JP2000287073A - Method for imbedding and detecting electronic watermark, device for imbedding and detecting electronic watermark, and storage medium with electronic watermark imbedding and detection program stored therein, and electronic watermark system and integrated circuit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect an electronic watermark from a partial image by adding a watermark pattern resulting from applying discrete inverse Fourier transform to a watermark series matrix, after revision to an original image in a form of tiles to generate an imbedded image and to quicken image processing. SOLUTION: An imbed object component designation section 130 uses key information 104 to generate imbed object component position information 131 and gives the information 131 to a component value revision section 140. An imbed series generating section 120 spreads position marker information and an electronic watermark 103 to an imbed series 121 and gives the result to the component value revision section 140. The component value revision section 140 uses the imbed object component position information 131 and the imbed series 121, to set a null matrix to a watermark coefficient matrix 141 as its initial value. The component value revision section 140 revises component values of the watermark coefficient matrix 141 that is a complex matrix set initially to the null matrix, and a discrete inverse Fourier transform section 150 applies discrete inverse Fourier transform to the watermark coefficient matrix 141 to generate a watermark pattern 151. Then a watermark pattern addition section 110 emphasizes the watermark pattern by using a strength parameter 102, to provide an output of an imbedded image 105 that results from adding the watermark pattern to the input original image 101 in a form of tiles.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a digital watermark embedding method, a digital watermark detecting method, a digital watermark embedding device, a digital watermark detecting device, a storage medium storing a digital watermark embedding program, and a digital watermark detecting program. In particular, the present invention relates to a storage medium, a digital watermarking system, and an integrated circuit, in which, in order to realize copyright protection of digital content, the information of the content itself is changed by a very small amount that cannot be perceived by humans, and A digital watermark embedding method, a digital watermark detection method, a digital watermark embedding device, and a digital watermark detection device for embedding watermark information, which is information for preventing human perception, from detecting the watermark information. And a storage medium storing a digital watermark embedding program, and Storage medium storing the digital watermark detection program, and an electronic watermark system, and an integrated circuit.

In recent years, transmission of digital contents over a network has been facilitated by the advancement of digital information technology (eg, JPEG [Joint Picture Coding Expert Group], etc.) for audio, video, and still images. It has become to. In addition, distribution of contents via a network has begun with the spread of information infrastructure such as the Internet.

[0003] However, since the content delivered through the network is digital information, it is easy to make a complete copy using personal content or the like. Further, in an information infrastructure such as the Internet, all users can be information senders, so that illegal copying or copyright infringement such as selling illegal copying may be performed. Furthermore, with the improvement in the processing capability of personal computers, it becomes easier to process and edit image and audio data on personal computers, and copyrights such as unintended tampering by the original copyright owner and unauthorized secondary use are Infringement can also take place. These threats of copyright infringement make the authors hesitate to provide information and hinder the distribution of information through networks.

[0004] Separately from this, from the current content production site, even if the user wants to correctly process a certain content and perform a legitimate secondary use, he does not know where the right is and where should he inquire? There is a problem such as "I don't know." The present invention stores a digital watermark embedding method, a digital watermark detecting method, a digital watermark embedding device, a digital watermark detecting device, and a digital watermark embedding program for embedding and detecting a digital watermark in view of these circumstances. The present invention relates to a storage medium, a storage medium storing a digital watermark detection program, a digital watermark system, and an integrated circuit.

[0005]

2. Description of the Related Art Conventionally, a copyright management / protection system using a digital watermark technique has been proposed. As a technology related to digital watermarking technology, “A. Piva et al.:”DCT
-based Watermark Recovering without Resorting to t
he Uncorrupted Orignal Image ", Proceedings of the
1997 International Conference on Image Processing
(ICIP'97), 1997 ”(hereinafter referred to as Ref. 1) and“ Takao
Nakamura, Hiroshi Ogawa, Youichi Takashima,: "A Wat
ermarking Technique for Still Image "NTT R & D Vol.4
7, No. 6, 1998 "(hereinafter referred to as Ref. 2) and" Wisetsu
it Piyapisuit, Kineo Matsui ,: "Block Characteristi
c in Color Image Watermarking Using Equivalent Sig
nal Transform "National Conference of the Institut
e of Electronics, Information and Communication Eng
ineers, D-11-42, 1999 "(hereinafter referred to as Ref. 3).

The method shown in Ref. 1 embeds watermark information by obtaining a frequency component matrix by orthogonally transforming the image size and changing the coefficient value of the intermediate frequency band in the transmission frequency component matrix. The change of the coefficient value uses a spread spectrum technique. Thereby, it is resistant to irreversible compression, color conversion, and contrast change of an image. Also, Ref. 1
In reading the watermark information, the watermark information is read by despreading the coefficient value of the intermediate frequency band in the frequency component matrix obtained by orthogonally transforming the image to be examined. As a result, the enlargement / reduction of the image can be dealt with by comparing the size of the image with the original digital content and performing detection after adjusting the size to the same size as the original digital content.

In the method shown in Ref. 1, at the time of reading, the initial values of random numbers used in spread spectrum are tried one by one, and the one with a high response is read as watermark information. Thus, partial cutout and the like can be detected if it is known where the cutout portion corresponds to the original digital content. Next, the method shown in Ref. 2 above divides an image into large blocks according to the block size used in general lossy compression, performs orthogonal transformation for each block, obtains a frequency component matrix, and normalizes this. I do. Then, a coefficient sequence to be embedded is selected based on the key information, and watermark information is embedded by quantizing the coefficient value. The watermark-embedded image is obtained by performing inverse normalization and inverse orthogonal transformation on the changed matrix. In addition, the method improves the tolerance by changing the low frequency band.

[0008] The method shown in Ref. 3 mentioned above applies a conversion in which a conversion matrix for equivalent signal conversion of a color image (for example, RGB ← → YCbCr, etc.) is slightly changed from the defined one according to watermark information. Perform embedding. At the time of watermark information detection in Ref.3, the detection target image is converted into an equivalent signal, and detection is performed based on whether or not these statistical values are peculiar due to embedding. Thus, the method shown in Ref. 3 is resistant to geometrical modification of the image.

In embedding a digital watermark in a digital image, there is a method of adding a pattern called a watermark pattern with equal strength over the entire image to obtain an image in which a digital watermark is embedded.

[0010]

SUMMARY OF THE INVENTION
The method shown in Ref.1 has the following problems. I. Low resistance to partial clipping. II. Processing is heavy because orthogonal transform of original image size is used. III. Image quality is significantly degraded due to watermark embedding.

IV. Since the trial is performed for all possible values of the watermark information when detecting the watermark information, the processing is heavy. The method described in Ref. 2 has resistance to lossy compression, but has a problem that resistance to color conversion, contrast change, and partial clipping is low. In addition, normalization aims at improvement of image quality / resistance by embedding strong in a complicated part of an image and weakly in a flat part, but does not function very effectively. When watermark information is detected in Ref.2, processing up to normalization at the time of embedding is performed, and watermark information is detected from the key information and the normalized frequency component matrix. Thereby, when the block size is the same as the image size, it has resistance to enlargement / reduction. However, it does not quantitatively evaluate the reliability of the detected watermark information. Also,
There is a problem that image quality degradation due to watermark embedding is large.

The method shown in Ref. 3 is very weak in irreversible compression and color conversion. Also, in the case of multi-bit embedding, since one bit is embedded for each block, there is a problem that resistance to geometrical modification is weakened. further,
When embedding a digital watermark, a method of embedding a pattern called a watermark pattern with equal strength is generally perceived as being more likely to deteriorate in a flat part of an image than in a complicated part.
There is a problem that a watermark pattern can be added to an image only with such an intensity that deterioration is not perceived in a flat portion.

The present invention has been made in view of the above points, and enables the embedding process to be performed at a high speed. In the detection process, there is no original image, and a partial image corresponds to any part of the original image. A digital watermark embedding method, a digital watermark detection method, a digital watermark embedding device, and a digital watermark detection device capable of detecting a digital watermark from a partial image even when it is unknown.
It is another object of the present invention to provide a storage medium storing a digital watermark embedding program, a storage medium storing a digital watermark detection program, and a digital watermark system and an integrated circuit.

It is a further object of the present invention to use a method of calculating the local complexity of an image based on visual characteristics to change the strength of a watermark pattern according to the local complexity of the image, and to relatively change the strength of the watermark pattern. A digital watermark embedding method, a digital watermark detecting method, a digital watermark embedding device, a digital watermark detecting device, a storage medium storing the digital watermark embedding program, and a digital watermark detecting program capable of improving the resistance of the watermark And a digital watermark system and an integrated circuit.

[0015]

FIG. 1 is a diagram for explaining a first principle of the present invention. The present invention (Claim 1)
Is a digital watermark embedding method that embeds a digital watermark in a digital image so that it is not perceived by human perception. The real component and the imaginary component of each coefficient value of the watermark coefficient matrix, which is a matrix, are independently changed (step 1), and the watermark sequence matrix after the change is subjected to discrete inverse Fourier transform to generate a watermark pattern (step 2). The watermark pattern is added to the original image in a tile shape (step 3) to generate an embedded image (step 4).

According to the present invention (claim 2), embedding target component position information, which is a sequence of a length for embedding an electronic watermark, is generated from key information, and position marker information and an electronic watermark are converted into an embedding sequence using the key information. Spread, using the embedding target component position information and the embedding sequence, change the component value of the watermark coefficient matrix whose initial value is a complex matrix of zero matrix, perform a discrete inverse Fourier transform of the watermark coefficient matrix, and generate a watermark pattern. , Emphasizing the watermark pattern with the intensity parameter,
An embedded image is generated by adding the emphasized watermark pattern to the input image in a tiled manner.

According to the present invention (claim 3), when generating the position information of the component to be embedded, a random number sequence having the key information as an initial value is generated, and the position information of the component to be embedded is generated from the random number sequence. According to the present invention (claim 4), when spreading position marker information and a digital watermark into an embedded sequence, a 0th spreading sequence with key information as an initial value is generated, and the spreading sequence is a sequence representing position marker information. And store it by multiplying it by a predetermined number, convert the digital watermark into a symbol expression, obtain the total number of symbols and each symbol value, and generate a spreading sequence whose initial value is the sum of the key information and the symbol value, As a sequence representing a symbol, an average value of stored values is obtained in addition to each item of the stored spread sequence, and a value obtained by subtracting the average value from each item is used as an embedded sequence.

According to the present invention (claim 5), when generating a spreading sequence, either +1 or -1 is used as the spreading sequence. According to the present invention (claim 6), when spreading position marker information and a digital watermark into an embedded sequence, a spread sequence is generated using key information as an initial value, and a predetermined parameter is multiplied by the spread sequence and stored. Means for converting the digital watermark into a symbol representation, adding a spread sequence corresponding to the symbol of the symbol representation to each item of the storage means, and subtracting an average value from each term of the storage means to obtain an embedded sequence. And

According to the present invention (claim 7), when generating a spreading sequence, either +1 or -1 is used as the spreading sequence. The present invention (claim 8) provides a digital watermark detection method for detecting a digital watermark from a detection target image in which a digital image is embedded so that the digital watermark is not perceived by human perception. When detecting the embedded digital watermark, a block is cut out from an arbitrary position of the detection target image (step 5), the block is subjected to discrete Fourier transform, a coefficient matrix is obtained (step 6), and designated by key information. The detected target component position information to be detected is generated (step 7), and each term of the detected target component position information is multiplied by the phase difference of the coefficient according to the amount of translation to generate a position marker sequence (step 8). Offset information, which is the amount of parallel movement when the start point of the watermark pattern that is present and the start point of a pixel block cut out from the detection target image match ( Step 9) for detecting digital watermark that is embedded from the pixel block cut out on the basis of the offset information (step 10).

According to the present invention (claim 9), when generating the detection target component position information from the key information, a random number sequence having the key information as an initial value is generated, and the detection target component position information is generated from the random number sequence. . According to the present invention (claim 10), when a position marker sequence is generated, a 0th spreading sequence with key information as an initial value is generated, and the spreading sequence is output as a position marker sequence.

According to the present invention (claim 11), when extracting offset information, offset candidate information is generated, a pixel block is cut out from the position of the offset candidate information of the detection target image, and the pixel block is subjected to discrete Fourier transform. A detection target coefficient matrix is generated, a detection target sequence is generated from the detection target component position information and the detection target coefficient matrix, and a position marker sequence, a correlation value of the detection target sequence, and offset candidate information are set, and the position marker detection information is obtained. The offset candidate information generated when the correlation value between the position marker sequence and the detection target sequence is maximum is extracted as offset information.

According to the present invention (claim 12), when detecting a digital watermark, a pixel block of N × N pixels is cut out from a position translated from the detection target image by offset candidate information, and the pixel block is subjected to discrete Fourier transform. Generates a detection target coefficient matrix, generates a detection target sequence from the detection target component position information and the detection target coefficient matrix, generates symbol detection information from the detection target sequence and key information, and generates a detection symbol from the symbol detection information. Then, a detection result is detected as a digital watermark from each detection symbol and output.

According to the present invention (claim 13), when detecting a digital watermark, a sequence for each symbol position is generated for each symbol position from the symbol detection information, and a maximum value among the sequences for each symbol position is obtained. Let the symbol candidate be a detected symbol,
The detection symbol is subjected to inverse symbol conversion to obtain a detection result.

According to the present invention (claim 14), when detecting a digital watermark, a pixel block is output by cyclically shifting a cut-out pixel block based on offset information. According to the present invention (claim 15), when detecting a digital watermark, if the detection target image is equal to or larger than N × N size, a pixel block of N × N pixels is cut out from the detection target image, and the detection target image is extracted. Is smaller than the N × N size, an average pixel value of the detection target image portion included in the N × N size of the detection target image is obtained, and a portion less than the N × N size is filled with the average pixel value. , A pixel block is generated, and the extracted pixel block is cyclically shifted to output the pixel block.

According to the present invention (claim 16), when detecting a digital watermark, a pixel block is divided from an offset position of a detection target image, a discrete Fourier transform is performed on the divided pixel block, and a detection target coefficient matrix is obtained. Then, a detection target sequence is generated from the detection target component position information and the detection target coefficient matrix, symbol detection information is generated from the detection target sequence and key information, and a pixel block is output from the symbol detection information.

According to the present invention (claim 17), when detecting a digital watermark, an image to be detected is divided into pixel blocks, the pixel blocks are cyclically shifted by offset information, and the cyclically shifted pixel blocks are detected. Performs discrete Fourier transform, generates a detection target sequence matrix, generates a detection target sequence from the detection target component position information and the detection target coefficient matrix, generates symbol detection information from the detection target sequence and the key information, and generates symbol detection information from the symbol detection information. A detection symbol is generated, and a digital watermark is detected from the detection symbol.

According to the present invention (claim 18), when detecting a digital watermark, an image to be detected is divided into pixel blocks, and the pixel blocks are cyclically shifted by an offset. , The portion less than the N × N size is filled in with the average pixel value, the pixel block is subjected to discrete Fourier transform, a detection target coefficient matrix is generated, and the detection target component position information and the detection target coefficient matrix , A symbol detection information is generated from the detection target sequence and the key information, each detection symbol is generated from the symbol detection information, and a digital watermark is detected from the detection symbol.

According to the present invention (claim 19), when a position marker sequence is generated, an image to be detected is divided into blocks according to offset information, and an added block to which the pixel values of the pixel blocks are added is output as a pixel block. . According to the present invention (claim 20), when a position marker sequence is generated, a detection target image is divided into blocks, a block obtained by adding pixel values of the divided blocks is set as an addition block, and the addition block is The pixel data is cyclically shifted by the offset information to generate a pixel block.

According to the present invention (claim 21), when a position marker sequence is generated, an image to be detected is divided into blocks. At this time, a pixel area less than one block around an edge of the image to be detected is determined. Then, an average value corresponding to the image to be detected at the time of division is obtained and filled, and the portion that is still insufficient is obtained by averaging the pixel values corresponding to the image to be detected, and the average value is filled to form a block. A block obtained by adding the pixel values is set as an addition block, and the addition block is cyclically shifted based on the offset information to generate a pixel block.

According to the present invention (claim 22), when the position marker sequence is generated, the average of the detection target sequence is changed to 0 after the detection target sequence is generated. According to the present invention (claim 23), when detecting a digital watermark, after generating a detection target sequence,
The average of the detection target series is changed to 0. The present invention (Claim 2
4) When the offset information is extracted at the time of detecting the position marker sequence, the maximum value of all the position marker detection information is obtained. If the maximum value is smaller than a predetermined threshold, the offset detection has failed. Is output, and when the maximum value is equal to or greater than the predetermined threshold value, offset information for obtaining the maximum value is output.

According to the present invention (claim 25), when detecting a digital watermark, symbol detection information is divided for each symbol position to generate a sequence for each symbol position, and the maximum value of the sequence for each symbol position is obtained. If the maximum value is smaller than the predetermined threshold, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output.If the maximum value is equal to or larger than the predetermined threshold, the signal is output. The symbol candidate at the maximum value is set as a detected symbol, and the detected symbol is converted into a digital watermark expression.

In the present invention (claim 26), a threshold determined based on the value of a normal distribution function having a mean of 0 and a variance of 1 is applied as the threshold. The present invention (claim 27) provides a digital watermark detection method for detecting a digital image from a detection target image in which a digital watermark is embedded such that the digital image is not perceived by human perception. When detecting a digital watermark, the detection target component position information is generated from the key information, a position marker sequence is generated from the key information, and the start point of the watermark pattern at the time of embedding and the detection target are detected based on the detected position marker detection information. Offset information indicating a deviation from the start point of the pixel block cut out from the image is detected, and a digital watermark embedded in the pixel block is detected using the offset information.

According to the present invention (claim 28), when generating detection target component position information, a random number sequence having key information as an initial value is generated, and based on the random number sequence, the length of the embedded sequence embedded is determined. And the detection target component position information having the same length as
According to the present invention (claim 29), when a position marker sequence is generated, a first spread sequence with key information as an initial value is generated, and the spread sequence is used as a position marker sequence.

According to the present invention (claim 30), when detecting the position marker detection information, the detection target image, the detection target component position information, and the position marker sequence are used to detect the position marker detection information.
A pixel block of size N is cut out, a discrete Fourier transform is performed on the pixel block, a detection target coefficient matrix is generated, a detection target sequence is generated from the offset information, the detection target coefficient matrix, and the detection target component position information, and a position marker is generated. A set of correlation values between the sequence and the detection target sequence is used as position marker detection information.

According to the present invention (claim 31), when a position marker sequence is generated, an image to be detected is divided into N × N size blocks, and an added block in which the pixel values of all blocks are added is a pixel block. And The present invention (claim 3
2) When detecting a digital watermark, an image to be detected is
Using the detection target component position information, offset information, and key information, a pixel block is cut out from the detection target image, the pixel block is subjected to discrete Fourier transform, a detection target coefficient matrix is generated, and the offset information, the detection target coefficient matrix, A detection target sequence is generated from the detection target component position information, a symbol detection result is generated from a correlation value between the detection target sequence and a sequence generated from the key information and the symbol candidate information, and a symbol detection result is generated based on each symbol position of the symbol detection result. Then, a symbol candidate value having the maximum correlation value between the sequence generated from the key information and the symbol candidate information and the detection target sequence is used as a detected symbol, and the result of converting the detected symbol into a digital watermark format is output.

According to the present invention (claim 33), when a position marker sequence is generated, when an N × N size block is cut out from a detection target image, and when the N × N size block cannot be cut out from the detection target image, Cuts out only the detection target image equivalent included when trying to cut out in N × N size,
For a portion that is not enough in the N × N size, an average of pixel values corresponding to the detection target image is obtained, and the portion that is missing in the average is filled.

According to the present invention (claim 34), when detecting a digital watermark, a pixel block is divided from an image to be detected.
Perform a discrete Fourier transform of the pixel block, generate a detection target coefficient matrix, generate a detection target sequence from the offset information, the detection target component position information, and the detection target coefficient matrix, generate a detection symbol from the detection target sequence and the key information, A sub-image is detected from the detected symbol.

According to the present invention (claim 35), when dividing a pixel block from an image to be detected, N ×
If an N-size block cannot be cut out from the detection target image when an N-size block is cut out, only a portion of the detection target image that is included when trying to cut out the N-by-N size is cut out, and N × N About the part where the size is not enough,
An average value of pixel values corresponding to the image to be detected is obtained and filled.

According to the present invention (claim 36), when detecting a position marker sequence and when detecting a digital watermark, an image to be detected is divided into N × N size blocks, and the pixel values of all the blocks are The added block obtained by the addition is defined as a pixel block. According to the present invention (claim 37), when a block of N × N size is cut out from an image to be detected when detecting a digital watermark, if the block cannot be cut out to N × N size from the image to be detected, N × N size is used. Cut out only the part equivalent to the detection target image included when trying to cut out in N size, and N × N
For the portion where the size is insufficient, an average value of the pixel values corresponding to the image to be detected is obtained, and the portion where the average value is insufficient is filled.

