direction, ny is the number of pixels in the vertical direction, and nf is the number of image frames of the content. The divided data is two-dimensional data when nf is 1, and three-dimensional data when nf is 2 or more. here,
The data of each divided block (block data) is
It is output to the block selection means 42.

The three-dimensional blocks are the first to
nfth frame, (nf + 1) th to (2 × nf)
Th frame, (2 × nf + 1) th to (3 × nf)
A frame such as the 1st to nfth frame, the 2nd to (nf + 1) th frame, and the 3rd to (nf + 2) th frame. The form may be generated from frames in which the number (nf) unit is overlapped.

The block selecting means 42 selects, from the two-dimensional or three-dimensional block data inputted from the block dividing means 41, the number of blocks (nb) which is a part of the parameter inputted from the outside and the block position (bpi). (I
= 0, ..., Nb-1)), the block data to be output is selected. Here, among the block data input from the block dividing means 41, the block data corresponding to the block position (bpi (i = 0, ..., Nb-1)) is selected and output.

The DCT section 22 performs DCT on the block data selected by the block selecting means 42 to generate a DCT coefficient (frequency component) and outputs it to the DCT coefficient selecting section 30A. The DCT unit 2
2 indicates that when the input block data is two-dimensional data, two-dimensional DCT is performed, and when the input block data is three-dimensional data, it is three.
Perform dimension DCT. The DCT coefficient generated here is D
It is output to the CT coefficient selection unit 30A.

The DCT coefficient selecting section 30A includes a DCT section 22.
DCT coefficients (m), which is a part of parameters input from the outside, among DCT coefficients input from
The DCT coefficient to be output is selected based on the T coefficient position (ACi (i = 0, ..., M-1)). Here, from the DCT coefficient generated by the DCT unit 22, the DCT coefficient
The DCT coefficient corresponding to the coefficient position (ACi (i = 0, ..., M-1)) is selected and output as the feature amount of the content.

With the above configuration, the content characteristic amount extraction apparatus 1C extracts the characteristic amount of the content based on the image frames of the content input in time series and the predetermined parameters input from the outside. You can It should be noted that the content feature amount extraction apparatus 1C can also realize each unit in a computer as each function program, or can combine each function program and operate as a content feature amount extraction program. Also, instead of DCT, DST, KLT, DF
It can also be realized by using an analysis method such as T.

(Fourth Embodiment: Operation of Content Feature Quantity Extracting Device) Next, with reference to FIGS. 4 and 11, an operation of the content feature quantity extracting device 1C according to the fourth embodiment will be described. To do. FIG. 11 is a flowchart for explaining the operation of the content feature quantity extraction device 1C.

In the content feature amount extraction device 1C, the partial area extraction unit 40 inputs the content in image frame units (step S30), and determines whether the number of image frames (nf) which is a part of the parameter has been input. Judgment (step S31), the number of frames (nf)
If the minutes have not been input (No), the process returns to step S30. On the other hand, when the number of frames (nf) is input, the input image frame is a three-dimensional image designated by the block size (nx, ny) and the number of frames (nf) that are part of the parameters. Divide into block size (nx × ny × nf) units (step S3)
2).

Then, the number of blocks (nb), which is a part of the parameter, and the block position (bpi (i = 0, ..., N
b-1)) corresponding block data is selected (step S33), and three-dimensional DC is applied to each block data.
T is performed to calculate the DCT coefficient (step S3
4).

Then, the DCT coefficient selecting unit 30A calculates the number of DCT coefficients (m), which is a part of the parameters, and the DCT coefficient position (ACi (i = 0, ..., M-1) from the DCT coefficients calculated in step S34. )) Is selected and output as a feature amount (step S3).
5).

By the above operation, the content feature amount extraction apparatus 1C can extract the feature amount of the content based on the image frames of the content input in time series and various parameters input from the outside. it can. Further, the feature quantity generated by the content feature quantity extraction device 1C becomes a feature quantity that does not change even when the content is compressed / expanded by MPEG or the like by selecting the DCT coefficient from the low frequency component, and the feature of the content It can function effectively as a quantity.

(Fifth Embodiment: Configuration of Content Authentication Data Generating Device) FIG. 5 is a block diagram showing the overall configuration of a content authentication data generating device according to a fifth embodiment of the present invention. As shown in FIG. 5, the content authentication data generation device 2 has a feature amount extraction means 50, a distributor identifier synthesis means 51, and a confirmation code generation means 52.
And the authentication data synthesizing means 53. The content authentication data generation device 2 is a device that generates authentication data for authenticating a content and a content distributor.