According to the present invention (claim 38), when detecting a position marker sequence, a process of changing the average of the detection target sequence to 0 after generating the detection target sequence is performed. According to the present invention (claim 39), upon detecting a digital watermark in claim 32, after generating the detection target sequence, a process of changing the average of the detection target sequence to 0 is performed. According to the present invention (claim 40), in detecting the digital watermark in claim 34, after generating the detection target sequence, a process of changing the average of the detection target sequence to 0 is performed.

According to the present invention (claim 41), when the offset information is detected, the maximum value of the position marker detection information is obtained.
When the maximum value is smaller than the predetermined threshold, a signal indicating that the detection of the offset information has failed is output. When the maximum value is equal to or more than the predetermined threshold, the offset information at the time of obtaining the maximum value is output. I do.

According to the present invention (claim 42), when detecting a digital watermark, the symbol detection result is divided for each symbol position, and the maximum value of the symbol position sequence for each symbol position is determined. If the value is smaller than the threshold value, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output. If the maximum value is equal to or greater than the predetermined threshold value, the symbol at which the maximum value is taken is output. Output candidates.

According to the present invention (claim 43), a threshold determined based on the value of a distribution function of a normal distribution having an average of 0 and a variance of 1 is used as the threshold. FIG. 2 is a first principle configuration diagram of the present invention. The present invention (Claim 44) is an electronic watermark embedding device for embedding an electronic watermark in a digital image so as not to be perceived by human perception. Changing means 1A for independently changing the real component and imaginary component of each coefficient value of a certain watermark coefficient matrix, watermark pattern generating means 2A for generating a watermark pattern by performing a discrete inverse Fourier transform on the changed watermark sequence matrix, It has an adding unit 3A for adding a watermark pattern to the original image in a tile shape, and an image generating unit 4A for generating an embedded image.

According to the present invention (claim 45), embedding target component designation means for generating embedding target component position information which is a sequence of lengths for embedding an electronic watermark from key information, and position marker information using key information. An embedding sequence generating unit that spreads an electronic watermark into an embedding sequence, and a component value changing unit that changes a component value of a watermark coefficient matrix whose initial value is a complex matrix of zero matrix, using the embedding target component position information and the embedding sequence. , Performs a discrete inverse Fourier transform of the watermark coefficient matrix,
Discrete inverse Fourier transform means for generating a watermark pattern,
Watermark pattern adding means for emphasizing the watermark pattern with the intensity parameter and generating an embedded image obtained by adding the emphasized watermark pattern to the input image in a tiled manner.

According to the present invention (claim 46), the embedding target component designating means includes a random number generating means for generating a random number sequence having key information as an initial value, and means for generating embedding component position information from the random number sequence. Including. The present invention (Claim 4
7) In the embedded sequence generating means, a 0th spread sequence with the key information as an initial value is generated, and the spread sequence is a sequence representing position marker information, which is stored by multiplying by a predetermined number.
The digital watermark is converted to a symbol representation, the total number of symbols and each symbol value are obtained, a spreading sequence is generated with the initial value obtained by adding the symbol value to the key information, and the spreading sequence stored as the symbol sequence is stored. Means for obtaining an average value of stored values in addition to each term of the series, and using a value obtained by subtracting the average value from each term as an embedded series.

According to the present invention (claim 48), the embedded sequence generating means uses either +1 or -1 as the spread sequence. The present invention (claim 49) provides an embedding sequence generating unit that generates a spreading sequence using key information as an initial value, multiplies the spreading sequence by a predetermined parameter, stores the multiplication in a storage unit, stores a digital watermark in a symbol. Means for converting to a representation, adding a spreading sequence corresponding to the symbol of the symbol representation to each term of the storage means, and subtracting an average value from each term of the storage means as an embedded sequence.

According to the present invention (claim 50), the embedded sequence generating means uses either +1 or -1 as the spread sequence. The present invention (Claim 51) is an electronic watermark detection apparatus for detecting an electronic watermark from a detection target image in which an electronic watermark is embedded so that a digital image is not perceived by human perception. Detection target component specifying means 5A for generating detection target component position information from the position information, position marker sequence generation means 6A for generating a position marker sequence from the key information, a position marker sequence, and a starting point of a watermark pattern at the time of embedding. Position marker detection means 7A for extracting offset information when the start point of a pixel block cut out from the detection target image coincides, and digital watermark detection for detecting a digital watermark embedded from the pixel block cut out based on the offset information Means 8A
And

According to the present invention (claim 52), the detection target component designating means 5A includes means for generating a random number sequence having key information as an initial value, and includes means for generating detection target component position information from the random number sequence. . The present invention (claim 53) includes a means for generating a 0th spreading sequence with key information as an initial value in the position marker sequence generating means 6A and outputting the spreading sequence as a position marker sequence.

According to the present invention (claim 54), the position marker detecting means 7A generates offset candidate information, cuts out a pixel block from the position of the offset candidate information of the detection target image, and performs a discrete Fourier transform on the pixel block. Means for generating a detection target coefficient matrix; means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; and a set of a position marker sequence, a correlation value of the detection target sequence, and the offset candidate information. Means for generating position marker detection information, and means for extracting offset candidate information when the correlation value between the position marker sequence and the detection target sequence is maximum as offset information.

According to the present invention (claim 55), in the digital watermark detecting means 8A, means for extracting a pixel block of N × N pixels from a position translated from the detection target image by the offset candidate information, Means for converting and generating a detection target coefficient matrix, means for generating a detection target sequence from detection target component position information and the detection target coefficient matrix, means for generating symbol detection information from the detection target sequence and key information, Means for generating each detection symbol from the detection information and means for outputting a detection result from each detection symbol as an electronic watermark are included.

According to the present invention (claim 56), in the digital watermark detecting means 8A, means for generating a sequence for each symbol position from each symbol position from the symbol detection information, The candidate symbol at the time of taking is set as a detected symbol, and the detected symbol is subjected to inverse symbol conversion.
Means for acquiring a detection result.

The present invention (Claim 57) includes means for outputting a pixel block by cyclically shifting an extracted pixel block by offset in the digital watermark detection means 8A. According to the present invention (claim 58), in the digital watermark detecting means 8A, the detection target image is N × N.
If the size is greater than or equal to the
Means for cutting out a pixel block of pixels; and, if the detection target image is smaller than N × N,
Means for obtaining an average pixel value of a detection target image portion included in the × N size, filling a portion less than the N × N size with the average pixel value, and generating a pixel block; Means for outputting a pixel block by shifting.

According to the present invention (claim 59), in the digital watermark detecting means 8A, means for dividing a pixel block from an offset position of a detection target image, discrete Fourier transform of the divided pixel block, and a detection target coefficient matrix Means for generating detection target component position information, means for generating a detection target sequence from the detection target coefficient matrix, means for generating symbol detection information from the detection target sequence and key information, and a pixel block from the symbol detection information. Output means.

According to the present invention (claim 60), the digital watermark detecting means 8A divides the image to be detected into pixel blocks and cyclically shifts the pixel blocks by an offset. Means for performing a discrete Fourier transform to generate a detection target sequence matrix, means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, and means for generating symbol detection information from the detection target sequence and key information , Means for generating each detected symbol from the symbol detection information, and means for detecting a digital watermark from the detected symbol.

According to the present invention (claim 61), the digital watermark detecting means 8A divides the image to be detected into pixel blocks, and cyclically shifts the pixel blocks with an offset. Means for filling a portion less than the N × N size with an average pixel value, and a discrete Fourier transform of the pixel block,
Means for generating a detection target coefficient matrix, means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, means for generating symbol detection information from the detection target sequence and key information, and Means for generating each detected symbol and means for detecting a digital watermark from the detected symbol are included.

According to the present invention (claim 62), the position marker sequence generating means 6A divides the image to be detected into pixel blocks by offset, and outputs an addition block obtained by adding the pixel values of the pixel blocks as a pixel block. Means. According to the present invention (claim 63), in the position marker sequence generating means 6A, means for dividing the detection target image into blocks, and a block obtained by adding the pixel values of the divided blocks as an addition block, Means for cyclically shifting the block by the offset to generate a pixel block is included.

According to the present invention (claim 64), the position marker sequence generating means 6A divides an image to be detected into blocks, and at this time, for a pixel area less than one block around an end of the image to be detected. Means for obtaining an average value corresponding to the image to be detected at the time of division, obtaining an average value of pixel values corresponding to the image to be detected for a portion that is not sufficient, filling the average value to form a block, The block includes a unit obtained by adding the values to be an addition block and a unit that cyclically shifts the addition block by an offset to generate a pixel block.

The present invention (claim 65) includes a means for changing the average of the detection target sequence to 0 after the detection target sequence is generated in the position marker sequence generation means 6A. The present invention (Claim 66) includes means for changing the average of the detection target sequence to 0 after generating the detection target sequence in the digital watermark detection means 8A. According to the present invention (claim 67), in the position marker detecting means 7A, the maximum value of all the position marker detection information is obtained, and if the maximum value is smaller than a predetermined threshold value, it indicates that the offset detection has failed. Means for outputting a signal, and when the maximum value is equal to or greater than the predetermined threshold value, outputting an offset when the maximum value is obtained.

According to the present invention (claim 68), the digital watermark detecting means 8A divides the symbol detection information for each symbol position to generate a sequence for each symbol position, and obtains the maximum value of the sequence for each symbol position. Means for outputting a signal indicating that detection of a digital watermark embedded in the detection target image has failed when the maximum value is smaller than a predetermined threshold value, and outputting the signal when the maximum value is equal to or more than the predetermined threshold value. Includes means for using the symbol candidate at the time of taking the maximum value as a detection symbol, and means for converting the detection symbol into a digital watermark expression.

According to the present invention (claim 69), a threshold determined based on the value of a normal distribution function having a mean of 0 and a variance of 1 is applied as the threshold. The present invention (claim 70)
Is a digital watermark detection device for detecting a digital watermark embedded in a digital image so as not to be perceived by human perception, and a detection target component designation means for generating detection target component position information from key information; A position marker sequence generating means for generating a position marker sequence from the key information, and a detected position marker detection information, which is used to determine a shift between a start point of a watermark pattern at the time of embedding and a start point of a pixel block cut out from a detection target image. It has position marker detection means for detecting offset information to be represented, and digital watermark detection means for detecting a digital watermark embedded in a pixel block using the offset information.

According to the present invention (claim 71), in the detection target component designating means, a means for generating a random number sequence with key information as an initial value is provided. Means for generating length detection target component position information is included. The present invention (claim 72) includes a means for generating a first spreading sequence with key information as an initial value in the position marker sequence generating means, and using the spreading sequence as a position marker sequence.

According to the present invention (claim 73), in the position marker detecting means, an N × N size pixel block is cut out from the detection target image using the detection target image, the detection target component position information, and the position marker series. Means for generating a detection target coefficient matrix by performing a discrete Fourier transform of the pixel block, means for generating a detection target sequence from offset information, a detection target coefficient matrix, and detection target component position information, and detecting a position marker sequence. Means for combining the correlation values with the target series to obtain position marker detection information.

According to the present invention (claim 74), in the digital watermark detecting means, means for extracting a pixel block from the detection target image using the detection target image, the detection target component position information, the offset information and the key information, Means for generating a detection target coefficient matrix by performing a discrete Fourier transform of the pixel block; means for generating a detection target sequence from offset information, a detection target coefficient matrix, and detection target component position information; and, from key information and symbol candidate information. Means for generating a symbol detection result from a correlation value between the generated sequence and the detection target sequence,
Means for determining a symbol candidate value having a maximum correlation value between a sequence generated from key information and symbol candidate information and a detection target sequence based on each symbol position of the symbol detection result as a detection symbol; And a means for outputting the result of conversion into the format.

According to the present invention (claim 75), in the position marker sequence generating means, when a block of N × N size is cut out from the image to be detected and the block cannot be cut out to N × N size from the image to be detected, Only the portion corresponding to the detection target image included when trying to cut out in the N × N size is cut out, and the portion that is not sufficient in the N × N size is obtained by averaging the pixel values corresponding to the detection target image. Includes means to fill in missing parts.

According to the present invention (claim 76), in the digital watermark detection means, means for dividing a pixel block from an image to be detected, means for performing a discrete Fourier transform on the pixel block to generate a coefficient matrix for a detection, and offset information Means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; means for generating a detection symbol from the detection target sequence and the key information; and detecting a sub-image from the detection symbol.

The present invention (Claim 77) provides a digital watermark detecting means that, when a block of N × N size is cut out from an image to be detected, if the block cannot be cut out to N × N size from the image to be detected, XN only the image corresponding to the image to be detected included when trying to extract in size
The portion that is not enough in the size xN includes means for obtaining an average value of pixel values corresponding to the image to be detected and filling the portion that is not enough with the average value.

According to the present invention (claim 78), in the position marker series generating means and the digital watermark detecting means, means for dividing the detection target image into N × N size blocks and the pixel values of all the blocks are added. Means for making the addition block a pixel block is included.

According to the present invention (claim 79), in the position marker sequence generating means and the digital watermark detecting means, when a block of N × N size is cut out from a detection target image, the block can be cut out to N × N size from the detection target image. Otherwise, N
Only the portion equivalent to the detection target image included when attempting to cut out in the size of × N is cut out, and the portion that is not sufficient in the size N × N is obtained by averaging the pixel values corresponding to the detection target image. Includes means to fill in missing parts.

According to the present invention (claim 80), the position marker sequence generating means includes means for performing a process of changing the average of the detection target sequence to 0 after generating the detection target sequence.
The present invention (claim 81) includes means for performing, in the digital watermark detection means, a process of changing the average of the detection target sequence to 0 after generating the detection target sequence.

According to the present invention (claim 82), the position marker detecting means determines the maximum value of the position marker detection information, and the detection of the offset information fails if the maximum value is smaller than a predetermined threshold value. And means for outputting offset information when the maximum value is obtained when the maximum value is equal to or greater than a predetermined threshold value.

According to the present invention (claim 83), in the digital watermark detecting means, means for dividing each detection target symbol generated from the detection target sequence and the key information into symbol positions, and a maximum of the symbol position sequence for each symbol position. Means for calculating a value, and when the maximum value is smaller than a predetermined threshold value, a signal indicating that detection of a digital watermark embedded in the detection target image has failed is output, and the maximum value is not less than a predetermined threshold value. In this case, means for outputting a symbol candidate when the maximum value is taken is included.

According to the present invention (claim 84), a threshold determined based on the value of a normal distribution function having a mean of 0 and a variance of 1 is used as the threshold. The present invention (claim 85) provides
A storage medium storing a digital watermark embedding program for embedding a digital watermark in a digital image so as not to be perceived by a human perception, wherein each of a key information and a watermark coefficient matrix which is a complex matrix is obtained from the digital watermark embedded in the digital image. A process of independently changing the real component and the imaginary component of the coefficient value, a process of generating a watermark pattern by performing a discrete inverse Fourier transform on the changed watermark sequence matrix, and a process of adding the watermark pattern to the original image in a tiled manner, Generating an embedded image.

According to the present invention (claim 86), an embedding target component designation process for generating an embedding target component position information which is a sequence of lengths for embedding an electronic watermark from key information, and a position marker information using key information. An embedding sequence generation process of spreading an electronic watermark into an embedding sequence, and a component value changing process of changing a component value of a watermark coefficient matrix whose initial value is a complex matrix of zero matrix using the embedding target component position information and the embedding sequence. A discrete inverse Fourier transform process for generating a watermark pattern by performing a discrete inverse Fourier transform on a watermark coefficient matrix, and emphasizing the watermark pattern with an intensity parameter, and adding the enhanced watermark pattern to the input image in a tiled manner. And a watermark pattern adding process for generating an embedded image.

According to the present invention (claim 87), the embedding target component designation process includes a random number generation process for generating a random number sequence with key information as an initial value, and a process for generating the embedding target component position information from the random number sequence. Including. According to the present invention (claim 88), in an embedded sequence generation process, a 0th spread sequence with key information as an initial value is generated,
The spread sequence is set as a sequence representing position marker information, stored by multiplying by a predetermined number, and the digital watermark is converted into a symbol expression.
Acquire the total number of symbols and each symbol value, generate a spreading sequence with the initial value obtained by adding the symbol value to the key information,
As a sequence representing a symbol, in addition to each item of the stored spread sequence, an average value of stored values is obtained, and a value obtained by subtracting the average value from each item is used as an embedded sequence.

In the present invention (claim 89), in the process of generating an embedded sequence, a spread sequence is generated using key information as an initial value, a predetermined parameter is multiplied by the spread sequence, the result is stored in storage means, and a digital watermark is generated. The method includes a process of converting into a symbol representation, adding a spreading sequence corresponding to the symbol of the symbol representation to each term of the storage means, and subtracting an average value from each term of the storage means as an embedded sequence.

According to the present invention (claim 90), in the embedded sequence generation process, +1 or −
Use any one of 1. The present invention (claim 91) is a storage medium storing a digital watermark detection program for detecting a digital image from a detection target image in which a digital watermark for preventing the digital image from being sensed by human perception is embedded. A detection target component designation process for generating detection target component position information from the key information; a position marker sequence generation process for generating a position marker sequence from the key information;
A position marker detection process for detecting a position marker sequence and extracting offset information when a start point of a watermark pattern at the time of embedding and a start point of a pixel block cut out from a detection target image match, and cut out based on the offset information A digital watermark detection process for detecting a digital watermark embedded from the pixel block.

The present invention (claim 92) includes a process of generating a random number sequence with key information as an initial value in the detection target component designation process, and a process of generating detection target component position information from the random number sequence. The present invention (claim 93)
Includes a process of generating a 0th spreading sequence with key information as an initial value in a position marker sequence generating process, and outputting the spreading sequence as a position marker sequence.

According to the present invention (claim 94), in a position marker detection process, offset candidate information is generated, a pixel block is cut out from the position of the offset candidate information of the detection target image, and a discrete Fourier transform is performed on the pixel block. A process of generating a detection target coefficient matrix, a process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, and a set of a position marker sequence, a correlation value of the detection target sequence, and the offset candidate information. , A process of generating position marker detection information, and a process of extracting offset candidate information when the correlation value between the position marker sequence and the detection target sequence is the maximum as offset information.

According to the present invention (claim 95), in the digital watermark detection process, a process of cutting out a pixel block of N × N pixels from a position translated from the detection target image by the offset candidate information, and a discrete Fourier transform of the pixel block A process of generating a detection target coefficient matrix; a process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; a process of generating symbol detection information from the detection target sequence and key information; The method includes a process of generating each detection symbol from information and a process of outputting a detection result from each detection symbol as a digital watermark.

According to the present invention (claim 96), in the digital watermark detection process, a process of generating a sequence for each symbol position from each symbol position from the symbol detection information, and taking the maximum value among the sequences for each symbol position. A symbol candidate at the time as a detected symbol, converting the detected symbol into an inverse symbol, and obtaining a detection result.

The present invention (Claim 97) includes, in the digital watermark detection process, a process of outputting a pixel block by cyclically shifting an extracted pixel block by an offset. The present invention (claim 98)
In the digital watermark detection process, when the detection target image is equal to or larger than N × N size, a process of cutting out a pixel block of N × N pixels from the detection target image, and a case where the detection target image is smaller than N × N size In this case, a process of obtaining an average pixel value of a detection target image portion included in the N × N size of the detection target image, filling a portion less than the N × N size with the average pixel value, and generating a pixel block When,
The process includes a process of outputting a pixel block by cyclically shifting the extracted pixel block.

According to the present invention (claim 99), in the digital watermark detection process, a process of dividing a pixel block from an offset position of a detection target image, a discrete Fourier transform of the divided pixel block, and a detection target coefficient matrix are performed. A process of generating, a process of generating a detection target sequence from the detection target component position information and a detection target coefficient matrix, a process of generating symbol detection information from the detection target sequence and key information, and outputting a pixel block from the symbol detection information Process.

According to the present invention (claim 100), in a digital watermark detection process, a detection target image is divided into pixel blocks, and the pixel blocks are cyclically shifted by an offset. Fourier transform, a process of generating a detection target sequence matrix, a process of generating a detection target sequence from detection target component position information and a detection target coefficient matrix, and a process of generating symbol detection information from the detection target sequence and key information, The method includes a process of generating each detected symbol from the symbol detection information and a process of detecting a digital watermark from the detected symbol.

According to the present invention (claim 101), in a digital watermark detection process, an image to be detected is divided into pixel blocks, and the pixel blocks are cyclically shifted with an offset. For a portion less than the N × N size, a process of filling the missing portion with an average pixel value, a process of performing a discrete Fourier transform on a pixel block to generate a detection target coefficient matrix, a detection target component position information and a detection target coefficient A process of generating a detection target sequence from a matrix, a process of generating symbol detection information from the detection target sequence and key information, a process of generating each detection symbol from the symbol detection information, and a process of detecting a digital watermark from the detection symbol. including.

According to the present invention (claim 102), in the position marker series generation process, a process of dividing a detection target image into blocks by an offset and a process of outputting an added block obtained by adding pixel values of a pixel block as a pixel block. And The present invention (claim 103) provides
In the position marker sequence generation process, a process of dividing the detection target image into blocks, and a block obtained by adding the pixel values of the divided blocks to an addition block, and cyclically shifting the addition block by an offset, Includes the process of creating blocks.

According to the present invention (claim 104), in the position marker series generation process, the detection target image is divided into blocks, and at this time, a pixel area less than one block around the end of the detection target image is divided. The process of obtaining an average value corresponding to the image to be detected at the time, obtaining the average value of the pixel values corresponding to the image to be detected in the missing portion, filling the average value into a block, and the pixel value of the divided block Are added to each other, and a process of cyclically shifting the added block by offset to generate a pixel block is included.

The present invention (Claim 105) includes a process of changing the average of the detection target sequence to 0 after the generation of the detection target sequence in the position marker sequence generation process. The present invention (claim 106) includes, in the digital watermark detection process, a process of changing the average of the detection target sequence to 0 after generating the detection target sequence. The present invention (claim 10
7) includes a process of, after generating a detection target sequence in the digital watermark detection process, changing the average of the detection target sequence to zero.

According to the present invention (claim 108), in the position marker detection process, the maximum value of all the position marker detection information is obtained, and if the maximum value is smaller than a predetermined threshold value, the offset detection fails. And outputting the offset when the maximum value is obtained when the maximum value is equal to or larger than the predetermined threshold value.

The present invention (claim 109) provides a digital watermark detection process in which symbol detection information is divided for each symbol position to generate a sequence for each symbol position, and a process for obtaining the maximum value of the sequence for each symbol position. When,
If the maximum value is smaller than the predetermined threshold, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output.If the maximum value is equal to or larger than the predetermined threshold, the signal is output. The process includes a process of setting a symbol candidate at the maximum value as a detection symbol and a process of converting the detection symbol into a digital watermark expression.

According to the present invention (claim 110),
A threshold determined based on the value of the distribution function of a normal distribution with mean 0 and variance 1 is applied. The present invention (claim 111) provides:
A storage medium storing a digital watermark detection program for detecting a digital watermark embedded in a digital image so as not to be perceived by human perception, and a detection target component for generating detection target component position information from key information A designation process, a position marker sequence generation process of generating a position marker sequence from key information, and a start point of a watermark pattern at the time of embedding and a start point of a pixel block cut out from a detection target image based on the detected position marker detection information. And a digital watermark detection process for detecting a digital watermark embedded in a pixel block using the offset information.

According to the present invention (claim 112), in the detection target component designation process, a process of generating a random number sequence using key information as an initial value is performed based on the random number sequence and has the same length as the length of the embedded sequence embedded. The method includes a process of generating position detection target component position information. The present invention (Claim 113)
Includes a process of generating a first spread sequence using key information as an initial value in a position marker sequence generation process, and using the spread sequence as a position marker sequence.

According to the present invention (claim 114), in the position marker detection process, the detection target image, the detection target component position information, and the position marker sequence are used to detect N
A process of cutting out a pixel block of × N size, a process of generating a detection target coefficient matrix by performing a discrete Fourier transform of the pixel block, and a generation of a detection target sequence from offset information, a detection target coefficient matrix, and detection target component position information And a process of setting a correlation value between the position marker sequence and the detection target sequence as a set to obtain position marker detection information.

According to the present invention (claim 115), in the digital watermark detection process, a pixel block is cut out from the detection target image using the detection target image, the detection target component position information, the offset information and the key information, A process of generating a detection target coefficient matrix by performing a discrete Fourier transform on a pixel block, a process of generating a detection target sequence from offset information, a detection target coefficient matrix, and detection target component position information, and a process of generating key information and symbol candidate information. A process of generating a symbol detection result from a correlation value between the generated sequence and the detection target sequence; and a process of setting a symbol candidate value having a maximum correlation value as a detection symbol based on each symbol position of the symbol detection result. Outputting the result of converting the detected symbols into a digital watermark format.

According to the present invention (claim 116), in a position marker sequence generation process, when a block of N × N size is cut out from a detection target image, N × N size blocks are extracted from the detection target image.
If the image cannot be cut out to the N size, only the image to be detected included in an attempt to cut out the image to the N × N size is cut out. And filling the missing parts with the average.

The present invention (claim 117) provides a digital watermark detection process in which a pixel block is divided from a detection target image, a process in which the pixel block is subjected to discrete Fourier transform to generate a detection target coefficient matrix, and offset information. A process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, a process of generating a detection symbol from the detection target sequence and the key information, and detecting a sub-image from the detection symbol.

According to the present invention (claim 118), in the digital watermark detection process, when a block of N × N size is cut out from an image to be detected, if the block cannot be cut out to N × N size from the image to be detected, N Only the portion equivalent to the detection target image included when attempting to cut out in the size of × N is cut out, and the portion that is not sufficient in the size N × N is obtained by averaging the pixel values corresponding to the detection target image. Includes the process of filling in missing parts.

According to the present invention (claim 119), in the position marker series generation process, the process of dividing the detection target image into N × N size blocks and the addition block in which the pixel values of all the blocks are added are performed in the pixel block. Process. According to the present invention (claim 120), in the digital watermark detection process, a process of dividing a detection target image into N × N size blocks and a process of setting an addition block obtained by adding pixel values of all blocks to a pixel block are used. Including.

According to the present invention (claim 121), in the digital watermark detection process, when a block of N × N size is cut out from an image to be detected, if the block cannot be cut out to N × N size from the image to be detected, N Only the portion equivalent to the detection target image included when attempting to cut out in the size of × N is cut out, and the portion that is not sufficient in the size N × N is obtained by averaging the pixel values corresponding to the detection target image. Includes the process of filling in missing parts.

According to the present invention (claim 122), in a position marker series generation process, when a block of N × N size is cut out from a detection target image, N × N size blocks are extracted from the detection target image.
If the image cannot be cut out to the N size, only the image to be detected included in an attempt to cut out the image to the N × N size is cut out. And a process of filling in a portion where the average value is insufficient.

The present invention (Claim 123) includes, in the position marker sequence generation process, a process of generating a detection target sequence and then performing a process of changing the average of the detection target sequence to zero. The present invention (claim 124) includes, in the digital watermark detection process, a process of performing a process of changing the average of the detection target sequence to 0 after generating the detection target sequence.

According to the present invention (claim 125), in the position marker detection process, the process of obtaining the maximum value of the position marker detection information and the step of detecting the offset information when the maximum value is smaller than a predetermined threshold value have failed. And outputting the offset information when the maximum value is obtained when the maximum value is equal to or greater than a predetermined threshold value.

According to the present invention (claim 126), in the digital watermark detection process, a process of dividing the symbol detection information generated from the detection target sequence and the key information into each symbol position, and a maximum of the symbol position sequence for each symbol position are performed. When the maximum value is smaller than a predetermined threshold, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output, and the maximum value is equal to or larger than the predetermined threshold. In such a case, a process of outputting a symbol candidate when the maximum value is obtained is included.

According to the present invention (claim 127), in the digital watermark detection process, a process of dividing each detection target symbol and a symbol detection information generated from key information into symbol positions, When the maximum value is smaller than a predetermined threshold, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output, and the maximum value is equal to or larger than the predetermined threshold. In such a case, a process of outputting a symbol candidate when the maximum value is obtained is included.

The present invention (claim 128) provides claim 126
In the above, a threshold determined based on the value of a distribution function of a normal distribution having an average of 0 and a variance of 1 is used as the threshold. The present invention (claim 129) is based on claim 127, wherein the threshold value is a threshold value determined based on the value of the distribution function of a normal distribution with mean 0 and variance 1. The present invention (Claim 130)
Is a digital watermark embedding device that embeds a digital watermark in a digital image so as not to be perceived by human perception, and a digital watermark detection device that detects the digital watermark. ,
Using the key information, an embedding target component specifying unit that generates embedding target component position information that is a sequence having the same length as the length of embedding the digital watermark from the key information for embedding,
An embedding sequence generating unit that spreads the position marker information and the digital watermark to the embedding sequence, and a component that changes the component value of a watermark coefficient matrix whose initial value is a complex matrix of zero matrix using the embedding target component position information and the embedding sequence Value changing means, discrete inverse Fourier transform of a watermark coefficient matrix to generate a watermark pattern, and a discrete inverse Fourier transform means for generating a watermark pattern, the watermark pattern is enhanced by an intensity parameter, and the enhanced watermark pattern is tiled on the input image. A watermark pattern adding unit that generates an embedded image obtained by addition; a digital watermark detection device includes a detection target component designating unit that generates detection target component position information from key information; and a position marker sequence from the key information. A position marker sequence generating means for generating a position marker sequence, detecting a position marker sequence, Position marker detecting means for extracting offset information at which the start point of the pixel block cut out from the output target image matches, and digital watermark detecting means for detecting a digital watermark embedded from the pixel block cut out based on the offset information. Have.

According to the present invention (claim 131), the digital watermark embedding device is constituted by an integrated circuit. According to the present invention (claim 132), the digital watermark detection device is configured by an integrated circuit. The present invention (claim 133) provides a digital watermark embedding device that embeds a digital watermark in a digital image so as not to be perceived by human perception, and a digital watermark detection device that detects the digital watermark. The digital watermark embedding apparatus uses an embedding target component designating unit that generates, from key information for embedding, embedding target component position information that is a sequence having the same length as the length of embedding an electronic watermark, and key information, An embedding sequence generating unit that spreads the position marker information and the digital watermark to the embedding sequence, and a component that changes the component value of the watermark coefficient matrix whose initial value is a complex matrix of zero matrix using the embedding target component position information and the embedding sequence Value changing means, discrete inverse Fourier transform means for performing a discrete inverse Fourier transform on a watermark coefficient matrix, and a watermark pattern, A watermark pattern adding means for emphasizing the watermark pattern with an intensity parameter and adding an enhanced watermark pattern to the input image in a tiled manner to generate an embedded image obtained. Detection target component specifying means for generating detection target component position information from the information; position marker sequence generation means for generating a position marker sequence from the key information; and a start point of the watermark pattern at the time of embedding based on the detected position marker detection information. Position marker detecting means for detecting offset information indicating a deviation from a start point of a pixel block cut out from the detection target image, and digital watermark detecting means for detecting a digital watermark embedded in the pixel block using the offset information. Having.

According to the present invention (claim 134), the digital watermark embedding device is constituted by an integrated circuit. The present invention (claim 13
In 5), the digital watermark detection device is configured by an integrated circuit. FIG. 3 is a diagram for explaining the second principle of the present invention. The present invention (claim 136) provides a digital watermark embedding method for embedding a digital watermark in a digital image, wherein the local complexity of the image is obtained (step 11). Is changed to a part that is less complex than that part (step 1).
2) Acquire an embedded image (step 13).

According to the present invention (claim 137), when creating a local complexity of an image having a large value in a texture region in an image and a small value in a flat region or an edge portion, When the difference between the image, each pixel, and an adjacent pixel is equal to or greater than a certain threshold, the edge and the texture area are measured by a counter that is incremented.

According to the present invention (claim 138), an edge / texture component value image obtained by measuring an edge and a texture area is subdivided into minute blocks, and the edge / texture component value in the block is divided for each subdivided block. , And the sum is used as a texture degree index of the block.

According to the present invention (claim 139), an edge / texture component value image obtained by measuring an edge and a texture region is subdivided into minute blocks, and the edge / texture component value in the block is divided for each subdivided block. , And a value obtained by mapping the sum with a predetermined function is used as a texture degree index of the block.

According to the present invention (claim 140), the edge / texture component value is divided into minute blocks using the edge / texture component value image and the texture degree index, and each edge / texture component in the divided block is divided. The texture component value image is obtained by multiplying the value by the texture degree index corresponding to the value. The present invention (claim 141) uses an edge / texture component value image and a texture degree index,
Divide the edge / texture component value into small blocks,
A texture component value image is obtained by multiplying each edge / texture component value in the divided block by the texture degree index corresponding to the value.

According to the present invention (claim 142), when the image is changed in accordance with the local complexity of the image, the difference between the input image and each pixel and the adjacent pixel is determined as the local complexity of the image.
An edge / texture component value image is obtained by measuring an edge and a texture area based on the value of a counter that is incremented when the threshold value is equal to or greater than a certain threshold value.
The texture component value image is used as the local complexity of the image.

According to the present invention (claim 143), when an image is changed in accordance with the local complexity of an image, an edge / texture component obtained by measuring an edge and a texture region as the local complexity of the image. The value image is subdivided into minute blocks, and the sum of the edge / texture component values in each subdivided block is obtained. The sum is used as the texture index of the block, and an image obtained by up-sampling the texture index is obtained. Is used as the local complexity of

According to the present invention (claim 144), when an image is changed in accordance with the local complexity of an image, an edge / texture obtained by measuring an edge and a texture area as the local complexity of the image. The component value image is subdivided into minute blocks, and the edge
The sum of the texture component values is obtained, the value obtained by mapping the sum with a predetermined function is used as the texture index of the block, and the upsampled texture index is used as the local complexity of the image.

The present invention (claim 145) uses an edge / texture component value image and a texture degree index as the local complexity of an image when changing the image in accordance with the local complexity of the image. Dividing the edge / texture component value into minute blocks, and multiplying each edge / texture component value in the divided block by a texture index corresponding to the texture / component value image to obtain a texture component value image; Use the image as the local complexity of the image.

The present invention (claim 146) provides a digital watermark embedding method for embedding a digital watermark in a digital image so as not to be perceived by human perception, so that another digital watermark is not perceived in the digital image. When embedding, the local complexity of the image is generated from the input image which is a digital image, the watermark pattern is generated from the digital watermark using the key information necessary for embedding, and the watermark pattern is generated according to the local complexity of the image. Is added to the input image to generate an information embedded image.

According to the present invention (claim 147), when generating the local complexity of an image, a difference counter is incremented when a difference between each pixel of an input image and an adjacent pixel is larger than a predetermined threshold value. Then, the value of the difference counter at the time when the processing for the adjacent pixel is completed is set as the edge / texture component value image, and the edge / texture component value image is set as the local complexity of the image.

According to the present invention (claim 148), when generating the local complexity of an image, the difference counter is incremented when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value. Then, the value of the difference counter at the time when the processing for the adjacent pixels is completed is defined as an edge / texture component value image, the edge / texture component image is subdivided into blocks, and the sum of the edge / texture component values of the block is calculated. The texture degree index of the block is used, and an up-sample of the texture degree index is used as the local complexity of the image.

According to the present invention (claim 149), when generating the local complexity of an image, the difference counter is incremented when the difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value. Then, the value of the difference counter at the time when the processing for the adjacent pixels is completed is defined as an edge / texture component value image, the edge / texture component image is subdivided into blocks, and the sum of the edge / texture component values of the block is calculated. The texture / index is multiplied by the edge / texture component image of each block, and a texture component value image is obtained as the local complexity of the image.

The present invention (claim 150) provides a digital watermark embedding method for embedding a digital watermark in a digital image so as not to be perceived by human perception, so that another digital watermark is not perceived in the digital image. When embedding, the input image, which is a digital image, is subdivided into blocks of a predetermined size, the local complexity of the image is generated, and key information necessary for embedding is used. Generate a watermark pattern, generate an adaptive watermark pattern emphasizing the corresponding orthogonal transform coefficient of the basic watermark pattern according to the amplitude value of the orthogonal variable coefficient of the block, and combine the basic pattern and the watermark pattern with the local complexity and intensity of the image. Add to the block according to the parameter,
Generate an information embedded image.

According to the present invention (claim 151), when the local complexity of an image is generated, the difference counter is incremented when the difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value. Then, the value of the difference counter at the time when the processing for the adjacent pixel is completed is set as the edge / texture component value image, and the edge / texture component value image is set as the local complexity of the image.

According to the present invention (claim 152), when the local complexity of an image is generated, a difference counter is incremented when a difference between each pixel of an input image and an adjacent pixel is larger than a predetermined threshold value. Then, the value of the difference counter at the time when the processing for the adjacent pixels is completed is defined as an edge / texture component value image, the edge / texture component image is subdivided into blocks, and the sum of the edge / texture component values of the block is calculated. The texture degree index of the block is used, and an up-sample of the texture degree index is used as the local complexity of the image.

According to the present invention (claim 153), a value obtained by mapping a texture index generated from an edge / texture component value image using a predetermined function is used as a texture index. According to the present invention (claim 154), when the local complexity of an image is generated, a difference counter is incremented by 1 when a difference between each pixel of an input image and an adjacent pixel is larger than a predetermined threshold value, The value of the difference counter at the time when the processing for the adjacent pixel is completed is defined as an edge / texture component value image,
The edge / texture component image is subdivided into blocks, and the sum of the edge / texture component values of the block is calculated as a texture index of the block. The component value image is obtained and is set as the local complexity of the image.

FIG. 4 is a block diagram showing the second principle of the present invention. The present invention (claim 155) is an electronic watermark embedding device for embedding an electronic watermark in a digital image so as not to be perceived by human perception. Local complexity creating means 1B for obtaining complexity, watermark pattern creating means 2B for creating a pattern representing a digital watermark using key information necessary for embedding, and watermark pattern creating means for an input image. An image adding unit 3B that adds the pattern and the local complexity of the image to generate an information-embedded image.

According to the present invention (claim 156), the local complexity creating means 1B adds 1 to a difference counter when the difference between each pixel of an input image and an adjacent pixel is larger than a predetermined threshold value. It includes means for setting the value of the difference counter when the processing for the adjacent pixels is completed as an edge / texture component value image, and means for setting the edge / texture component value image as the local complexity of the image.

According to the present invention (claim 157), the local complexity creating means 1B adds 1 to a difference counter when the difference between each pixel of an input image and an adjacent pixel is larger than a predetermined threshold value. Means for setting the value of the difference counter when the processing for the adjacent pixel is completed as an edge / texture component value image; subdividing the edge / texture component image into blocks; taking the sum of the edge / texture component values of the block; It includes a means for setting the texture degree index of the block and a means for setting an up-sample of the texture degree index as the local complexity of the image.

According to the present invention (claim 158), in the local complexity generating means 1B, means for generating an edge / texture component value image from an input image and means for generating a texture index from the edge / texture component value image. And means for mapping the texture index using a predetermined function to obtain the local complexity of the image.

According to the present invention (claim 159), the local complexity creating means 1B adds 1 to a difference counter when a difference between each pixel of an input image and an adjacent pixel is larger than a predetermined threshold value. Means for setting the value of the difference counter when the processing for the adjacent pixel is completed as an edge / texture component value image; subdividing the edge / texture component image into blocks; taking the sum of the edge / texture component values of the block; Means for setting a texture index of the block; and means for multiplying the edge / texture component image of each block by the texture index to obtain a texture component value image and setting the local complexity of the image.

The present invention (Claim 160) is an electronic watermark embedding device for embedding an electronic watermark in a digital image so as not to be perceived by human perception. Block dividing means for subdividing the input image into blocks of a predetermined size, basic watermark pattern generating means for creating a basic watermark pattern representing the digital watermark from key information used for embedding and a digital watermark to be embedded in the input image, Adaptive watermark pattern creating means for creating an adaptive pattern that emphasizes the corresponding orthogonal transform coefficient of the basic watermark pattern according to the amplitude value of the orthogonal transform coefficient of the block divided by the block dividing means,
A local complexity generating means for obtaining a local complexity of an image from a block, and a basic watermark pattern and an adaptive watermark pattern for each block are classified into blocks according to an intensity parameter representing the local complexity of the image and the strength of embedding. And add
Adding means for generating an information embedded image.

According to the present invention (claim 161), the local complexity generating means adds 1 to the difference counter when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value. It includes means for setting the value of the difference counter when the processing for the pixel is completed as an edge / texture component value image, and means for setting the edge / texture component value image as the local complexity of the image.

According to the present invention (claim 162), the local complexity generating means adds 1 to the difference counter when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value. Means for setting the value of the difference counter at the time of completion of the processing for the pixel as an edge / texture component value image; subdividing the edge / texture component image into blocks; taking the sum of the edge / texture component values of the block; It includes a means for setting the texture degree index of the block, and a means for setting an up-sample of the texture degree index as the local complexity of the image.