The feature quantity extraction means 50 extracts the feature quantity of the content based on the image frames input in time series of the content. This feature amount extraction means 50
Can be realized using, for example, the content feature amount extraction device described in FIGS. 1 to 4. In this case, the feature amount extraction unit 50 is configured to use the block size (nx, ny),
Number of frames (nf), DCT block size (nk, n
l, nm), DCT coefficient number (m), DCT coefficient position (ACi (i = 0, ..., m-1)), block number (n
b), block position (bpi (i = 0, ..., Nb-
The feature amount of the content is extracted by inputting parameters such as 1)). The feature amount of this content is output to the distributor identifier synthesizing unit 51.

The distributor identifier synthesizing means 51 uses the combination number inputted from the outside for the feature quantity of the content inputted from the feature quantity extracting means 50 and the content distributor identifier for specifying the original distributor of the content. To generate combined data. As the content distributor identifier, a numerical value or the like that can recognize the original distributor of the content can be used. The combined data generated here is output to the confirmation code generating means 52.

Here, a synthesizing method for synthesizing the feature amount of the content and the content distributor identifier will be described with reference to FIG. FIG. 6 shows the content feature amount (DCT
It is a figure showing an example of composition of i (i = 0, ..., M-1) and a content distributor identifier (CHID). Figure 6
In the combination method 1 (composition number 1) of (1), the content distributor identifier (CHID) is set to the feature amount of the content (DC
It is added immediately before Ti (i = 0, ..., M-1)). Figure 6
In the synthesizing method 2 (combining number 2) of (2), the content distributor identifier (CHID) is set to the content feature amount (DC
It is added immediately after Ti (i = 0, ..., M-1)).

In the composition method 3 (composition number 3) of FIG. 6C, the content distributor identifier (CHID) is divided,
Content feature amount (DCTi (i = 0, ..., m−
1)) Add inside. In the combination method 4 (composition number 4) of FIG. 6 (4), the content distributor identifier (CHID) and the feature amount of the content (DCTi (i = 0, ..., M-1))
And are randomly combined.

Note that the side that actually authenticates the content (for example, the content certification authority) holds this synthesizing method as known information due to a contract with the original distributor of the content, and based on the synthesizing number. , The content distributor identifier (CHID) and the content feature amount (DCTi (i = 0, ..., M-1)) can be extracted. Returning to FIG. 5, the description will be continued.

The confirmation code generating means 52 uses the SHA data input from the distributor identifier synthesizing means 51 to generate SHA.
A hash value is generated by a hash function such as (Secure Hash Algorithm). This hash value is output to the authentication data synthesizing means 53. The authentication data synthesizing means 53 includes the hash value generated by the confirmation code generating means 52 and the feature quantity extracting means 5.
The parameter input to 0 and the distributor identifier synthesizing unit 51
The combination number and the combination number input to are combined and output as authentication data.

On the side that actually authenticates the content (for example, the content authentication station), the authentication data can be generated from the content by using the parameter and the synthesis number synthesized here. Thereby, for example, the content certification authority compares the original certification data of the content registered in advance with the certification data of the content newly generated by the parameter and the composition number to certify the content. Can be done.

The configuration of the content authentication data generating device 2 according to the present invention has been described above based on the embodiment, but the present invention is not limited to this. For example, when the parameter and the composition number are registered in advance as known information by the contract between the original distributor of the content and the content certification authority, the feature amount extraction means 50 inputs the parameter. Instead, the operation may be performed based on an internal parameter value in which the same value as the parameter is set in advance, and the distributor identifier synthesizing unit 51 may also be configured to operate based on a preset synthesis number. Absent. In this case, the authentication data synthesizing means 53 may be removed from the configuration of the content authentication data generating device 2 and the hash value generated by the confirmation code generating means 52 may be used as the authentication data.

Further, when the authentication data for authenticating only the substance of the content is generated without authenticating the original distributor of the content, the distributor identifier synthesizing means 51 is removed from the configuration, and the characteristic amount extracting means 50 generates it. The feature amount may be input to the confirmation code generation unit 52 to generate the authentication data. It should be noted that the content authentication data generation device 2 can realize each unit in a computer as each function program, or can combine each function program and operate as a content authentication data generation program.