According to the present invention (claim 163), the local complexity generating means includes means for generating an edge / texture component value image from an input image, and means for generating a texture index from the edge / texture component value image. And means for mapping the texture degree index using a predetermined function to obtain the local complexity of the image.
Electronic watermark embedding device according to 0.

According to the present invention (claim 164), the local complexity generating means adds 1 to the difference counter when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value. Means for setting the value of the difference counter at the time of completion of the processing for the pixel as an edge / texture component value image; subdividing the edge / texture component image into blocks; taking the sum of the edge / texture component values of the block; Means for setting a texture degree index of a block, and means for multiplying an edge / texture component image of each block by the texture degree index to obtain a texture component value image and setting the local complexity of the image.

The present invention (claim 165) is based on claim 155.
Electronic watermark embedding device is configured by an integrated circuit. According to the present invention (claim 166), the digital watermark embedding device of claim 160 is constituted by an integrated circuit. The present invention (Claim 16
7) is a storage medium storing a digital watermark embedding program for embedding a digital watermark in a digital image so that the digital watermark is not perceived by human perception, and determines a local complexity of the image from an input image to be input. The required local complexity creation process, the watermark pattern creation process for creating a pattern representing the digital watermark, and the input image, adding the pattern created in the watermark pattern creation process and the local complexity of the image to obtain information And an image adding process for generating an embedded image.

According to the present invention (claim 168), in the local complexity generation process, when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value, the difference counter is incremented by one, and The process includes a process of setting the value of the difference counter at the time when the process for the pixel is completed as an edge / texture component value image, and a process of setting the edge / texture component value image as a local complexity of the image.

According to the present invention (claim 169), in the local complexity creating process, when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value, the difference counter is incremented by one, and A process of setting the value of the difference counter at the time when the processing for the pixel is completed to an edge / texture component value image, subdividing the edge / texture component image into blocks, taking the sum of the edge / texture component values of the block, The process includes a process of setting a texture degree index of a block and a process of setting an upsampled texture degree index to a local complexity of an image.

The present invention (Claim 170) provides a process for generating an edge / texture component value image from an input image and a process for generating a texture index from the edge / texture component value image in the local complexity generation process. , The texture index is mapped using a predetermined function to obtain the local complexity of the image.

According to the present invention (claim 171), in the local complexity generating process, when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value, the difference counter is incremented by one, and A process of setting the value of the difference counter at the time when the processing for the pixel is completed to an edge / texture component value image, subdividing the edge / texture component image into blocks, taking the sum of the edge / texture component values of the block, The process includes a process of setting a texture degree index of a block and a process of multiplying the edge / texture component image of each block by the texture degree index to obtain a texture component value image and setting the local complexity of the image.

The present invention (claim 172) is a storage medium storing a digital watermark embedding program for embedding a digital watermark in a digital image so that the digital watermark is not perceived by human perception. , And a block dividing process of subdividing the input image into blocks of a predetermined size, and a basic watermark that creates a basic watermark pattern representing the digital watermark from key information used for embedding and a digital watermark to be embedded in the input image. A pattern generation process, an adaptive watermark pattern creation process for creating an adaptive pattern that emphasizes the corresponding orthogonal transform coefficient of the basic watermark pattern according to the amplitude value of the orthogonal transform coefficient of the block divided in the block division process, and Local complexity creation process to find local complexity of image An adaptive watermark pattern and the basic watermark pattern, according to the intensity parameter indicative of the local complexity and embedding strength of the image, and an addition process that adds the Buroku, generates information-embedded-image.

According to the present invention (claim 173), in the local complexity generation process, when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value, the difference counter is incremented by one, and The process includes a process of setting the value of the difference counter at the time when the process for the pixel is completed as an edge / texture component value image, and a process of setting the edge / texture component value image as a local complexity of the image.

According to the present invention (claim 174), in the local complexity generation process, when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value, the difference counter is incremented by one, and A process of setting the value of the difference counter at the time when the processing for the pixel is completed to an edge / texture component value image, subdividing the edge / texture component image into blocks, taking the sum of the edge / texture component values of the block, The process includes a process of setting a texture degree index of a block and a process of setting an up-sample of the texture degree index as a local complexity of an image.

The present invention (claim 175) provides a process for generating an edge / texture component value image from an input image and a process for generating a texture index from the edge / texture component value image in the local complexity generation process. , The texture index is mapped using a predetermined function to obtain the local complexity of the image.

According to the present invention (claim 176), in the local complexity generation process, when the difference between each pixel of the input image and the adjacent pixel is larger than a predetermined threshold value, the difference counter is incremented by one, and A process of setting the value of the difference counter at the time when the processing for the pixel is completed to an edge / texture component value image, subdividing the edge / texture component image into blocks, taking the sum of the edge / texture component values of the block, The process includes a process of setting a texture degree index of a block and a process of multiplying the edge / texture component image of each block by the texture degree index to obtain a texture component value image and setting the local complexity of the image.

Thus, according to the present invention, the real component and the imaginary component of each coefficient value of the watermark coefficient matrix, which is a complex matrix, are independently changed from the key information and the digital watermark, and the changed watermark coefficient matrix is discrete. By performing the inverse Fourier transform, a watermark pattern is generated, and an embedded image can be obtained by adding the watermark pattern to the original image in a tile shape.

According to the present invention, it is possible to create local complexity of an image having a relatively large value in a texture region and a relatively small value in a flat region or an edge portion. Become. As a result, information embedding that is more adaptive to the visual characteristics than the conventional technology can be realized,
The watermark resistance is relatively increased. Further, according to the present invention, at the time of embedding, since a pattern reflecting the pattern of the original image is added, it is possible to further improve the image quality, that is, further improve the relative tolerance.

Further, according to the present invention, an edge / texture component value image obtained by treating edges and textures equally, a texture degree index obtained from a local density of the edge / texture component value image, and the former 2 By generating a texture component value image obtained from a single
Alternatively, a texture component value image having a relatively small value can be created in an edge portion, information embedding can be expressed more adaptively to visual characteristics than in the conventional technology, and the resistance to digital watermarking relatively increases.

Further, according to the present invention, upon detection, a block is cut out from an arbitrary position in the image to be detected, a discrete Fourier transform is performed on the block to obtain a coefficient matrix, and then component values specified by key information are obtained. A sequence is created, and a search is made for the amount of translation by correlating a sequence obtained by multiplying each term of the component value sequence by the phase difference of the coefficient according to the amount of translation with a sequence meaning the start point of the watermark pattern, As a result of the search,
Assume that the translation amount when the maximum response value is output is the translation amount from the start point of the watermark pattern to the start point of the cut-out block, and using this translation amount,
Further, it becomes possible to detect the embedded digital watermark.

Further, according to the present invention, it is not necessary to perform a discrete Fourier transform for each scan when searching for the amount of parallel movement, and it is sufficient to perform a discrete Fourier transform only once before the search. It is possible to reduce the detection processing time as compared with the moving amount search (a method of performing orthogonal transformation and scanning while shifting the coordinates).

[0148]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The embodiments of the present invention will be described below with reference to the accompanying drawings. [First Embodiment] In this embodiment, a first example of embedding watermark information will be described.

FIG. 5 shows the configuration of the watermark embedding device according to the first embodiment of the present invention. Digital watermark embedding device 1 shown in FIG.
00 denotes a watermark pattern adding unit 110, an embedding sequence generating unit 120, an embedding target component specifying unit 130, and a component value changing unit 14.
0, comprising a discrete inverse Fourier transform unit 150. The digital watermark embedding device 100 receives an input image 101, an intensity parameter 102, a digital watermark 103, and key information 104 as input, and finally outputs an embedded image 105.

Next, an outline of the operation of the above configuration will be described. FIG. 6 is a flowchart showing the operation of the watermark embedding device according to the first embodiment of the present invention. Step 110) The information embedding device 10 receiving the input
In 0, an input image 101, an intensity parameter 102, a digital watermark 103, and key information 104 are input. First, the embedding target component specifying unit 130 generates the embedding target component position information 131 using the key information 104, and sends this to the component value changing unit 140.

Step 120) The embedding sequence generation unit 120 spreads the position marker information and the digital watermark 103 to the embedding sequence 121 using the key information 104, and sends this to the component value changing unit 140. Step 130) The component value changing section 140 obtains the embedding target component position information 13 acquired from the embedding target component specifying section 130.
Using 1 and the embedding sequence 121, the initial value is set to a zero matrix.

Step 140) Component value changing section 140
Changes the component values of the watermark coefficient matrix 141, which is a complex matrix of zero matrices. Step 150) The discrete inverse Fourier transform unit 150 creates a watermark pattern 151 by performing a discrete inverse Fourier transform on the watermark coefficient matrix 141. Step 160) The watermark pattern adding unit 110 emphasizes the watermark pattern 151 with the intensity parameter 102, outputs an embedded image 105 obtained by adding the watermark pattern 151 to the input image 101 in a tile shape, and ends the information embedding process. I do.

First, the embedding target component designating section 1 in FIG.
30 will be described. FIG. 7 shows a configuration of the embedding target component designation section according to the first embodiment of the present invention. The embedding target component designation unit 130 shown in FIG. 1 receives the key information 104 as input and generates the embedding target component position information 131.
32. FIG. 7 is a flowchart illustrating the operation of the embedding target component specifying unit according to the first embodiment of this invention.

Step 111) Component 1 to be embedded
30 receives the key information 104 and generates a random number sequence using the random number generator 132 with the key information 104 as an initial value. Step 112) The embedding target component designating unit 130 further calculates the length n of an embedding sequence described later based on the generated random number sequence.
The embedding target component position information 131, which is a sequence having the same length as, is generated.

The embedding target component position information 131 has the following configuration. L _k = (x _k , y _k , z _k ) (0 ≦ k <n) x _k : the coefficient to be embedded in the x direction y _k : the coefficient to be embedded in the y direction z _k : the virtual reality of the coefficient to be embedded The flag to indicate (whether a real component or an imaginary component)
In each of the 31 terms, the selection is made such that there is no overlap (all of the above three components coincide with each other for the two terms L _k and L _m ). Also, in order to add noise to the digital watermark and achieve irreversible compression resistance, the band is limited so that coefficients are not selected from the high frequency band (for example,

(Equation 1)

(R is a filtering band) depending on the constraint conditions as described above. Furthermore, the symmetry of the coefficient matrix required for the signal obtained by the discrete inverse Fourier transform to be a real value F (u, v) = F ^* (Nu, Nv) (where F (u, v ) Is a sample value f (x, y) (0 ≦ x
(N, 0 ≦ y <N) Fourier transform) (N is the vertical and horizontal sizes of the watermark coefficient matrix described later) (* indicates the complex conjugate)
It is necessary to consider that only the positive order is selected for the direction order or the y direction. The embedding target component position information 131 is sent to the component value changing unit 140.

Next, the configuration of the embedded sequence generating section 120 in the present embodiment will be described. FIG. 9 shows the configuration of the embedded sequence generation unit according to the first embodiment of the present invention. The embedded sequence generation unit 120 has a spreading sequence generator 122,
The digital watermark 103 and the key information 104 are input. Basically, it is based on the direct spread system in the spread spectrum communication system. Specifically, in the process shown in FIG.
An embedded sequence 121, which is a real number sequence of length n, is generated.

FIG. 10 is a flowchart showing the operation of the embedded sequence generation unit according to the first embodiment of the present invention. Step 121) Diffusion of position marker: The spreading sequence generator 122 can generate K types of sequences of length n, and sets each spreading sequence to r ^(k) and (0 ≦ k <K−1). I will represent it. As the spreading sequence, for example, an M-sequence or the like is used.

Spread sequence generator 122 uses key information 104 as an initial value, and generates a spread sequence {r _i ⁽⁰⁾ } of the ^0th length n.
(0 ≦ i <n) is generated, multiplied by a predetermined parameter weight α (≧ 1), and a real-valued buffer {m _i } of length n (0 ≦ i <n) To be stored. Expressing this by the _{formula, m i: = α · r} i (0) (0 ≦ i <n) (: = indicates that substitutes the value of the right side to the left side) and a.

Step 122) Conversion of Digital Watermark into Symbol Expression: The embedded sequence generation unit 120 converts the digital watermark 103 into a symbol expression. For example, assuming that the digital watermark is information composed of four 8-bit characters,
When each bit is converted into a symbol, it can be converted into four symbols (the value of each symbol is 0 to 255, corresponding to the ASCII code of a character). At this time, the number of symbols is K
It is smaller and the maximum possible value of each symbol is chosen to be smaller than n. Let the value of each symbol be s _j (1 ≦ j ≦ J
(<K)), the maximum value of the possible values of s _j is represented by M (≦ n).

Step 123) Diffusion of digital watermark:
A description will be given of a j-th process in which embedded sequence generation section 120 adds a spreading sequence corresponding to each symbol s _j to {m _i }. Using a value obtained by adding the symbol value s _j to the key information 104 as an initial value, the spreading sequence generator 122 sets the j-th length n
To generate a spreading sequence {r _i ^(j) } (0 ≦ i <n). Next, embedded sequence generation section 120 adds each item of {r _i ^(j) } to each item of {m _i }. Expressed in _{equation, m i: = m i +} r i (j) (0 ≦ i <n) This is repeated for 1 ≦ j ≦ J.

Step 124) Embedding sequence generator 120
Outputs the result of subtracting the average value from each term of {m _i } as an embedded sequence 121 so that the average value of {m _i } obtained up to the above step 123 becomes 0, and changes the component value. Send to section 140. Expressing the subtraction process by an expression,

(Equation 2)

[0163] _{m i: = m i - ave} (0 ≦ i <n)
And Next, the operation of the component value changing unit 140 according to the present embodiment will be described. The component value changing unit 140 receives the embedding target component position information 131 and the embedding sequence 121 as inputs, and
A watermark coefficient matrix is generated according to the procedure shown in FIG. FIG.
5 is a flowchart of the operation of the component value changing unit according to the first embodiment of the present invention.

Step 131) The component value changing section 140 prepares a watermark coefficient matrix F (u, v), which is a complex matrix of N × N size, and sets the initial value to a zero matrix. Step 141) The component value changing unit 140 determines that the k-th embedding target component position information 131 is L _k = (x _k , y
_k , z _k ) to modify the elements of F (u, v).

Step 142) When z _k is a value representing a real component: The component value changing unit 140 adds _mk to the real component of F (x _k , y _k ). Moreover, to maintain the symmetry of Fourier transform _{coefficients, F (N-x k,} N-y k) to the real component of the addition of m _k. Expressed by the _{formula, F (x k, y k} ): = F (x k, y k) + m k F (N-x k, N-y k): = F (N-x k, N-
y _k ) + m _k .

[0166] Step 143) When z _k is a value that represents the imaginary component: component value changing unit 140, F (x _k, the imaginary component of y _k) is added m _k. Further, m _k is subtracted from the imaginary component of F (N−x _k , N−y _k ) so as to maintain the symmetry of the Fourier transform coefficients. When expressed by an equation, F (x _k , y _k ): = F (x _k , y _k ) + m _k · i F (N−x _k , N−y _k ): = F (N−x _k , N−
y _k ) + m _k · i (i is an imaginary unit) This is sequentially performed for k = 0... n−1.

Next, the operation of the discrete inverse Fourier transform unit 150 in this embodiment will be described. FIG. 12 is a diagram for explaining the operation of the discrete inverse Fourier transform unit according to the first embodiment of the present invention. The discrete inverse Fourier transform unit 150 performs a discrete inverse Fourier transform on the watermark coefficient matrix 141 to obtain N × N
A watermark pattern 151 which is a real matrix of a size (an imaginary component is discarded) is obtained and sent to the watermark pattern adding unit 110.

Next, the operation of the watermark pattern adding unit 110 in this embodiment will be described. FIG. 13 is a diagram for explaining the operation of the watermark pattern adding unit according to the first embodiment of this invention. The watermark pattern adding unit 110 emphasizes the watermark pattern 151 with the value of the intensity parameter 102, adds it to the input image 101 in a tile shape, and generates and outputs the embedded image 105. Expression: {I _xy }: input image (0 ≦ x <Width, 0 ≦ y <Height) (Width is the horizontal size of the input image, Height is the vertical size of the input image) {W _ij }: watermark pattern ( 0 ≦ i, j <N) power: intensity parameter, I ′ _xy : = I _xy + power W _{X% N, y% N} (0 ≦ x <Width, 0 ≦ y <Height) (a% b is , A represents the remainder when a is divided by b). The watermark pattern adding unit 110
{I ′ _xy } is output as the embedded image 105. When the value of the intensity parameter 102 is 1, the watermark pattern is added without emphasis. That is, this is equivalent to a case where no intensity parameter is used.

As described above, the operation of the digital watermark embedding device of the present embodiment is completed. In the digital watermark embedding process using this embodiment, the N × N two-dimensional discrete inverse Fourier transform is performed only once. In the conventional technology, N ×
It is necessary to perform the discrete Fourier transform and the discrete inverse Fourier transform having the size of N by the number of blocks when the image is divided into blocks of N × N pixels. For example, when the number of blocks is b, the ratio of the embedding processing amount between the first embodiment and the conventional technology is 1: 2b, and the embedding processing can be performed at high speed by using the first embodiment.

[Second Embodiment] Next, in this embodiment, a process for detecting a digital watermark embedded in the digital watermark embedding apparatus shown in the first embodiment from a detection target image will be described. I do. FIG. 14 shows the configuration of a digital watermark detection device according to the second embodiment of the present invention.

The digital watermark detection device 200 is the same as the first digital watermark detection device described above.
This is an apparatus for detecting an embedded digital watermark from an image in which a process such as irreversible compression has been partially cut out from an embedded image obtained using the method of the embodiment. The digital watermark detection device 200 includes a detection target component designating unit 21
0, position marker sequence generation unit 220, position marker detection unit 2
30 and a digital watermark detection unit 240.

FIG. 15 is a flowchart showing the operation of the digital watermark detection apparatus according to the second embodiment of the present invention. Step 210) The digital watermark detection device 200 receives the detection target image 201 and the key information 202 used at the time of embedding. First, the detection target component specifying unit 210 generates the detection target component position information 211 from the key information in the same manner as the embedding target component specifying unit 130 in the above-described first embodiment.
30 and the digital watermark detection unit 240.

Step 220) In the position marker sequence generation section 220, the position marker sequence 22
1 is sent to the position marker detector 230. Step 230) The position marker detection unit 230 cuts out a pixel block of size N × N (N is the size of the watermark pattern in the first embodiment) from the detection target image 201, and the starting point of the cut-out pixel block. The start point of the pixel block, which coincides with the start point of the watermark pattern at the time of embedding (the upper left position of the pixel block), is output as offset information 231, and the digital watermark detection unit 2
Send to 40.

Step 240) Digital watermark detecting section 24
0 indicates that a pixel block of N × N size is cut out from an offset position specified by the detection target image 201, key information 202, detection target component position information 211, and offset information 231, and a digital watermark embedded in this pixel block is extracted. It detects and outputs the detection result 203. Next, the detection target component designation unit 210 according to the present embodiment will be described.

FIG. 16 shows the configuration of the detection target component designating section according to the second embodiment of the present invention, and FIG. 17 is a flowchart showing the operation of the detection target component designating section according to the second embodiment of the present invention. is there. Step 211) The detection target component designation unit 210 has a random number generator 212, receives the key information 202, and generates a random number sequence using the key information 202 as an initial value using the random number generator 212.

Step 212) On the basis of the generated random number sequence, the length n of the embedded sequence used in the first embodiment is used.
The detection target component position information 211 which is a sequence having the same length as the above is generated. The detection target component position information 211 has the following configuration. L _k = (x _k , y _k , z _k ) (0 ≦ k <n) x _k : coefficient to be detected in the x direction y _k : coefficient to be detected in the y direction z _k : flag representing the imaginary reality of the coefficient to be detected ( Real component or imaginary component) The detection target component designation unit 210 operates in exactly the same way as the embedding target component designation unit 130 in the first embodiment described above. That is, using the same key information as input, the embedding target component position information 131 generated by the embedding target component specifying unit 130 of the first embodiment and the detection target component generated by the detection target component specifying unit 210 of the present embodiment The position information 211 is exactly the same. Generated detection target component position information 2
11 is sent to the position marker detection unit 230.

Next, a description will be given of the position marker sequence generation unit 220 in this embodiment. FIG. 18 shows a second embodiment of the present invention.
FIG. 19 shows the configuration of the position marker sequence generation unit of the embodiment of FIG.
9 is a flowchart illustrating an operation of the position marker sequence generation unit according to the second example of the present invention. Step 221) The position marker sequence generation unit 220 has a spread sequence generator 222, and receives the key information 202 as an input and uses the spread sequence generator 222 to generate the key information 202.
Is the 0-th spreading sequence of length n {r _i ⁽⁰⁾ }
(0 ≦ i <n) is generated, and this spread sequence is directly used as a position marker sequence 221 ({p _i } (0 ≦ i <n) (p
_i : = r _i ⁽⁰⁾ (0 ≦ i <n)). This is sent to the position marker detection unit 230. Position marker sequence generator 22
As the spreading sequence generator 222 in 0, one that performs the same operation as the spreading sequence generator 122 in the embedded sequence generator 120 in the first embodiment described above is used.