(Fifth Embodiment: Operation of Content Authentication Data Generation Device) Next, with reference to FIG. 5 and FIG.
The operation of the content authentication data generation device 2 according to the fifth embodiment will be described. FIG. 12 is a flowchart for explaining the operation of the content authentication data generation device 2.

The content authentication data generating device 2 extracts the characteristic amount of the content by the characteristic amount extracting means 50 based on the parameter for generating the authentication data (step S40). Then, the distributor identifier synthesizing means 5
According to 1, the feature amount and the distributor identifier that identifies the original distributor that is the distributor of the content are combined based on the combination number that specifies the combining method (step S41).
The confirmation code generating means 52 calculates a hash value from the combined data combined in step S41 using a hash function (step S42). Then, the authentication data synthesizing unit 53 generates authentication data by synthesizing the hash value, the parameter, and the synthesis number (step S43).

By the above operation, the content authentication data generating apparatus 2 generates the characteristic amount of the content and combines the characteristic amount with the distributor identifier for specifying the original distributor as the distributor of the content. It is possible to generate authentication data necessary for performing content authentication and content distributor authentication.

(Structure of Content Authentication System) FIG. 7
FIG. 1 is a diagram showing a schematic configuration of a content authentication system. As shown in FIG. 7, the content authentication system includes a certificate authority 60 that authenticates the content and the original distributor of the content and a distributor management that manages the content distributor identifier of the original distributor of the content via the network 90. It is composed of a device 70 and a user terminal device 80 for requesting content authentication.

The certification authority 60 comprises a certification control means 61, a certification data generation means 62, and a storage means 63. The certificate authority 60 is a general computer, and includes a CPU (not shown), a memory, communication control means for accessing the network 90, and the like.

The certificate authority 60 authenticates the content and the original distributor of the content in response to the content authentication request from the user terminal device 80. The authentication control means 61 controls requests and responses with the distributor management device 70 and the user terminal device 80. The authentication control means 61 acquires a content distributor identifier for authenticating the original distributor of the content from the distributor management device 70. Further, the authentication control unit 61 notifies the user terminal device 80 of the authentication result of the authentication data generation unit 62, which will be described later, in response to the content authentication request from the user terminal device 80.

Further, the authentication data generating means 62 generates authentication data based on the content received from the user terminal device 80 and the content distributor identifier. The authentication data generating means 62 can use the content authentication data generating device of FIG.

Further, the accumulating means 63 accumulates the authentication data (original authentication data 63a) of the content to be distributed in accordance with a contract with the original distributor of the content in advance.
It is composed of a general storage medium. The original authentication data 63a is generated by the same procedure as the authentication data generating means 62. Further, the original authentication data 63a is compared with the authentication data generated by the authentication data generation means 62, and the comparison result is notified to the user terminal device 80 by the authentication control means 61. The storage means 6
It is assumed that 3 holds an address or the like for identifying the distributor management device corresponding to the content (not shown).

The distributor management device 70 comprises a distributor identifier response means 71 and a storage means 72. The distributor management device 70 is a general computer, and includes a CPU (not shown), a memory, communication control means for accessing the network 90, and the like.

This distributor management device 70 notifies the certificate authority 60 of the content distributor identifier 72a incorporated in the authentication data when the original distributor of the contents generates the authentication data of the contents. is there. The distributor identifier responding unit 71 notifies the certificate distributor 60 of the content distributor identifier 72a based on the content distributor identifier request from the certificate authority 60.

The accumulating means 72 accumulates the above-mentioned content distributor identifier 72a, and is composed of a general recording medium. The content distributor identifier 72a is incorporated by the original distributor of the content when generating the authentication data of the content, and may be a value unique to the original distributor of the content or a numerical value managed for each content. May be used.

The user terminal device 80 has the authentication requesting means 81.
And storage means 82. This certificate authority 60 is a general computer, and is not shown in the figure.
It has a PU, a memory, a communication control means for accessing the network 90, and the like. A personal computer or a workstation can be used as the user terminal device 80.

The authentication requesting means 81 is a content authenticator for the content 82a distributed by the content distributor.
A request for content distributor authentication is made to the certificate authority 60, and the authentication result is displayed on a display device or the like (not shown). Further, the storage means 81 stores the content distributed by the content distributor, and is composed of a general storage medium. This content 82a
May be delivered directly from the original distributor of the content, or may be delivered from a third party. In this content authentication system, the user terminal device 80
There may be more than one.

(Content Authentication Method) Next, a content authentication method in the content authentication system will be described with reference to FIGS. FIG. 13 is a flowchart showing an operation procedure for authenticating content. The operation will be described below based on each step.

(Authentication Information Transmission Step) First, the user terminal device 80 transmits a “contents authentication request” for requesting content authentication to the authentication station 60 (step S51). Then, the certificate authority 60 having received the “content authentication request” transmits a “content & content distributor name request” requesting the content and the distributor name of the content to the user terminal device 80 (step S52). ). Based on this request, the user terminal device 80
Sends the substance of the content and the name of the distributor of the content to the certificate authority 60 by "content & content distributor name response".

(Distributor Identifier Obtaining Step) Next, the certificate authority 60 requests the content distributor identifier from the corresponding distributor management device 70 based on the content distributor name "content distributor identifier". The request is transmitted (step S54). Then, the distributor management device 70 that has received the “content distributor identifier request” is stored in the storage means 72 and is distributed to the contents distributor identifier 72.
a is transmitted to the certificate authority 60 as a "content distributor identifier response" (step S55).

(Authentication Data Generation Step) The certification authority 60, which has received the content distributor identifier, generates the certification data from the contents received in step S53 by the certification data generation means 62, and outputs the certification data and the original certification data 63a. To compare. Note that, here, the description will be made on the assumption that the authentication data of the content uses the authentication data generated by the content authentication data generation device of FIG.

First, the certificate authority 60 extracts a parameter, a composite number, and a hash value from the original authentication data 63a stored in the storage means 63 corresponding to the content received in step S53 (step S56). . Then, an authentication hash value is generated for the received content based on the parameters extracted from the original authentication data 63a, the composition number, and the content distributor identifier 72a received in step S55 (step S57).

(Authentication Data Comparison Step) Next, the authentication station 60 compares the authentication hash value with the hash value extracted from the original authentication data 63a in step S56 (step S58). If the comparison result is different, it is authenticated that the content requested by the user has been tampered with.

(Authentication result reply step) Then, the certification authority 60 sends a "contents certification result notification" to the user terminal device 80 which has requested certification of the contents based on the comparison result. Through the above steps, the content user can perform content authentication and content distributor authentication by making an inquiry to the certificate authority, so that unauthorized content can be detected without burdening the user. You can

[0110]

As described above, the content feature amount extraction device, the program thereof, the content authentication data generation device, the program thereof, and the content authentication method according to the present invention have the following excellent effects.

According to the first and sixth aspects of the present invention, the content feature amount extraction device and the program accurately capture the temporal and planar changes of the image frame of the moving image content. It is possible to generate a feature amount that can specify the content. Further, even when the moving image content is compressed by MPEG or the like and transmitted, the characteristic amount can function as the characteristic amount of the moving image content, and thus can be used as authentication data of the moving image content.

According to the second aspect of the present invention, the content feature amount extraction apparatus uses the reduced image generated by using the average value of the pixel values of each divided area obtained by dividing the image frame of the moving image content as the representative pixel value. Since the feature amount is generated, the feature amount does not change even if the moving image content is compressed / expanded, and can be used as authentication data for authenticating the content.

According to the third aspect of the present invention, since the content feature quantity extraction device generates one reduced image from a plurality of image frames, it is possible to reduce the amount of calculation for generating the frequency component. The extraction time can be shortened. In addition, the content feature amount extraction device compresses moving image content in order to generate a feature amount from a reduced image generated by using an average value of pixel values of a divided area group including divided areas of a plurality of image frames as a representative pixel value. -The feature amount does not change even if it is expanded, and it can be used as authentication data for authenticating content.

According to the invention described in claim 4, the content feature amount extraction apparatus can generate a feature amount that accurately captures the planar feature of the moving image content, and the feature amount is the content. It can be used as authentication data for authentication.

According to the invention described in claim 5, the content feature amount extraction apparatus can generate a feature amount which accurately captures the planar and temporal features of the moving image content, and the feature amount is , Can be used as authentication data for authenticating content.

According to the seventh and eighth aspects of the present invention, the content feature amount extraction device and program specify the moving image content by accurately grasping the temporal change of the image frame of the moving image content. It is possible to generate a possible feature amount. In addition, the content feature amount extraction device and the program select the DCT coefficient from the low-frequency component to compress the moving image content by MPEG or the like and transmit the moving image content. Since it can effectively function as a quantity, it can be used as authentication data for moving image content.