Next, the position marker detecting section 230 in this embodiment will be described. FIG. 20 shows the configuration of the position marker detection unit according to the second embodiment of the present invention. The position marker detection unit 230 shown in FIG.
2. Block generator 233, discrete Fourier transformer 23
4, a detection target sequence generation unit 235, a position marker detection information generation unit 236, and an offset information generation unit 237.

FIG. 21 is a flowchart showing the operation of the position marker detecting section according to the second embodiment of the present invention. Step 231) The position marker detection unit 230 receives the detection target image 201, the detection target component position information 211, and the position marker sequence 221 as input, and first, in the offset candidate information generation unit 232, the offset candidate information 204 (a,
b) are sequentially generated from (0, 0) to (N-1, N-1).

Step 232) Hereinafter, the processes from the block generation unit 233 to the position marker detection information generation unit 236 are performed for each offset candidate information 204. FIG.
FIG. 23 is a diagram for explaining the operation of the block generation unit according to the second embodiment of the present invention, and FIG. 23 is a flowchart of the operation of the block generation unit according to the second embodiment of the present invention. Hereinafter, the operation of the block generation unit 233 will be described with reference to FIGS.

Step 232-1) Block generator 2
33 is a detection target image 201 and offset candidate information 20
Enter 4. Step 232-2) A pixel block of N × N size is cut out from a position offset from the upper left of the detection target image 201 by the offset candidate information 204 (a, b).

Step 232-3) The cut-out pixel block is sent to the discrete Fourier transform unit 234. Step 233) The discrete Fourier transform unit 234 performs a discrete Fourier transform on the pixel block to perform the detection target coefficient matrix 20
6 is sent to the detection target sequence generation unit 235.

Step 234) Detection target sequence generation section 2
At 35, the detection target sequence 207 is obtained from the detection target component position information 211 and the detection target coefficient matrix 206 as follows.
Get. FIG. 24 is a flowchart illustrating the operation of the detection target sequence generation unit according to the second embodiment of this invention. Step 234-1) The detection target sequence generation unit 235
The detection target component position information 211 is _represented by L _k (x _k , y _k ,
z _k ) (0 ≦ k <n), and the detection target coefficient matrix 206 is represented by F
If (u, v) (0 ≦ u <N, 0 ≦ v <N) and the detection target sequence 207 is represented by {q _k } (0 ≦ k <n), the processing is performed by the following branch.

Step 234-2) When z _k is a value representing a real component: q _k : = (real component value of F (x, y)). Step 234-3) When z _k is a value representing an imaginary component: q _k : = (imaginary component value of F (x, y)). The processing of the above steps 234-2 and 234-3 is sequentially performed for k = 0... N-1.

Step 234-4) The detection target sequence 207 {q _k } obtained above is output. Step 235) Next, the position marker detection information generation unit 2
At 36, the correlation value between the position marker sequence 221 and the detection target sequence 207 is obtained by the following equation, and this is combined with the offset candidate information 204 to obtain the position marker detection information 208.
Output as corr _ab . The position marker detection information 2
08 is obtained by the following equation.

(Equation 3)

The processing from the block generation unit 233 to the position marker detection information generation unit 236 is performed by setting the offset candidate information (a, b) 204 to (0, 0) to (N-1, N-1).
The position marker detection information 208 is sequentially transmitted to the offset information generation unit 237. Step 236) The offset information generation unit 237 outputs all the input position marker sequences 221 and the detection target sequence 2 out of all the input position marker detection information 208.
07 is the maximum when the correlation value of the offset candidate information (a,
b) is output as offset information 231 and sent to the digital watermark detection unit 240.

The offset information 231 output from the position marker detecting section 230 indicates how far the starting point of the watermark pattern at the time of embedding is shifted from the upper left of the detection target image 201. Next, the digital watermark detection unit 240 according to the present embodiment will be described. FIG. 25 shows the configuration of the digital watermark detection unit according to the second embodiment of the present invention.
The digital watermark detection unit 240 shown in the figure includes a block generation unit 241, a discrete Fourier transform unit 242, a detection target sequence generation unit 243, a symbol detection unit 244, and a detection result generation unit 24.
5 is comprised.

FIG. 26 is a flowchart showing the operation of the digital watermark detection unit according to the second embodiment of the present invention. Step 241) The digital watermark detection unit 240 receives the detection target image 201, the detection target component position information 211, the offset information 231 and the key information 202, and
The block generation unit 241 cuts out an N × N size pixel block from a position offset based on the offset information (a, b) from the upper left of the detection target image, and sends it to the discrete Fourier transform unit 242.

Step 242) Discrete Fourier Transform Unit 2
42 generates a detection target coefficient matrix 247 by performing a discrete Fourier transform on the pixel block 205, and sends this to the detection target sequence generation unit 243. Step 243) The detection target sequence generation unit 243 obtains the detection target sequence 248 by the same processing as the detection target sequence generation unit 235 in the position marker detection unit 230, and sends it to the symbol detection unit 244.

Step 244) Symbol detector 244
Calculates a correlation value between a sequence generated from the key information 202 and each piece of symbol candidate information and a detection target sequence 248, and generates symbol detection information 249. Step 245) After obtaining all the symbol detection information 249, the detection result generation unit 245 sets the symbol detection information 2
For each of the 49 symbol positions, the symbol candidate value when the correlation value is the maximum is set as a detected symbol.

Step 246) After determining detection symbols from all symbol positions, the detection result generation section 245
Outputs the result of inversely converting the detected symbol to the digital watermark format as a detection result 203. Next, the symbol detection unit 244 of the digital watermark detection unit 240 will be described. FIG. 27 shows the configuration of the symbol detector of the second embodiment of the present invention.

The symbol detecting section 244 shown in the figure includes a symbol candidate generating section 2441 and a symbol sequence generating section 244.
2. It is composed of a symbol detection information generating section 2443, which receives the key information 202 and the detection target sequence 248 as input, and sets M symbol candidates for each of J symbol positions to be detected in advance. The symbol detection information 249 is generated and output to the detection result generation unit 245. Hereinafter, detection of the j-th (1 ≦ j ≦ J) symbol will be described.

FIG. 28 is a flowchart showing the operation of the symbol detecting section according to the second embodiment of the present invention. Step 244-1) The symbol candidate generation unit 2441 of the symbol detection unit 244 sets the symbol candidate c from 0 to M-
1 sequentially, and for each symbol candidate, step 2
The processing from 44-2 to step 244-3 is performed. M
Represents the maximum value of the symbol value by the symbol conversion in the first embodiment described above.

Step 244-2) The processing performed by the symbol sequence generation section 2442 will be described. FIG.
Shows the configuration of the symbol sequence generator of the second embodiment of the present invention, and FIG. 30 is a flowchart showing the operation of the symbol sequence generator of the second embodiment of the present invention. Step 244-2-1) Symbol Sequence Generation Unit 244
2, a spread sequence generator 2447 inputs a key information 202 and a symbol candidate c2444, and sets a value obtained by adding c and the key information 202 as an initial value.
^{_{i (j)} (0 ≦}} i <n) to generate, which used directly as the symbol sequence ^{_{2445 {p i (j)}}} (0 ≦ i <n) (p i j): = r i (j) (0 ≦ i <n)). The symbol sequence is sent to symbol detection information generation section 2443.

Step 244-3) The symbol detection information generation section 2443 receives the detection target sequence 248, the symbol sequence 2445, the symbol candidate c2444, and the symbol position j currently being processed and inputs the symbol sequence and the detection target sequence by the following formula. 248 correlation values are obtained, and the correlation value, the symbol candidate, and the symbol position are grouped to form symbol detection information
rr _c ^(j) 249 is generated.

(Equation 4)

The symbol detection information 249 is sent to the detection result generator 245. Next, the detection result generation unit 245 according to the present embodiment will be described. FIG. 31 shows the configuration of the detection result generator according to the second embodiment of this invention. The detection result generation unit 245 shown in FIG.
1, a detection symbol generation unit 2452 and an inverse symbol conversion unit 2453.

FIG. 32 is a flow chart showing the operation of the detection result generator according to the second embodiment of the present invention. Step 245-1) The detection result generation section 245 receives the symbol detection information 249 as an input, and first, the symbol generation information for each symbol position
9 is divided for each symbol position j, and the sequence corr _c ^(j) , (0 ≦ c <M) 2454 for each symbol position of length M is j
= 1 to J are generated and sent to the detected symbol generator 2452.

Step 245-2) The detected symbol generation section 2452 receives the sequence 2454 for each symbol position as input, and outputs the sequence 245 for each symbol position for each symbol position j.
4, a symbol candidate c having the largest correlation value is found, and a detected symbol s _j (1 ≦ j <J) 2455 is generated. Step 245-3) After obtaining all s _j , the inverse symbol conversion unit 2453 converts the symbol expression to the original digital watermark expression (for example, the inverse conversion corresponding to the conversion in the first embodiment is 4). This is a process of converting the values of one detection symbol (the value of each detection symbol is 0 to 255) into ASCII codes, each of which is regarded as an ASCII code, and generates and outputs a detection result 203. . The detection result 203 represents a digital watermark embedded in the detection target image 201.

With the above processing, the processing of the digital watermark detection device in the second embodiment ends. In the digital watermark detection processing according to the second embodiment, a block to be detected is shifted from a detection target image while being shifted, a discrete Fourier transform is performed to obtain a detection target sequence, and a correlation between the detection target sequence and the position marker sequence is obtained. By finding the peak, it is found how much the starting point of the watermark pattern at the time of embedding is offset from the upper left of the detection target image. This makes it possible to correctly find the position of the watermark pattern even when the detection target image is a partial image (size is 2N × 2N or more) obtained by cutting out the embedded image in the first embodiment from an arbitrary position. When the offset amount is found, the digital watermark is detected from the pixel block starting from that position, so that the embedded digital watermark can be correctly detected.

In the digital watermark detection processing according to the second embodiment, the original image (the input image according to the first embodiment) is unnecessary. As described above, the digital watermark detection according to the second embodiment can detect a digital watermark from an image obtained by cutting out an embedded image from an arbitrary position without an original image, which is impossible with the related art. It becomes possible. [Third Embodiment] Next, as a third embodiment of the present invention,
In the above-described second embodiment, an example will be described in which a clipped pixel block is cyclically shifted based on offset information.

In the following description of this embodiment, parts other than those described below are the same as those of the above-described second embodiment. First, the position marker detection unit 23 in the present embodiment
The operation of the block generation unit 233 during 0 will be described. In addition,
The configuration is the same as the configuration shown in FIG. FIG. 33 is a flowchart illustrating the operation of the block generation unit according to the third embodiment of this invention.

Step 301) Position marker detector 23
The block generation unit 233 of 0 inputs the detection target image 201 and the offset candidate information 204. Step 302) First, as illustrated in FIG. 34, the block generation unit 233 performs N × N ×
Cut out N size blocks. Step 303) Next, the block generation unit 233 performs a cyclic shift on the block based on the offset candidate information 204 (moving a portion that protrudes from the block by parallel movement to the opposite side) to perform pixel shift. Obtain block 205.

Step 304) The obtained pixel block 205 is output. When this pixel block is represented by an equation, {B _ij }: a block obtained by cutting out an N × N size from the upper left of the detection target image 201, a pixel obtained by a cyclic shift based on offset candidate information (a, b) The block is: B ′ _ij = B _{(i + a)% N, (j + b)% N} (0 ≦ i <N, 0 ≦ j <N, x% y represents the remainder of x divided by y ) And outputs {B ′ _ij } as a pixel block.

Next, the operation of the block generation unit 241 in the digital watermark detection unit 240 according to the third embodiment will be described below. Digital watermark detection unit 24 in the present embodiment
0, the block generation unit 241 receives the detection target image 201 and the offset information 231 as inputs, and outputs the position marker detection unit 2
Similarly to the block generation unit 233 of FIG. 30 (inputting offset information as offset candidate information), a pixel block is output using cyclic shift.

In the case of the digital watermark detection processing using this embodiment, the size required for the image to be detected is increased from 2N × 2N or more to N × N or more, and is smaller than that of the second embodiment. Detection from the detection target image becomes possible. [Fourth Embodiment] Next, in a fourth embodiment of the present invention, when detecting a digital watermark, the size of a pixel block in a case where the detection target image is equal to or larger than N × N size or smaller than N × N size. An example of clipping will be described.

First, the block generator 233 of the position marker detector 230 will be described. Except for the parts described below, the configuration is the same as that of the above-described second and third embodiments. FIG. 35 is a diagram for explaining the operation of the block generation unit according to the fourth embodiment of the present invention, and FIG. 36 is an operation of the block generation unit of the position marker detection unit according to the fourth embodiment of the present invention. It is a flowchart which shows.

Step 401) The detection target image 201 and the offset candidate information 204 are input to the block generation unit 233 of the position marker detection unit 230 in this embodiment. Step 402) The block generation unit 233 determines whether the detection target image 201 is equal to or larger than the N × N size. If so, the process proceeds to step 403; otherwise, the process proceeds to step 404.

Step 403) Block generator 233
Cuts out an N × N size block from the upper left of the detection target image. Step 404) The detection target image 201 is small and N ×
If the image cannot be cut out to the N size, the block generation unit 233
Cuts out only the portion corresponding to the detection target image (FIG. 35A) included when trying to cut out N × N, and calculates the average value of the pixel values corresponding to the detection target image for the portion that is less than the N × N size. And fill it with this.

Step 405) Block generator 233
Obtains a pixel block by cyclically shifting the block obtained by the above processing in the same manner as in the third embodiment. Step 406) The block generator 233 outputs the obtained pixel block 205. Hereinafter, the operation of the block generation unit 241 in the digital watermark detection unit 240 will be described.

The block generation unit 241 of the digital watermark detection unit 240 in the third embodiment described above
01 and the offset information 231 are input, and a pixel block 246 is obtained and output by the same processing as that of the block generation unit 233 of the position marker detection unit 230 (inputting the offset information as offset candidate information). In the case of the digital watermark detection processing using the present embodiment, the size required for the detection target image becomes an arbitrary size from N × N or more as compared with the third embodiment, and the detection from the smaller detection target image can be performed. It becomes possible.

[Fifth Embodiment] Next, a fifth embodiment of the present invention will be described. Except for the parts described below, the configuration is the same as that of the second embodiment. In this embodiment,
The digital watermark detection unit will be described. FIG. 37 shows the configuration of the digital watermark detection unit according to the fifth embodiment of the present invention.

[0212] The digital watermark detection section 500 shown in the figure comprises a block division section 510, a discrete Fourier transform section 520, a detection target sequence generation section 530, a symbol detection section 540, and a detection result generation section 550. FIG. 38 is a flowchart showing the operation of the digital watermark detection unit according to the fifth embodiment of the present invention.

Step 510) Digital watermark detecting section 50
0 is the detection target image 201 and the detection target component position information 21
39, the offset information 231 and the key information 202 are input. First, the block dividing unit 510 shifts the detection target image 201 from the position offset by the offset information (a, b) 231 from the upper left as shown in FIG. ×
The pixel block is divided into T pixel blocks of N size, and numbers 0 to T−1 are assigned to the pixel blocks 511 obtained by the division.

The t-th (0 ≦ t <T) -th pixel block 51
The following processing is performed for No. 1. Step 520) The discrete Fourier transform unit 520 performs a discrete Fourier transform on the t-th pixel block 511 to obtain a detection target coefficient matrix 521. Step 530) The detection target sequence generation unit 530 generates a detection target sequence 531 from the detection target component position information 211 and the detection target coefficient matrix 521.

Step 540) Symbol detection section 540
Generates the symbol detection information 546 by calculating the correlation value between the sequence generated from the key information 202 and each symbol candidate information and the detection target sequence 531. Step 550) After obtaining the symbol detection information 546 for all t, the detection result generation unit 550 sets, for each symbol position of the symbol detection information 546, the symbol candidate value having the maximum correlation value as the detection symbol.

Step 560) Detection result generator 550
Determines the detected symbols from all symbol positions, and outputs the detected symbols as a detection result 203 obtained by inversely converting the detected symbols into a digital watermark format. Next, the symbol detection unit 540 in the present embodiment will be described. FIG. 40 shows the configuration of the symbol detector of the fifth embodiment of the present invention. Symbol detection section 540 shown in FIG.
Are the symbol candidate generator 541 and the symbol sequence generator 5
42, a symbol detection information generation unit 543. The symbol detection unit 540 receives the key information 202 and the detection target sequence 531 (obtained from the t-th pixel block) as input, and sets a predetermined number J of detected symbols. The symbol detection information 546 for the M symbol candidates 544 is generated for each symbol position.
Output to 0. Hereinafter, detection of the j-th (1 ≦ j ≦ J) symbol will be described.

FIG. 41 is a flowchart showing the operation of the symbol detecting section according to the fifth embodiment of the present invention. Step 551) The symbol candidate generation unit 541 sequentially generates the symbol candidates c from 0 to M-1, and performs the following steps 552 to 553 for each symbol candidate. M represents the maximum value of the symbol value by the symbol conversion in the first embodiment.

Step 552) Symbol sequence generator 5
42 will be described. FIG. 42 shows the configuration of the symbol sequence generator of the fifth embodiment of the present invention, and FIG. 43 is a flowchart showing the operation of the symbol sequence generator of the fifth embodiment of the present invention. Step 552-1) The symbol sequence generation unit 542 determines
The spread sequence generator 5 receives the key information 202 and the symbol information c as input and sets a value obtained by adding c and the key information 202 as an initial value.
421, the j-th spreading sequence of length n {r _i ^(j) } (0 ≦
i <n), which is directly used as a symbol sequence 545
Used as {p _i ^(j) } (0 ≦ i <n) (p _i ^(j) :
= R _i ^(j) (0 ≦ i <n)). The symbol sequence 545 is sent to the symbol detection information generator 543.

Step 553) The symbol detection information generation section 543 generates the detection target sequence 531 and the symbol sequence 545.
And the symbol candidate c and the symbol position j currently being processed are input, the correlation value between the symbol sequence 545 and the detection target sequence 531 is obtained by the following equation, and the correlation value, the symbol candidate, the symbol position, and the pixel block position are grouped. generating a symbol detection information ^{_{corr c (j) (t)}} 546 Te.

(Equation 5)

[0220] Symbol detection information 546 is sent to detection result generating section 550. Next, the detection result generation unit 550 of the present embodiment will be described. FIG. 44 shows the configuration of the detection result generator according to the fifth embodiment of the present invention. The detection result generation unit 550 shown in FIG.
1, detection symbol generation section 552, inverse symbol conversion section 55
3

FIG. 45 is a flowchart showing the operation of the detection result generator according to the fifth embodiment of the present invention. Step 561) The detection result generation unit 550 receives the symbol detection information 546 as input, and detects a symbol as the detection result 556.
(Digital watermark). First, each symbol position for each sequence generation unit 551 divides the symbol detection information 546 for each symbol position j, the length M × each symbol position for each sequence corr c of ^{_{T (j) (t),}} (0 ≦ c < M, 0 ≦ t <T) 5
54 is generated for each of j = 1 to J and sent to the detected symbol generation unit 552.

Step 562) Detected symbol generator 5
52 receives the symbol position sequence 554 as input, finds the symbol candidate c at which the maximum correlation value is obtained in each symbol position sequence 554 for each symbol position j, and detects the detected symbol s _j (1 ≦ j < J). Step 563) After obtaining all s _j , the inverse symbol conversion unit 553 converts the symbol expression into the original digital watermark expression (for example, four inverse detections corresponding to the conversion in the first embodiment are performed). A symbol (the value of each detected symbol is 0 to 255) is regarded as an ASCII code and is converted into four 8-bit characters, and a detection result 203 is generated and output.

The detection result 203 indicates a digital watermark embedded in the image to be detected. In the digital watermark detection processing according to the present embodiment, the detection target image (size of 2N × 2N or more) 201 is divided into blocks, and the symbol having the maximum correlation value among all the blocks is detected. As compared with the above-described second embodiment (detection is performed only from one block), more accurate detection is possible.

[Sixth Embodiment] Next, a digital watermark detection section will be described as a sixth embodiment of the present invention. Except for the parts described below, the configuration is the same as that of the third embodiment. FIG. 46 shows the configuration of the digital watermark detection unit according to the sixth embodiment of the present invention. Digital watermark detection unit 6 shown in FIG.
00 denotes a block dividing unit 610 and a discrete Fourier transform unit 6
20, detection target sequence generation section 630, symbol detection section 64
0, a detection result generation unit 650.