According to the ninth or tenth aspect of the present invention, the content authentication data generation device and program accurately grasp the temporal change of the image frame of the moving image content, and the moving image content is MPEG. Since the authentication data is generated by the feature amount that does not change even when the content is compressed, it is possible to effectively authenticate the alteration of the content. Further, since the content distributor identifier set by the content distributor is incorporated in the authentication data, the content distributor authentication can be performed by confirming the content distributor identifier.

According to the invention described in claim 11, in the content authentication method, the authentication data generated by the feature amount of the content and the content distributor identifier set by the original distributor of the content is registered in the certificate authority. By setting
When authenticating the content, the user of the content only needs to make an inquiry to the certification authority, and the user does not need a dedicated device, a program, or the like, and can be surely authenticated. In addition, the certificate authority,
Since the feature amount that does not change even if the content is compressed by MPEG or the like is registered as the authentication data, it is possible to perform authentication even if the content is compressed after being transmitted by the MPEG or the like and transmitted.

[Brief description of drawings]

FIG. 1 is a block diagram showing an overall configuration of a content feature quantity extraction device according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing an overall configuration of a content feature quantity extraction device according to a second embodiment of the present invention.

FIG. 3 is a block diagram showing an overall configuration of a content feature quantity extraction device according to a third embodiment of the present invention.

FIG. 4 is a block diagram showing an overall configuration of a content feature quantity extraction device according to a fourth embodiment of the present invention.

FIG. 5 is a block diagram showing an overall configuration of a content authentication data generation device according to a fifth embodiment of the present invention.

FIG. 6 is an explanatory diagram illustrating an example of a method of combining a feature amount of content and a content distributor identifier.

FIG. 7 is a block diagram showing a schematic configuration of a content authentication system.

FIG. 8 is a flowchart illustrating the operation of the content feature amount extraction device according to the first embodiment of the present invention.

FIG. 9 is a flowchart illustrating an operation of the content feature quantity extraction device according to the second embodiment of the present invention.

FIG. 10 is a flowchart illustrating an operation of the content feature quantity extraction device according to the third embodiment of the present invention.

FIG. 11 is a flowchart illustrating an operation of the content feature quantity extraction device according to the fourth embodiment of the present invention.

FIG. 12 is a flowchart illustrating an operation of the content authentication data generation device according to the fifth embodiment of the present invention.

FIG. 13 is a flowchart illustrating the operation of the content authentication method according to the present invention.

[Explanation of symbols]

1, 1A, 1B, 1C ... Content feature quantity extraction device 2 ... Content authentication data generator 10, 10A ... Reduced image generator 11 ... Two-dimensional block dividing means 12: three-dimensional block dividing means 13 ... Two-dimensional block average value calculation means 14 ... Three-dimensional block average value calculation means 15 ... Image synthesizing means 20: Two-dimensional DCT part 21,21A: 3D DCT section 22 ... DCT section 30, 30A ... DCT coefficient selection section 40 ... Partial area extraction unit 41: Block dividing means 42 ... Block selection means 50 ... Feature amount extraction means 51 ... Distributor identifier synthesizing means 52 ... Code generation means for confirmation 53 ... Authentication data synthesizing means 60 ... Certificate Authority 61 ... Authentication control means 62 ... Authentication data generating means 63: storage means 63a: Original authentication data 70: Distributor management device 71. Distributor identifier response means 72 ... Storage means 72a ... Content distributor identifier 80: User terminal device 81 ... Authentication request means 82: storage means 82a …… Contents 90: Network

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Ituro Murota             1-10-11 Kinuta, Setagaya-ku, Tokyo, Japan             Broadcasting Association Broadcast Technology Institute (72) Inventor Tsuyoshi Otake             1-10-11 Kinuta, Setagaya-ku, Tokyo, Japan             Broadcasting Association Broadcast Technology Institute F term (reference) 5C059 KK38 KK43 MA00 MA23 PP04                       RB01 RC35 SS06 UA02 UA05                 5C064 BA01 BB02 BC17 BC22 BD08                 5J064 AA02 BA16 BC01 BC16 BC21                       BC25 BC29 BD04 CA03                 5L096 FA23 GA19

Claims (11)