FIG. 47 is a flowchart showing the operation of the digital watermark detection unit according to the sixth embodiment of the present invention. Step 610) The digital watermark detection unit 600 receives the detection target image 201, the detection target component position information 211, the offset information 231 and the key information 202, and outputs the digital watermark as the detection result 203. First, the block dividing unit 610 converts the detection target image 201 as shown in FIG.
It is divided into N × N size blocks from the upper left, and T blocks obtained by cyclically shifting each block using the offset information (a, b) 231 in the same manner as the block generation unit 233 of the position marker detection unit 230 N × N pixel blocks are generated. Further, the block dividing unit 610
Assigns numbers 0 to T-1 to the pixel blocks obtained by the division. t (0 ≦ t <T) -th pixel block 60
The processing from step 620 to step 640 is repeated for 1.

Step 620) Discrete Fourier Transform Unit 6
20 obtains a detection target coefficient matrix 602 by performing a discrete Fourier transform on the t-th pixel block 601. Step 630) The detection target sequence generation unit 630 obtains the detection target sequence 603 by the same processing as the detection target sequence generation unit 235 of the position marker detection unit 230, and sends the detection target sequence 603 to the symbol detection unit 640.

Step 640) Symbol detection section 640
Generates a symbol detection information 604 by calculating a correlation value between a sequence generated from the key information 202 and each symbol candidate information and a detection target sequence 603. T-th pixel block 6
The process from step 620 is repeated until 01. Step 650) After obtaining the symbol detection information 604 for all the pixel blocks 601 of t, the detection result generation unit 650 detects the symbol candidate value at the time when the correlation value is the maximum for each symbol position of the symbol detection information 604. And

Step 660) Detection result generator 650
Determines detection symbols from all symbol positions, and then outputs the result of inverse conversion of the detection symbols into a digital watermark format as a detection result 203. Next, the symbol detection unit 640 according to the present embodiment will be described. FIG. 49 shows the configuration of the symbol detector of the sixth embodiment of the present invention. Symbol detection section 640 shown in FIG.
Are the symbol candidate generator 641 and the symbol sequence generator 6
43, a symbol detection information generator 645.

[0229] The symbol detector 640 determines the key information 202 and the detection target sequence 6 (obtained from the t-th pixel block).
03 is input, and symbol detection information 646 is output.
Symbol detection information 604 for M symbol candidates 642 is generated for each of the J symbol positions to be detected, which is set in advance, and output to detection result generation section 650. Hereinafter, detection of the j-th (1 ≦ j <J) symbol will be described.

FIG. 50 is a flowchart showing the operation of the symbol detector of the sixth embodiment of the present invention. Step 641) The symbol candidate generation unit 641 sequentially generates the symbol candidates c from 0 to M−1, and performs the processes from step 642 to step 643 for each symbol candidate 642. M represents the maximum value of the symbol value by the symbol conversion in the first embodiment.

Step 642) Symbol sequence generator 6
43 is shown in FIG. 51, and its operation is shown in FIG. Step 642-1) The symbol sequence generation unit 643 calculates
The spread information generator 6431 receives the key information 202 and the symbol candidate c642 as input and sets a value obtained by adding c and the key information 202 as an initial value. The spread sequence generator 6431 has a j-th spread sequence {r _i ^(j) } of length n.
(0 ≦ i <n) is generated and used as it is as a symbol sequence 644 {p _i ^(j) } (0 ≦ i <n) (p _{i
^{(j): = r i (}} j) (0 ≦ i <n)). Symbol sequence 64
4 is sent to the symbol detection information generator 645.

Step 643) The symbol detection information generation section 645 generates the detection target sequence 603 and the symbol sequence 644.
, The symbol candidate c and the symbol position j currently being processed are input, the correlation value between the symbol sequence 644 and the detection target sequence 603 is obtained by the following equation, and the correlation value, the symbol candidate, the symbol position, and the pixel block position are grouped. generating a symbol detection information ^{_{corr c (j) (t)}} 604.

(Equation 6)

[0233] Symbol detection information 604 is sent to detection result generating section 650. Next, the detection result generation unit 650 according to the present embodiment will be described. FIG. 53 shows the configuration of the detection result generator according to the sixth embodiment of the present invention. The detection result generation unit 650 shown in FIG. 11 includes a sequence generation unit 651 for each symbol position, a detection symbol generation unit 653, and an inverse symbol conversion unit 655.

FIG. 54 is a block diagram showing a detection result according to the sixth embodiment of the present invention.
6 is a flowchart illustrating an operation of a fruit generation unit. Step 651) The detection result generation unit 650 sets the symbol
With the detection information 604 as an input, first, the system for each symbol position
In the column generation unit 651, the symbol detection information 604 is
Each symbol position of length M × T, divided for each symbol position j
Every series corr _c ^{(j) (t)}(0 ≦ c <M, 0 ≦ t <T) 6
52 is generated and detected for each j from j = 1 to j
This is sent to the symbol generation unit 653.

Step 652) Detected symbol generator 6
53 receives a symbol position sequence 652 as input, finds a symbol candidate c at the time of taking the maximum correlation value in each symbol position sequence 652 for each symbol position j, and detects a detected symbol s _j (1 ≦ j). <J) 654 is generated. The above processing is repeated until all s _j are obtained. Step 653) After obtaining all the detected symbols s _j , the inverse symbol conversion unit 655 converts the symbol expression to the original digital watermark expression (for example, the inverse conversion corresponding to the conversion in the first embodiment is 4). This is a process of converting the values of one detection symbol (the value of each detection symbol is 0 to 255) into four 8-bit characters by regarding each as an ASCII code, and generates and outputs a detection result 203. The detection result 203 represents a digital watermark embedded in the detection target image.

In the digital watermark detection processing according to this embodiment,
The minimum size of the image to be detected is (N × N), and detection can be performed from a partial image smaller than that of the fifth embodiment.
Since the number of blocks used for detection is larger than in the fifth embodiment, more accurate detection is possible as compared with the fifth embodiment. Further, detection with higher accuracy is possible as compared with the third embodiment (detection is possible only from one block).

[Seventh Embodiment] An electronic watermark detection unit will be described as a seventh embodiment of the present invention. Except for the parts described below, the configuration is the same as that of the above-described fourth embodiment. FIG. 55 shows the configuration of the digital watermark detection unit according to the seventh embodiment of the present invention. Digital watermark detection unit 700 shown in FIG.
Is a block division unit 710 and a discrete Fourier transform unit 72
0, detection target sequence generation section 730, symbol detection section 74
0, a detection result generation unit 750.

FIG. 56 is a flowchart showing the operation of the digital watermark detecting section according to the seventh embodiment of the present invention. Step 710) The digital watermark detection unit 700 receives the detection target image 201, the detection target component position information 211, the offset information 231 and the key information 202 as input, and first, the block division unit 710 converts the detection target image 201 into the one shown in FIG. As shown, the block is divided into N × N size blocks from the upper left,
As in the case of No. 3, T N × N pixel blocks obtained by cyclically shifting each block using the offset information (a, b) are generated. At this time, for a block that cannot be cut out to the N × N size around the end of the detection target image 201, only a portion corresponding to the detection target image included in the case of the N × N cutout is cut out, and the N × N size is sufficient. For the portion that does not exist, the average value of the pixel values corresponding to the detection target image is obtained and filled with the average value, and then the cyclic shift is performed. The block division unit 710 assigns numbers 0 to T−1 to the pixel blocks obtained by the division.
For the t (0 ≦ t <T) -th pixel block 701,
The following steps 720 to 740 are repeated.

Step 720) Discrete Fourier transform unit 7
20 obtains a detection target coefficient matrix 702 by performing a discrete Fourier transform on the t-th pixel block. Step 730) The detection target sequence generation unit 730 obtains the detection target sequence 703 by the same processing as the detection target sequence generation unit 235 of the position marker detection unit 230, and sends it to the symbol detection unit 740.

Step 740) Symbol detector 740
Generates a symbol detection information 704 by calculating a correlation value between a sequence generated from the key information 202 and each symbol candidate information and a detection target sequence 703. Step 750) After obtaining the symbol detection information 704 for all t, the detection result generation unit 750 sets, for each symbol position of the symbol detection information 704, the symbol candidate value with the maximum correlation value as the detected symbol.

Step 760) After determining detection symbols from all symbol positions, the detection result generation section 750
Outputs the result of inversely converting the detected symbol to the digital watermark format as a detection result 203. Next, the symbol detection unit 740 according to the present embodiment will be described. FIG. 58 shows the configuration of the symbol detector of the seventh embodiment of the present invention. The symbol detector 740 shown in FIG.
Are the symbol candidate generator 741 and the symbol sequence generator 7
43, a symbol detection information generator 745.

[0242] The symbol detecting section 740 detects the key information 20.
2 and the detection target sequence 703 (obtained from the t-th pixel block) are input, and the symbol detection information 704 on the M symbol candidates 742 is set for each predetermined symbol position of the number j of symbols to be detected. Is generated and output to the detection result generation unit 750. Hereinafter, j (1 ≦ j ≦
The J-th symbol detection will be described.

FIG. 59 is a flowchart showing the operation of the symbol detector of the seventh embodiment of the present invention. Step 741) The symbol candidate generation unit 741 sequentially generates the symbol candidates c from 0 to M−1, and performs the processing of steps 742 to 743 for each symbol candidate. M represents the maximum value of the symbol value in the first embodiment described above.

Step 742) Symbol sequence generator 7
The configuration and operation of 43 will be described. FIG. 60 shows the seventh embodiment of the present invention.
FIG. 61 shows the configuration of the symbol sequence generation unit of the embodiment of FIG.
15 is a flowchart illustrating an operation of the symbol sequence generation unit according to the seventh example of the present invention. Step 742-1) The symbol sequence generation unit 743 determines
The spread sequence generator 7 receives the key information 202 and the symbol candidate c as input and sets a value obtained by adding c and the key information 202 as an initial value.
At 431, the j-th spreading sequence of length n {r _i ^(j) } (0 ≦
i <n), and this is directly used as a symbol sequence 744
Used as {p _i ^(j) } (0 ≦ i <n) (p _i ^(j) :
= R _i ^(j) (0 ≦ i <n)). The symbol sequence 744 is sent to the symbol detection information generator 745.

Step 743) The symbol detection information generation section 745 generates the detection target sequence 703 and the symbol sequence 744.
, The symbol candidate c and the symbol position j currently being processed are input, the correlation value between the symbol sequence 744 and the detection target sequence 703 is calculated by the following equation, and the correlation value, the symbol candidate, the symbol position, and the pixel block position are grouped. , it generates the symbol detection information ^{_{corr c (j) (t)}} 746.

(Equation 7)

[0246] Symbol detection information 746 is sent to detection result generating section 750. Next, the detection result generation unit 750 according to the present embodiment will be described. FIG. 62 shows the configuration of the detection result generator according to the seventh embodiment of the present invention. The detection result generation unit 750 shown in FIG. 7 includes a sequence generation unit 751 for each symbol position, a detection symbol generation unit 753, and an inverse symbol conversion unit 755.

FIG. 63 shows the detection result of the seventh embodiment of the present invention.
It is a flowchart of operation | movement of a fruit generation part. Step 751) The detection result generation unit 750 sets the symbol
With the detection information 704 as input, first, the system for each symbol position
In the column generation unit 751, the symbol detection information 704 is
Each symbol position of length M × T, divided for each symbol position j
Every series corr _c ^{(j) (t)}, (0 ≦ c <M, 0 ≦ t <T)
752 for each j from j = 1 to J
The output symbol generation unit 753 is sent.

Step 752) Detected symbol generator 7
53 receives a symbol position sequence 752 as input, finds a symbol candidate c at which the maximum correlation value is obtained in each symbol position sequence 752 for each symbol position j, and detects a detected symbol s _j (1 ≦ j ≦ J) Generate 754. Step 7
51 and the process of step 752 are repeated until all s _j are obtained. Step 753) After obtaining all s _j , the inverse symbol conversion unit 755 converts the symbol expression to the original digital watermark expression (for example, the inverse conversion corresponding to the conversion in the first embodiment is 4). This is a process of converting each of the values of one detection symbol (the value of each detection symbol is 0 to 255) into an 8-bit character of 4 characters assuming that the value is an ASCII code, and generates and outputs a detection result 203. .

The detection result 203 represents a digital watermark embedded in the image to be detected. In the digital watermark detection processing according to the present embodiment, since there is no restriction on the minimum size of the detection target image, detection from a partial image smaller than that in the above-described sixth embodiment becomes possible, and the number of blocks used for detection is reduced. Since the number is larger than that in the sixth embodiment, it is possible to perform detection with higher accuracy than in the sixth embodiment. Further, detection with higher accuracy is possible as compared with the fourth embodiment (detection is performed only from one block).

[Eighth Embodiment] Next, as an eighth embodiment of the present invention, an image to be detected is divided into blocks by offset information, and an added block to which the pixel value of the pixel block is added is used as a pixel block. The position marker detection unit that outputs as is described below. Parts other than the following description are the same as in the above-described second embodiment.

FIG. 64 shows the configuration of the block generator of the position marker detector according to the eighth embodiment of the present invention. The block generation unit 800 of the position marker detection unit 230 includes a block division unit 810 and a block addition unit 820. FIG.
It is a flowchart which shows operation | movement of the block generation part of the position marker detection part of 8th Example of this invention.

Step 810) Block generator 800
66, the detection target image 201 and the offset candidate information 204 are input, and as shown in FIG. N ×
Divide into N pixel blocks. Step 820) Next, the block adder 820 adds all these blocks to generate an addition block 802. B _ij ^(t) : t-th block (0 ≦ i <N, 0 ≦ j <
When N, 0 ≦ t <T), the addition block A _ij (0 ≦ i <N, 0 ≦ j <N)
Is

(Equation 8)

Is obtained. Step 830) The block generator 800 outputs the addition block 802 as a pixel block 803 to the discrete Fourier transformer 234. Next, the block generation unit 241 of the digital watermark detection unit 240 in this embodiment will be described with reference to FIG.

The digital watermark detection section 240 in this embodiment
The block generation unit 241 receives the detection target image 201 and the offset information 231 as inputs, and outputs the position marker detection unit 230
(The offset information is input as the offset candidate information), the detection target image is shifted from the position shifted by the offset information (a, b) to T N
Divide into × N blocks. Next, the block generation unit 24
Within 1, all these blocks are added to generate an added block. This addition block is output to the discrete Fourier transform unit 242 as a pixel block 246.

In the digital watermark detection processing in this embodiment, the detection target image (2N × 2N size or more) is divided into N × N size blocks, and the position marker detection and the digital watermark Perform detection. The watermark pattern, which is a repetitive pattern of N × N size, is emphasized T times by the addition of the T blocks, but since the pattern of the original image has low correlation between the blocks, the addition is gradually canceled by the addition. That is,
The influence of the original image is reduced by the detection from the addition block, and at the same time, the watermark pattern is emphasized, so that the detection with higher accuracy than in the second embodiment can be performed.

Although the processing amount involved in the detection is increased by the amount of the addition block as compared with the second embodiment, this processing amount can be compared with the processing amounts of other parts of the detection processing.
It is so small that it can be neglected, and 1 as in the fifth embodiment.
Compared to the case of detecting from one block at a time, processing can be performed at a speed faster by the number of blocks. [Ninth Embodiment] Next, as a ninth embodiment of the present invention,
The detection target image is divided into blocks, a block obtained by adding the pixel values of the divided blocks is set as an addition block, and the addition block is cyclically shifted by offset information to perform a process of generating a pixel block. The position marker detection unit and the digital watermark detection unit will be described.
The parts other than those described below are the same as those in the third embodiment.

FIG. 67 shows the configuration of the block generator of the position marker detector according to the ninth embodiment of the present invention. The block generation unit 900 shown in FIG. 9 includes a block division unit 910, a block addition unit 920, and a cyclic shift unit 930. FIG. 68 is a flowchart showing the operation of the block generator of the position marker detector according to the ninth embodiment of the present invention.

Step 910) Block generator 900
Receives the detection target image 201 and the offset candidate information 204, and the block division unit 910 divides the detection target image 201 into T blocks of N × N pixels from the upper left as shown in FIG. Step 920) Next, the block adder 920 adds all these blocks to generate an addition block 902. B _ij ^(t) : t-th block (0 ≦ i <N, 0 ≦ j <
When N, 0 ≦ t <T), the addition block A _ij (0 ≦ i <N, 0 ≦ j <N)
902 is

(Equation 9)

Is obtained. Next, the operation of the cyclic shift unit 930 of the block generation unit 900 of the position marker detection unit 230 (FIG. 20) in this embodiment will be described. FIG. 70 is a diagram for explaining the operation of the cyclic shift unit according to the ninth embodiment of the present invention.

As shown in the figure, in the cyclic shift section 930, the addition block 902 is cyclically shifted based on the offset candidate information 204 to generate a pixel block 246, which is output to the discrete Fourier transform section 234. . In the expression, A _ij : the pixel block 246 obtained by cyclically shifting the addition block (0 ≦ i <N, 0 ≦ j <N) by the offset candidate information (a, b) 204 is represented by C _ij = A _{(i + a)% N, (j + b)% N} (0 ≦ i <N, 0 ≦ j <N).

Next, the block generation unit of the digital watermark detection unit 240 in this embodiment will be described with reference to FIG. The block generation unit 241 of the digital watermark detection unit 240 in this embodiment receives the detection target image 201 and the offset information 231 as input, and performs the same operation as the block generation unit 900 in the position marker detection unit 230 (using offset information as offset candidate information). Input), the detection target image 201
Is divided into T blocks of N × N pixels from the upper left.

Next, in a block adder in the block generator 241, all these blocks are added.
Generate an addition block. Then, the block generation unit 24
In the cyclic shift unit 1, the addition block is cyclically shifted based on the offset information to generate a pixel block 246, and outputs the pixel block 246 to the discrete Fourier transform unit 234.

In the digital watermark detection processing according to this embodiment,
The required minimum size of the detection target image is N × N, and detection from the detection target image smaller than in the eighth embodiment described above is possible. Further, since the number of divided blocks T is larger than that in the eighth embodiment, the influence of the original image is smaller in the detection, and the watermark pattern is further emphasized. High detection is possible.

In addition, the influence of the original image on detection is reduced by the addition of blocks, and the watermark pattern is emphasized. Therefore, detection with higher accuracy than in the third embodiment is possible. Further, the processing amount associated with the digital watermark detection increases by the amount of generating the addition block as compared with the third embodiment, but this processing amount is negligible when compared with the processing amounts of other parts of the detection processing. Compared with the case where detection is performed from one block at a time as in the sixth embodiment, processing can be performed at high speed by the number of blocks.

[Tenth Embodiment] Next, a tenth embodiment of the present invention will be described.
As an example, the position marker detection unit 230 will be described. The parts other than those described below are the same as those in the fourth embodiment. In the present embodiment, the position marker detection unit 2
In step 30, the detection target image is divided into blocks. At this time, an average value corresponding to the detection target image at the time of division is obtained for a pixel area less than one block around the edge of the detection target image. This is different from the ninth embodiment in that the average of the pixel values corresponding to the image to be detected is obtained and the average value is filled in to form a block.

First, the block generation unit of the position marker detection unit 230 in this embodiment will be described. FIG.
17 shows a configuration of a block generation unit of a position marker detection unit according to a tenth embodiment of the present invention. Block generator 100 shown in FIG.
0 indicates the block dividing unit 1010 and the block adding unit 102
0, a cyclic shift unit 1030.

FIG. 72 is a flowchart showing the operation of the block generator of the position marker detector according to the tenth embodiment of the present invention. Step 1010) The block generation unit 1000 receives the detection target image 201 and the offset candidate information 204 as inputs, and the block division unit 1010 converts the detection target image 201 into blocks of N × N pixels from the upper left as shown in FIG. Divide into pieces. At this time, for a pixel area less than one block around the edge of the detection target image 201,
As shown in FIG. 74, the average value of the detection target image equivalent (FIG. 74A) included when trying to cut out by N × N is obtained, and the missing portion is the average of the pixel values equivalent to the detection target image 201. Find the value and fill it with this.

Step 1020) Next, the block adder 1020 adds all these blocks to generate an addition block 1012. B _ij ^(t) : t-th block (0 ≦ i <N, 0 ≦ j <
When N, 0 ≦ t <T), the addition block A _ij (0 ≦ i <N, 0 ≦ j <N)
1012 is

(Equation 10)

Is obtained. Next, the operation of the cyclic shift unit of the block generation unit of the position marker detection unit 230 in this embodiment will be described. FIG. 75 is a diagram for explaining the operation of the cyclic shift unit according to the tenth embodiment of the present invention.

In cyclic shift section 1030,
Addition block 1012 based on offset candidate information 204
Is cyclically shifted to generate a pixel block, and this is output to the discrete Fourier transform unit 234. _Aij : A _ij : a pixel block obtained by cyclically shifting the addition block (0 ≦ i <N, 0 ≦ j <N) by the offset candidate information (a, b) 204 is C _ij = A _{( i + a)% N, (j + b)% N} (0 ≦ i <N, 0 ≦ j <N).

Next, a block generation unit (not shown) of the digital watermark detection unit according to the tenth embodiment of the present invention will be described. In the present embodiment, the block generation unit of the digital watermark detection unit receives the image to be detected and the offset information as input, and, like the block generation unit of the position marker detection unit (inputs the offset information as offset candidate information), The image is divided into T N × N pixel blocks from the upper left.

Next, all these blocks are added in a block adding section to generate an added block. Then, the cyclic shift unit cyclically shifts the addition block based on the offset information to generate a pixel block, and outputs the pixel block to the discrete Fourier transform unit. In the digital watermark detection processing according to the present embodiment, the required minimum size of the detection target image is an arbitrary size, and detection can be performed from a detection target image smaller than that in the ninth embodiment.

Furthermore, since the number of divided blocks T is larger than in the ninth embodiment, the influence of the original image on detection is smaller and the watermark pattern is more emphasized. Highly accurate detection becomes possible. In addition, the influence of the original image on the detection is reduced by the addition of the blocks, and the watermark pattern is emphasized, so that the detection can be performed with higher accuracy than the fourth embodiment.

Although the processing amount involved in the digital watermark detection is increased by the amount of generating the addition block as compared with the fourth embodiment, this processing amount can be compared with the processing amounts of other parts of the detection processing. It is so small that it can be ignored, and can be processed as fast as the number of blocks compared to the case of detecting from one block at a time as in the seventh embodiment. [Eleventh Embodiment] Next, an electronic watermark detection apparatus will be described as an eleventh embodiment of the present invention.

The digital watermark detecting apparatus according to the present embodiment uses the method of the first embodiment to cut out a part of an embedded image in which a digital watermark is embedded and perform processing such as irreversible compression. Is a device for detecting an embedded digital watermark. FIG. 76 shows the configuration of the digital watermark detection apparatus according to the eleventh embodiment of the present invention. The digital watermark detection device 1100 shown in the figure includes a position marker detection unit 1110, a position marker sequence generation unit 1120, a detection target component specification unit 11
30 and a digital watermark detection unit 1140.

FIG. 77 is a flow chart showing the operation of the digital watermark detecting apparatus according to the eleventh embodiment of the present invention. Step 1110) The digital watermark detection device 1100
Detection target image 201 and key information 202 used at the time of embedding
Is entered. First, in the detection target component specification unit 1130 of the digital watermark detection device 1100, the embedding target component specification unit 1 in the first embodiment described above is obtained from the key information 202.
30, the detection target component position information 11
03 is sent to the position marker detection unit 1110 and the digital watermark detection unit 1140.

Step 1120) The position marker sequence generation unit 1120 determines the position marker sequence 110 from the key information 202.
2 is generated and sent to the position marker detection unit 1110. Step 1130) The position marker detection unit 1110 cuts out a pixel block of N × N (N is the size of the watermark pattern in the first embodiment) size from the upper left of the detection target image 201, and starts from the start point of the cut out pixel block. The offset information 1101 indicating how much the starting point of the watermark pattern at the time of embedding is shifted is output and sent to the digital watermark detection unit 1140.

Step 1140) Digital watermark detection unit 1
Reference numeral 140 denotes an input of a detection target image 201, key information 202, detection target component position information 1103, and offset information 1101, and cuts out an N × N size pixel block from the upper left of the detection target image 201; And the detection target component position information 1103,
The digital watermark embedded in the pixel block is detected, and this is output as a detection result 203.

Next, the detection target component specifying section 1130 in this embodiment will be described. FIG. 78 shows the first embodiment of the present invention.
FIG. 79 shows the configuration of a detection target component designation unit according to the first embodiment, and FIG.
23 is a flowchart showing the operation of the detection target component specifying unit according to the eleventh embodiment of the present invention. Step 1111) The detection target component designation unit 1130
The key information 202 is input, and a random number sequence is generated using the random number generator 1131 with the key information 202 as an initial value.

(Step 1112) Further, based on the generated random number sequence, the detection target component specifying unit 1130 detects the detection target component position which is a sequence having the same length as the length n of the embedding sequence used in the first embodiment. The information 1103 is generated. The detection target component position information 1103 has the following configuration. L _k = (x _k , y _k , z _k ) (0 ≦ k <n) x _k : coefficient to be detected in the x direction y _k : coefficient to be detected in the y direction z _k : flag representing the imaginary value of the coefficient to be detected ( Real component or imaginary component) The detection target component specification unit 1130 performs exactly the same operation as the embedding target component specification unit in the first embodiment. That is, using the same key information as input, the embedding target component position information 131 generated by the embedding target component specifying unit 130 of the first embodiment and the detection target component specifying unit 113 of the present embodiment
The detection target component position information 1103 generated at 0 is exactly the same. Generated detection target component position information 110
3 is sent to the position marker detection unit 1110.

Next, the position marker sequence generation unit 1120 in this embodiment will be described. FIG. 80 shows the configuration of the position marker sequence generator of the eleventh embodiment of the present invention, and FIG. 81 is a flowchart showing the operation of the position marker sequence generator of the eleventh embodiment of the present invention. Step 1121) Position marker sequence generation unit 1120
Receives the key information 202 as an input, and
1, a spreading sequence {r _i ⁽⁰⁾ } (0 ≦ i <n) having a zeroth length n with the key information 202 as an initial value is generated, and this spreading sequence is directly used as a position marker sequence {p _i } (0 ≦ i
<N) 1102 to ^{_{(p i: = r i (}} 0) (0 ≦ i <
n)). As the spreading sequence generator 1121 of the position marker sequence generating unit 1120, one that performs the same operation as the spreading sequence generator 122 of the embedded sequence generating unit 120 in the first embodiment is used.

Next, the position marker detection unit 111 of this embodiment
0 will be described. FIG. 82 shows the configuration of the position marker detection unit according to the eleventh embodiment of the present invention. A position marker detection unit 1110 shown in FIG. 11 includes a block generation unit 1111, a discrete Fourier transform unit 1113, and a detection target sequence generation unit 111.
5, offset candidate information generating section 1117, position marker detection information generating section 1118, offset information generating section 1119
Consists of

FIG. 83 is a flow chart showing the operation of the position marker detecting section according to the eleventh embodiment of the present invention. Step 1131) The position marker detection unit 1110 receives the detection target image 201, the detection target component position information 1103, and the position marker sequence 1102 as input, and first, the block generation unit 1111 performs N × N size detection from the upper left of the detection target image 201. The pixel block 1104 is cut out.

Step 1132) Next, the discrete Fourier transform unit 1113 performs a discrete Fourier transform on the pixel block 1104 to obtain a detection target coefficient matrix 1105. Step 1133) Next, the offset candidate information generation unit 1117 converts the offset candidate information (a, b) to (0,
0) to (N-1, N-1). Step 1134) Hereinafter, the detection target sequence generation unit 111
5 to the position marker detection information generation unit 1118 are performed for each offset candidate information 1106.
The detection target sequence generation unit 1115 receives the detection target component position information 1103, the detection target coefficient matrix 1105, and the offset candidate information 1106, and generates a detection target sequence 1107 as follows. FIG. 84 is a flowchart illustrating the operation of the detection target sequence generation unit of the position marker detection unit according to the eleventh embodiment of the present invention.

Step 1134-1) The detection target sequence generation unit 1115 converts the detection target coefficient matrix 1105 to F (u,
v) (0 ≦ u <N, 0 ≦ v <N), and the detection target component position information 1103 is _represented by L _k = (x _k , y _k , z _k ) (0 ≦ k <
n), the offset candidate information 1106 is (a, b),
When the detection target sequence 1107 is represented by {q _k } (0 ≦ k <n), when z _k is a value representing a real component, step 113 is executed.
The process proceeds to 4-2, and when z _k is a value representing an imaginary component, the process proceeds to step 1134-3. Step 1134-2)

[Equation 11]

The processing described above is performed for k = 0... N-1. Step 1134-3)

(Equation 12)

The above processing is performed for k = 0... N-1. Step 1134-4) detection object sequence {q _k} 110
7 is output. Step 1135) Next, the position marker detection information generation unit 1118 obtains a correlation value between the position marker sequence 1102 and the detection target sequence 1107 by the following equation, and combines this with the offset candidate information 1106 to form the position marker detection information co.
Output as rr _ab 1108.

(Equation 13)

The processing from the detection target sequence generation unit 1115 to the position marker detection information generation unit 1118 is performed when the offset candidate information (a, b) is (0, 0) to (N−1, N−
The position marker detection information 1108 is sequentially sent to the offset information generation unit 1119 while repeating the above-described steps 1). Step 1136) Offset information generating section 1119
Is the offset candidate information (a, b) 1 when the correlation value between the position marker sequence 1102 and the detection target sequence 1107 is the maximum among all the input position marker detection information 1108.
106 is output as offset information 1101 and sent to the digital watermark detection unit 1140.

The offset information 1101 output from the position marker detection unit 1110 indicates how far the starting point of the watermark pattern at the time of embedding from the upper left of the detection target image 201 is shifted. Next, the digital watermark detection unit 1140 in this embodiment will be described. FIG. 85
Shows the configuration of an electronic watermark detection unit according to the eleventh embodiment of the present invention. The digital watermark detection unit 1140 shown in the figure includes a block generation unit 1141, a discrete Fourier transform unit 1143, a detection target sequence generation unit 1145, a symbol detection unit 1147,
It comprises a detection result generator 1149.

FIG. 86 is a flow chart showing the operation of the digital watermark detecting section according to the eleventh embodiment of the present invention. Step 1141) The digital watermark detection unit 1140 receives the detection target image 201, the detection target component position information 1103, the offset information 1101, and the key information 202, and first, the block generation unit 1141 causes the detection target image 2
An N × N pixel block 1142 is cut out from the upper left of 01, and sent to the discrete Fourier transform unit 1143.

Step 1142) The discrete Fourier transform unit 1143 performs a discrete Fourier transform on the pixel block 1142 to generate a detection target coefficient matrix 1144, and sends this to the detection target sequence generation unit 1145. Step 1143) The detection target sequence generation unit 1145
Detection target sequence generation unit 111 of position marker detection unit 1110
A detection target sequence 1146 is obtained by the same processing as in step 5 (using offset information 1101 as offset candidate information), and sent to symbol detection section 1147.

Step 1144) Symbol detector 11
47 generates symbol detection information 1148 by obtaining a correlation value between a sequence generated from the key information 202 and each symbol candidate information and the detection target sequence 1146. Step 1145) All symbol detection information 114
After obtaining the symbol No. 8, the detection result generation unit 1149 sets, for each symbol position of the symbol detection information 1148, the symbol candidate value at the time when the correlation value is the maximum as the detected symbol.

Step 1146) After the detected symbols are determined from all the symbol positions, the result of inversely converting the detected symbols into the digital watermark format is output as the detection result 203. Next, the symbol detector 1147 in the present embodiment will be described. FIG. 87 shows the configuration of the symbol detector of the eleventh embodiment of the present invention. Symbol detector 1 shown in FIG.
147 denotes a symbol candidate generation unit 11471, a symbol sequence generation unit 11473, and a symbol detection information generation unit 1147.
5 is comprised. The symbol detection unit 1147 receives the key information 202 and the detection target sequence 1146 as inputs, and outputs symbol detection information 1 for M symbol candidates for each of a predetermined number of J symbol positions to be detected.
1476, and outputs the result to the detection result generation unit 1149. Hereinafter, detection of the j-th (1 ≦ j ≦ J) symbol will be described.

FIG. 88 is a flowchart showing the operation of the symbol detector of the eleventh embodiment of the present invention. Step 1144-1) Symbol Candidate Generation Unit 1147
1 sequentially generates symbol candidates c from 0 to M−1,
Steps 1144-2 to 1144 for each symbol candidate
The processing up to 3 is performed. M represents the maximum value of the symbol value by the symbol conversion in the first embodiment.

Step 1144-2) The symbol sequence generator 11473 will be described. FIG. 89 shows the configuration of the symbol sequence generator of the eleventh embodiment of the present invention, and FIG. 90 is a flowchart showing the operation of the symbol sequence generator of the eleventh embodiment of the present invention. Step 1144-2-1) Symbol Sequence Generation Unit 11
473 receives the key information 202 and the symbol candidate c as input,
Using the value obtained by adding c and the key information 202 as an initial value, a spreading sequence {r
^{_{i (j)} (0 ≦}} i <n) to generate, is used as a directly symbol sequence ^{_{11474 {p i (j)}}} (0 ≦ i <n) (p i (j): = r i (j ⁾ (0 ≦ i <n)). The symbol sequence 11474 is converted to a symbol detection information generation unit 1147.
Send to 5.

Step 1144-3) The symbol detection information generation section 11475 receives the detection target sequence 1146, the symbol sequence 11474, the symbol candidate c, and the symbol position j currently being processed and detects the symbol sequence 11474 by the following equation. The correlation value of the target series 1146 is obtained,
Generating a symbol detection information corr _c ^(j) 1148 correlation values and symbol candidate and the symbol position as a set.

[Equation 14]

[0297] Symbol detection information 1148 is sent to detection result generating section 1149. Next, the detection result generation unit 1149 according to the present embodiment will be described. FIG. 91 shows the configuration of the detection result generation unit according to the eleventh embodiment of the present invention. The detection result generation unit 1149 shown in FIG. 7 includes a sequence generation unit 11491 for each symbol position, a detection symbol generation unit 1149.
3. It is composed of an inverse symbol converter 11495.

FIG. 92 is a flow chart showing the operation of the detection result generator according to the eleventh embodiment of the present invention. Step 1145-1) The detection result generation unit 1149
Inputs the symbol detection information 1148, first, in each symbol position for each sequence generating unit 11491 divides the symbol detection information 1148 for each symbol position j, lengths each symbol position for each sequence corr _c of M ^{(j), (0} ≤c <M) 1
1492 is generated for each of j = 1 to J and sent to the detected symbol generation unit 11493.

Step 1145-2) The detected symbol generation section 11493 receives the symbol position sequence 11492 as an input, and for each symbol position j, determines a symbol candidate c for obtaining the maximum correlation value in each symbol position sequence. The detected symbol s _j (1 ≦ j <J) 11494 is generated. Step 1145-3) After obtaining all s _j , the inverse symbol converter 11495 converts the symbol expression to the original digital watermark expression (for example, the inverse conversion corresponding to the conversion in the first embodiment is 4). The value of each of the two detected symbols (the value of each detected symbol is 0 to 255) is represented by A
This is a process of converting the four characters into 8-bit characters assuming the SCII code), and generates and outputs a detection result 203.

The detection result 203 indicates that the detection target image 20
1 represents a digital watermark embedded in the digital watermark. Thus, the processing of the digital watermark detection device in the eleventh embodiment ends. In the digital watermark detection according to the present embodiment, a block is cut out from the upper left of the detection target image, a discrete Fourier transform is performed to obtain a detection target coefficient matrix 1105, and then an offset and a coefficient are added to the component specified by the detection target component position information 1103. Multiply by the phase difference determined by the order of
A detection target sequence 1107 is generated. This is due to the property F (u, v) ← → f of the Fourier transform regarding translation.
(X, y) (← → indicates Fourier transform / inverse transform)

(Equation 15)

Is used. Accordingly, only one discrete Fourier transform is required at the time of detection, and processing can be performed at higher speed than the detection processing of the second to tenth embodiments. (Comparison of the processing time by the computer simulation in the block size of N = 128 is the second embodiment: the eleventh embodiment = 65: 1).

Also, a partial image (size: N) of the embedded image in the first embodiment cut out from an arbitrary position
× N or more) can be detected. [Twelfth Embodiment] Next, as a twelfth embodiment of the present invention, when the detection target image is divided into blocks in the block generation unit of the position marker detection unit, the average value corresponding to the detection target image is calculated. A description will be given of an example in which an average value of pixel values corresponding to an image corresponding to a detection target image is calculated for a missing portion, and a portion where the insufficient value is filled is blocked. Except for the parts described below, the configuration is the same as that of the above-described eleventh embodiment.

FIG. 93 is a view for explaining the operation of the block generator of the position marker detector according to the twelfth embodiment of the present invention. Position marker detection unit 1110 in the present embodiment
Receives the detection target image 201 and cuts out a block of N × N size from the upper left of the detection target image 201. At this time, if the detection target image 201 is too small to be cut out to the N × N size, only the detection target image equivalent (A in the figure) included in the N × N cutout is cut out, and the N × N size is cut out. The portion that is insufficient in size is obtained by averaging the pixel values corresponding to the image to be detected and filling it.

The block obtained by the above processing is output as a pixel block as in the eleventh embodiment. Next, an operation of generating a block by the digital watermark detection unit in the present embodiment will be described. The block generation unit of the digital watermark detection unit in the present embodiment receives a detection target image as an input, and obtains and outputs a pixel block by the same processing as the block generation unit in the position marker detection unit.

In the case of the digital watermark detection processing using this embodiment, the size required for the image to be detected can be any size from N × N or more as compared with the above-described eleventh embodiment. Can be detected. [Thirteenth Embodiment] Next, as a thirteenth embodiment of the present invention, a digital watermark detection section, a symbol detection section, and a detection result generation section will be described. Except for the parts described below, the configuration is the same as that of the above-described eleventh embodiment.

FIG. 94 shows the configuration of the digital watermark detecting section according to the thirteenth embodiment of the present invention. The digital watermark detection unit 2000 shown in FIG. 14 includes a block division unit 2100, a discrete Fourier transform unit 2200, a detection target sequence generation unit 2300, a symbol detection unit 2400, and a detection result generation unit 2500. FIG. 95 is a flowchart showing the operation of the digital watermark detection unit according to the thirteenth embodiment of the present invention.

Step 2100) Digital watermark detection unit 2
000 receives the detection target image 201, the detection target component position information 211, the offset information 231, and the key information 202 as input, first, the block division unit 2100 converts the detection target image 201 into N from the upper left as shown in FIG. × N
It is divided into T pixel blocks of a size. Step 2200) Numbers 0 to T−1 are assigned to the individual pixel blocks 2001 obtained by the division. t (0 ≦
The following steps 2300 to 2400 are repeated for the (t <T) -th pixel block 2001.

Step 2300) The discrete Fourier transform unit 2200 performs a discrete Fourier transform on the t-th pixel block 2001 to obtain a detection target coefficient matrix 2002. Step 2400) The detection target sequence generation unit 2300
The detection target sequence 2003 is obtained by the same processing as the detection target sequence generation unit of the position marker detection unit, and sent to the symbol detection unit 2400.

Step 2500) Symbol detection unit 24
00 generates a symbol detection information 2004 by obtaining a correlation value between a sequence generated from the key information 202 and each symbol candidate information and a detection target sequence 2003. Step 2600) After obtaining the symbol detection information 2004 for all t, the detection result generation unit 2500 sets, for each symbol position of the symbol detection information 2004, the symbol candidate value with the maximum correlation value as the detection symbol, and After the detected symbol is determined from the symbol position, the result of inversely converting the detected symbol into a digital watermark format is output as a detection result 203.

Next, the symbol detecting section 2 in the present embodiment
400 will be described. FIG. 97 shows the configuration of the symbol detector of the thirteenth embodiment of the present invention. The symbol detection unit 2400 shown in FIG.
It comprises a symbol sequence generation unit 2420 and a symbol detection information generation unit 2430. The symbol detection unit 2400
Key information 202 and (obtained from the t-th pixel block)
The detection target sequence 2003 is input, and symbol detection information 2004 for M symbol candidates is generated for each of the J symbol positions to be detected in advance and output to the detection result generation unit 2500.

FIG. 98 is a flow chart showing the operation of the symbol detector of the thirteenth embodiment of the present invention. Step 2410) The symbol candidate generation unit 2410 sequentially generates the symbol candidates c2401 from 0 to M−1, and performs steps 2420 to 24 for each symbol candidate.
Step 30 is repeated. M represents the maximum value of the symbol value by the symbol conversion in the first embodiment.