[Claims] 1. A content feature amount extraction device for extracting a feature amount of a moving image content based on image frames of the moving image content input in chronological order, wherein: Reduced image generation means for generating a reduced image, frequency component conversion means for converting the reduced image generated by the reduced image generation means into frequency components, and part or all of the data converted by the frequency component conversion means Is output as a feature amount of the moving image content, and a content feature amount extraction device comprising: 2. The reduced image generating means generates a reduced image by extracting the average value of the pixel values of the divided areas obtained by dividing the image frame for each plurality of pixel data as a representative pixel value. The content feature quantity extraction device according to claim 1, which is characterized. 3. The reduced image generation means generates, from the plurality of input image frames, divided regions obtained by dividing each image frame for each plurality of pixel data, and a plurality of divided regions existing at the same position of each image frame are generated. The content feature amount extraction device according to claim 1, wherein a reduced image is generated by extracting an average value of pixel values of a divided area group configured by the divided areas as a representative pixel value. 4. The frequency component conversion means converts the reduced image generated by the reduced image generation means into frequency components by two-dimensional orthogonal transformation. 2. The content feature amount extraction device according to item 1. 5. The frequency component conversion means converts the plurality of reduced images generated by the reduced image generation means into frequency components by three-dimensional orthogonal transformation. The content feature amount extraction device according to any one of 1. 6. A computer for extracting a feature amount of moving image content based on image frames of the moving image content input in chronological order, and generating a reduced image of the image frame from the image frame. A reduced image generating means, a frequency component converting means for converting the reduced image generated by the reduced image generating means into frequency components, and a part or all of the data converted by the frequency component converting means, A content feature amount extraction program that causes the feature amount output unit to output as a feature amount of. 7. A content feature quantity extraction device for extracting a feature quantity of a moving image content based on an image frame of the moving image content input in chronological order, wherein: Partial area extracting means for extracting at least one partial area which is a rectangular area, frequency component converting means for converting the partial area extracted by the partial area extracting means into frequency components, and this frequency component converting means A content feature quantity extraction device comprising: a feature quantity output means for outputting a part or all of the converted data as a feature quantity of the moving image content. 8. A computer for extracting a feature amount of moving image content based on image frames of the moving image content input in chronological order, and a rectangular area which is a part of the image frame from the image frame. A partial area extracting means for extracting one or more partial areas, a frequency component converting means for converting the partial areas extracted by the partial area extracting means into frequency components, and one of the data converted by the frequency component converting means. A content feature amount extraction program, which causes all or all of them to function as feature amount output means for outputting the feature amount of the moving image content. 9. The moving image content is distributed from the original distributor of the moving image content based on the moving image content and a distributor identifier that identifies the original distributor who is the distributor of the moving image content. A content authentication data generation device for generating authentication data for authenticating whether or not it is a characteristic, and a characteristic amount extraction means for extracting a characteristic amount of the moving image content based on the image frame of the moving image content; A distributor identifier combining unit that combines the feature amount of the moving image content extracted by the amount extracting unit and the distributor identifier, and a confirmation that content authentication data is generated from the data combined by the distributor identifier combining unit. A content authentication data generation device, comprising: 10. The moving image content is distributed from the original distributor of the moving image content based on the moving image content and a distributor identifier that identifies the original distributor who is the distributor of the moving image content. A computer for generating authentication data for authenticating whether or not it is a characteristic amount extraction means for extracting a characteristic amount of the moving image content based on the image frame of the moving image content; A distributor identifier synthesizing unit for synthesizing the feature amount of the moving image content and the distributor identifier, and a confirmation code generating unit for generating content authentication data from the data synthesized by the distributor identifier synthesizing unit. A content authentication data generation program characterized by causing it to function. 11. The moving image content is distributed from the original distributor of the moving image content based on the moving image content and a distributor identifier that identifies the original distributor who is the distributor of the moving image content. Authentication data generation means for generating authentication data for authenticating whether or not
A content authentication method for authenticating moving image content at an authentication station, comprising: a storage unit for pre-registering original authentication data generated by the original distributor in the same procedure as the authentication data generating unit. An authentication information transmitting step of transmitting the moving image content and identification information for identifying an original distributor of the moving image content to the certificate authority, based on the identification information transmitted by the certificate authority in the authentication information transmitting step. A distributor identifier acquisition step of acquiring the distributor identifier; an authentication data generation step of the authentication station generating authentication data based on the moving image content and the distributor identifier; An authentication data comparison step of comparing the authentication data generated in this authentication data generation step with the original authentication data registered in advance, Content Authentication method for authenticating result and replies step comprising the returning a comparison result of the step as an authentication result of the content.