Step 2420) Here, the symbol sequence generation section 2420 will be described. FIG. 99 shows a configuration of a symbol sequence generation unit according to the thirteenth embodiment of the present invention;
FIG. 100 shows the operation of the symbol sequence generation unit according to the thirteenth embodiment of the present invention. Step 2421) The symbol sequence generation unit 2420
The key information 202 and the symbol candidate c2401 are input and c
The spreading sequence generator 2421 generates a j-th spreading sequence {r _i ^(j) } (0 ≦ i <n) having a length of n and uses the value as a symbol used as a sequence ^{_{{p i (j)} (}} 0 ≦ i <n) 2402 (p i (j): = r i (j) (0 ≦ i <n)). The symbol sequence 2402 is converted to a symbol detection information generation unit 2430.
Send to

Step 2430) The symbol detection information generating section 2430 calculates the detection target sequence 2003, the symbol sequence 2402, the symbol candidate, and the symbol position j currently being processed.
As input, the following equation obtains the correlation value of the detection object sequence 2003 and the symbol sequence 2402 by a, and the correlation values and symbol candidate symbol positions and the pixel block position a set symbol detection information ^{_{corr c (j) (t)}} 2403 Generate

(Equation 16)

[0314] Symbol detection information 2004 is sent to detection result generation section 2500. Next, the detection result generation unit 2500 according to the present embodiment will be described. FIG. 101 shows the configuration of the detection result generation unit according to the thirteenth embodiment of the present invention.
The detection result generation unit 2500 shown in FIG. 17 includes a sequence generation unit 2510 for each symbol position, a detection symbol generation unit 2520,
It comprises an inverse symbol conversion unit 2530.

FIG. 102 is a flowchart showing the operation of the detection result generator according to the thirteenth embodiment of the present invention. Step 2510) The detection result generation unit 2500 receives the symbol detection information 2004 as input, and first, the symbol generation
004 is divided for each symbol position j, and a sequence corr _c ^{(j) (t)} (0 ≦ c <M, 0 ≦
t <T) 2501 is generated for each j from j = 1 to j and sent to the detected symbol generation unit 2520.

[0316] Step 2520) The detected symbol generation unit 2520 receives the sequence 2501 for each symbol position, finds a symbol candidate c at the time of taking the maximum correlation value in the sequence for each symbol position for each symbol position j, and A detected symbol s _j (1 ≦ j <J) 2502 is generated. The above processing is repeated until all s _j are obtained. Step 2530) After obtaining all s _j , the inverse symbol conversion unit 2530 converts the symbol expression to the original digital watermark expression (for example, four inverse detections corresponding to the conversion in the first embodiment are performed). The value of each symbol (the value of each detected symbol is 0 to 255)
This is a process of converting the four characters into 8-bit characters assuming the II code) to generate and output the detection result 203. The detection result 203 represents a digital watermark embedded in the detection target image.

In the digital watermark detection processing according to the present embodiment,
Since the detection target image (the size of N × N or more) is divided into blocks and the symbol at which the maximum correlation value is obtained among all the blocks is detected, the eleventh embodiment (1)
Detection can be performed with higher accuracy than that of detection from only one block). Further, since the number of discrete Fourier transforms required in the detection processing is only the number of blocks, processing can be performed at a higher speed than in the fifth embodiment.

[Fourteenth Embodiment] Next, a fourteenth embodiment of the present invention will be described.
As an embodiment of the present invention, a description will be given of a process in a case where an N × N size block cannot be cut out from a detection target image in a block division unit of a digital watermark detection unit. FIG.
It is a figure for explaining operation of a block division part of a digital watermark detection part of a 14th example of the present invention.

The block dividing unit 2100 divides the detection target image 201 into T blocks of N × N size from the left. At this time, around the edge of the detection target image 201, N × N
For a block that cannot be cut out into a size, only the portion corresponding to the detection target image included when trying to cut out by N × N is cut out, and the portion smaller than the N × N size is the average value of the pixel values corresponding to the detection target image. And fill it with this. Numbers 0 to T−1 are assigned to the respective pixel blocks obtained by the division.

In the information detection according to the present embodiment, the required minimum size of the image to be detected can be any size, and detection from the image to be detected smaller than in the thirteenth embodiment can be performed. The number of blocks used for detection is the first.
Since the number is larger than that of the third embodiment, more accurate detection can be performed as compared with the thirteenth embodiment. Further, the number of discrete Fourier transforms required in the detection processing is only the number of blocks, so that the processing can be performed at a higher speed than in the sixth embodiment.

[Fifteenth Embodiment] Next, a fifteenth embodiment of the present invention will be described.
As an example, a position marker detection unit and a digital watermark detection unit will be described. Except for the parts described below,
This is common to the eleventh embodiment. FIG. 104 shows the configuration of the block generator of the position marker detector according to the fifteenth embodiment of the present invention. Block generator 3000 shown in FIG.
Are the block dividing unit 3100 and the block adding unit 3200
Consists of

FIG. 105 is a flow chart showing the operation of the block generator of the position marker detector according to the fifteenth embodiment of the present invention. Step 3100) The block generation unit 3000 receives the detection target image 201, and the block division unit 3100 divides the detection target image into blocks of T N × N pixels from the upper left as shown in FIG.

(Step 3200) Next, the block adder 3200 adds all these blocks to generate an addition block 3002. When this is represented by an equation, B _ij ^(t) : the t-th block (0 ≦ i <N, 0 ≦ j <
When N, 0 ≦ t <T), the addition block A _ij (0 ≦ i <N, 0 ≦ j <N)
Is

[Equation 17]

Is obtained. This addition block 30
02 is output as the pixel block 205 to the discrete Fourier transform unit 1113. In the information detection according to the present embodiment, the detection target image 201 (N × N size or more) is divided into N × N size blocks, and a position marker detection and a digital watermark detection are performed from the block obtained by adding all the blocks.

The watermark pattern, which is a repetitive pattern of N × N size, is emphasized T times by the addition of T blocks, but the pattern of the original image has a low correlation between the blocks, and thus is gradually canceled by the addition. Will be done. That is, the influence of the original image is reduced by the detection from the addition block, and at the same time, the watermark pattern is emphasized. Therefore, detection with higher accuracy than in the eleventh embodiment is possible.

Although the processing amount involved in the detection is increased by the amount of the addition block as compared with the eleventh embodiment, this processing amount can be compared with the processing amounts of other parts of the detection processing. , Which is negligibly small, and can be processed at a speed equal to the number of blocks, as compared to the case of detecting from one block at a time as in the thirteenth embodiment. Further, the number of discrete Fourier transforms required for the detection process is 1
Since it is only once, processing can be performed at a higher speed than in the second to tenth embodiments.

[Sixteenth Embodiment] Next, a sixteenth embodiment of the present invention will be described.
As an example, a block generation unit of the position marker detection unit will be described. Except for the parts described below, the configuration is the same as that of the twelfth embodiment. FIG. 107 shows the configuration of the block generation unit of the position marker detection unit according to the sixteenth embodiment of the present invention. The block generation unit 4000 shown in FIG.
100 and a block adder 4200.

FIG. 108 is a flow chart showing the operation of the block generator of the position marker detector according to the sixteenth embodiment of the present invention. Step 4100) The block generation unit 4000 receives the detection target image 201, and the block division unit 4100 divides the detection target image 201 into T N × N pixel blocks from the upper left as shown in FIG. At this time, for a block that cannot be cut out to the N × N size around the edge of the detection target image 201, only a portion of the detection target image that is included when trying to cut out at the N × N is cut out,
The process of obtaining the average value of the pixel values corresponding to the image to be detected and filling the portion that is smaller than the N × N size is performed. Step 4200) Next, the block adder 4200
Adds all these blocks to generate an addition block 4002. When this is represented by an equation, B _ij ^(t) : t
Block (0 ≦ i <N, 0 ≦ j <N, 0 ≦ t <
T), the addition block A _ij (0 ≦ i <N, 0 ≦ j <
N)

(Equation 18)

Is obtained. Step 4300) The adding block 4002 is output to the discrete Fourier transform unit 1113 as a pixel block. Next, a block generation unit (not shown) in the digital watermark detection unit according to the present embodiment will be described.

[0330] In the present embodiment, the block generation unit in the digital watermark detection unit receives a detection target image as input, and, in the same manner as the block generation unit 4000 in the position marker detection unit, converts the detection target image from the top left into T N N Divide into blocks of × N pixels. Next, the block adding unit adds all these blocks to generate an added block. This addition block is output as a pixel block to the discrete Fourier transform unit.

In the information detection according to the present embodiment, the required minimum size of the detection target image is an arbitrary size, and detection from a smaller detection target image is possible as compared with the fifteenth embodiment. Further, as compared with the fifteenth embodiment, since the number of divided blocks T is larger, the influence of the original image on detection is smaller, and the watermark pattern is further emphasized. High detection is possible.

Also, by adding blocks, the influence of the original image on detection is reduced, and the watermark pattern is emphasized, so that detection with higher accuracy than in the twelfth embodiment is possible. Also, the processing amount associated with the detection increases by the amount of generating the addition block as compared with the twelfth embodiment.
This processing amount is negligibly small when compared with the processing amounts of the other parts of the detection processing. In addition, compared with the case where detection is performed from one block as in the fourteenth embodiment, the number of processing times is equal to the number of blocks. Only high-speed processing is possible.

[Seventeenth Embodiment] A seventeenth embodiment of the present invention will now be described.
As an example, a detection target sequence generation unit of the position marker detection unit will be described. Except for the parts described below, they are the same as the above-described second to sixteenth embodiments. In this embodiment, the detection target sequence generation unit of the position marker detection unit in each of the second to sixteenth embodiments generates the detection target sequence as described in each embodiment. After the final stage, the processing for setting the average of the detection target sequence to 0 is performed as follows, and the detection target sequence subjected to this processing is used for the subsequent processing. {Q _i }: a detection target sequence before processing (0 ≦ i <n),

[Equation 19] , And the average is changed to 0 by q _i : = q _i −ave (0 ≦ i <n). According to the present embodiment, the average of the detection target sequence is 0, and each term has a positive or negative (or 0) value. As a result, in the subsequent calculation of the correlation value, the correlation value other than when the position marker detection information has a valid peak is suppressed to be low, so that the position marker detection information is not compared with the case where the processing of the present embodiment is not performed. Effective peaks can be detected more accurately.

[Eighteenth Embodiment] Next, an eighteenth embodiment of the present invention will be described.
As an example of (1), a detection target sequence generation unit of the digital watermark detection unit will be described. Except for the parts described below, they are the same as the above-described second to seventeenth embodiments. In the present embodiment, the detection target sequence generation unit in the digital watermark detection unit in the above-described second to seventeenth embodiments performs the final generation of the detection target sequence as described in each embodiment. After the step, a process of setting the average of the detection target sequence to 0 is performed as described below, and the detection target sequence subjected to this process is used for the subsequent processing.

{Q _i }: detection target sequence before processing (0 ≦ i <n),

(Equation 20) Is calculated, and the average is changed to 0 by q _i : = q _i −ave (0 ≦ i <n). According to the present embodiment, the average of the detection target sequence is 0, and each term has a positive or negative (or 0) value. Thereby, in the subsequent calculation of the correlation value, the correlation value other than when the symbol detection information has a valid peak is suppressed to be low, so that the validity of the symbol detection information is smaller than when the processing of the present embodiment is not performed. Peaks can be detected with higher accuracy.

[Nineteenth Embodiment] Next, a nineteenth embodiment of the present invention will be described.
As an example, the offset information generation unit of the position marker detection unit will be described. Except for the parts described below, they are the same as the above-described second to eighteenth embodiments. FIG. 110 is a flowchart showing the operation of the offset information generation unit of the position marker detection unit according to the nineteenth embodiment of the present invention.

Step 5100) With the position marker detection information as an input, the offset information generation unit first obtains the maximum value O _max of all the position marker detection information. Step 5200) Next, a predetermined threshold value β
The processing is divided as follows. If O _max <β, the process proceeds to step 5300; otherwise, the process proceeds to step 5400.

Step 5300) A detection result indicating that the detection of the offset information has failed is output to the outside of the information detecting device, and all the processes in the information detecting device are terminated. Step 5400) Output the offset candidate information (a, b) at the time of taking the maximum value O _max as offset information,
Send to digital watermark detection unit.

According to the nineteenth embodiment, the detection is not continued unless the position marker detection information, that is, the correlation value between the position marker sequence and the detection target sequence is equal to or larger than the threshold value β. As a result, an image without embedded information can be used as a detection target image, or when an embedded image is severely damaged and cannot be detected already, detection is forcibly performed and an incorrect detection result is output. Instead, the detection result of the content indicating that the detection has failed is output, so that the reliability of the output detection result increases.

[Twentieth Embodiment] The twentieth embodiment of the present invention will now be described.
As an example, the detection result generation unit of the digital watermark detection unit will be described. Except for the parts described below, they are the same as the above-described second to nineteenth embodiments. FIG.
11 is a flowchart illustrating the operation of the detection result generation unit of the digital watermark detection unit according to the twentieth embodiment of the present invention.

Step 6100) The detection result generation section sets
With the symbol detection information as an input, the symbol position information is first divided into symbol position j in each symbol position sequence generation unit. Step 6200) The detection result generation unit determines that j (1 ≦ j ≦
The maximum value s _max ^(j) of the J) -th sequence for each symbol position is obtained, and the process branches as follows according to a predetermined threshold value γ.

Step 6300) If s _max ^(j) <γ, the flow proceeds to step 6400; otherwise, the flow proceeds to step 6500. Step 6400) A detection result indicating that the detection of the digital watermark embedded in the detection target image has failed is output, and all the processing in the detection result generation unit ends.

Step 6500) If s _max ^(j) ≧ γ, the symbol candidate c for taking the maximum value s _max ^(j)
Is the j-th detected symbol s _j (1 ≦ j ≦ J). Step 6600) After obtaining all s _j , convert the symbol expression to the original digital watermark expression (for example, the values of four detected symbols (the value of each detected symbol is 0 to 255) in the first embodiment) Are ASCII
This is a process of converting four characters into 8-bit characters assuming that the code is a code), and generates and outputs a detection result.

According to the twentieth embodiment, the detection is not continued unless the symbol detection information, that is, the correlation value between each symbol sequence and the detection target sequence is equal to or larger than the threshold value γ. As a result, an image in which no information is embedded can be used as a detection target image, or when an embedded image is severely damaged and cannot be detected, it is forcibly detected to output an incorrect detection result. Instead, the detection result of the content indicating that the detection has failed is output, so that the reliability of the output detection result increases.

[Twenty-first Embodiment] Finally, the second embodiment of the present invention is described.
A first embodiment will be described. Except for the parts described below, the configuration is the same as in the first and twentieth embodiments. In the present embodiment, the spread sequence generator of the embedded sequence generator in the first embodiment, the spread sequence generator of the position marker sequence generator in the twentieth embodiment, and the spread sequence of the symbol sequence generator in the digital watermark detector are described. The generator randomly and sequentially generates a value of +1 or -1 with a probability of 1/2, respectively (for example, Miyagawa, Iwatari, Imai: "Code Theory")
Shokodo, "M-sequence generation described on page 128", a random sequence consisting of binary values of "+1" and "-1" obtained using a k-th generator polynomial (k is a sufficiently large integer), etc. In this embodiment, the following is used as a calculation formula for calculating the correlation in the symbol detection information generation unit of the symbol detection unit of the digital watermark detection unit in the twentieth embodiment: {p _i }: Symbol sequence (0 ≦ i <n) {q _i }: When the detection target sequence (0 ≦ i <n), the correlation value of these two sequences is

(Equation 21) Considering the distribution of correlation values in the symbol detection information generation unit,

(Equation 22)

Is a series that follows a distribution with an average of 0 (according to the eighteenth embodiment) and a variance of 1 / n. Also, p _i (0 ≦ i
<N) represents the value of +1 or -1 with probability 1 /
Take 2 The sequence follows a distribution with mean 0 and variance 1. n
Is large enough,

(Equation 23) Is a series that follows a distribution with mean 0 and variance 1 / n. Here, using the central limit theorem in stochastic statistics, when n is sufficiently large,

(Equation 24)

Is a normal distribution N (0, 0
Follow 1). FIG. 112 shows the values of corr and the theoretical values N by computer simulation of the twenty-first embodiment of the present invention.
It is a graph showing (0, 1) (n = 1024), and it can be seen that corr follows N (0, 1) as inferred. If we know that corr obeys N (0,1),
The threshold value γ in the twentieth embodiment can be determined by accuracy (a scale indicating how much a detected peak is significant). For example, if the accuracy is 99.999% (the probability that a peak above the threshold appears by chance is 10 ⁻⁵ ),
From the normal distribution table, the threshold value γ at that time is 4.26489. FIG. 113 is a graph of symbol detection information by computer simulation according to the twenty-first embodiment of the present invention. As shown in the figure, four symbols (one symbol is 0 to 25)
As a result of detecting "A", "B",
Significant peaks are seen in symbol values representing “C” and “D”. The above γ is used to evaluate how significant this peak is (not a value by chance). In the example of FIG. 113, since each peak value is 4.26489 or more, these peaks with an accuracy of 99.999 or more may be determined to be significant.

According to the twenty-first embodiment, since the distribution of the symbol detection information can be made to be a normal distribution with a mean of 0 and a variance of 1, the symbol detection can be performed by specifying the accuracy. It becomes possible. As a result, it is possible to grasp how likely the detection result is, to increase the reliability of the detection result, and to improve the convenience. Computer simulation was performed to confirm the effectiveness of the proposed method of the present invention.
As a test image, an 8-bit grayscale image of 512 × 512 pixels (Lena, sailboat, bab
oon) was used. The values of the parameters used in the simulation are N = 128, n = 1023, α = 1.
_{4, M = 256 (u n} , v n) are

(Equation 25)

The watermark information is 32 bits (ASCII character string "ABCD") selected from the medium frequency band specified by
It is. Here, the image quality of the watermarked image will be described. FIG. 114 shows the PSNR when the embedding strength power is changed. However, the value of power in the graph is power
= Expressed as a relative value based on the time of ξ. The difference in PSNR with respect to power depending on the image is due to local weighting.

For the parallel movement / cutting, first, a watermarked image (Lena, power = ξ, PSNR = 42.107)
873) shows the result of detecting the watermark information from FIG. At the offset value (0,0), the peak (14.761
541) has been detected. The peaks (1) also appear in the symbols “A”, “B”, “C”, and “D” of the watermark information.
1.180982 to 9.108589), and the watermark information has been successfully detected.

Next, an attempt was made to detect watermark information from a partial image cut out from each image. Offset detection success rate (successful when a correct offset is detected) when changing the size N × N of the partial image (start point is randomly selected for each size), symbol detection success rate (all FIG. 116 shows that the symbol is correctly detected and the minimum value of the four peak values is larger than the threshold value 4.26489. For partial images larger than N = 256, detection of watermark information is almost certainly possible. The watermark information is embedded for each block of 128 × 128 pixels. However, in an image including an edge as shown in FIG. 117 (a), some frequency component values have large values, It is considered to be a large noise due to the image, and the detection has failed. Conversely, a relatively flat image as shown in FIG. 117B or a texture image has been successfully detected. In the case of the detection using the difference from the original image, the detection is considered to be successful because noise due to the original image is canceled.

Next, JPEG compression will be described. Watermarked images (Lena, sailboat, babo
on, power = ξ) was irreversibly compressed by JPEG to detect watermark information. FIG. 118 shows the offset response peak value and the minimum value of the symbol response peak when the quality parameter is changed.
Detection was successful up to the point where the broken line in the figure was discontinued. Considering the threshold described in the above description of the translation / cutting, the effective tolerance is considered to be slightly lower than that in FIG. Next, gradation conversion will be described. An attempt was made to detect watermark information from an image obtained by performing gradation conversion on a watermarked image (Lena, power = ξ). When the image is dropped to a binary image, the offset response peak value obtained from the image (FIG. 119 (a)) dithered by the Floyd-Steinberg method is 13.616614, and the symbol response peak values are 9.39720 to 6.825329. Information was successfully detected. In addition, a simple binary quantized image (FIG. 119)
The offset response peak value obtained from (b)) is 10.388254 and each symbol response peak value is 8.388254.
133438-5.680073, and the watermark information was successfully detected.

Next, detection from the print screen will be described. A watermarked image (Lena, power = ξ) was printed by a printer, and an attempt was made to detect watermark information from an image read by a scanner. A watermarked image converted to 72dpi PostScript data is converted to a black and white laser printer (RICO
HSP-10PS Pro / 6F, 600dpi), printed using a scanner (EPSON GT-6000) at 72dpi, 16-bit grayscale, and detected from an image reduced to 8-bit grayscale. At this time, the offset response peak value was 9.080504, and the symbol response peak values were 8.139693 to 5.615208, and the watermark information was successfully detected. In addition, since it was found that the aspect ratio slightly changed from the combination of the printer and the scanner, in the detection when this was corrected, the offset response peak value was 11.701824, and each symbol response peak value was
11.701824, each symbol response peak value is 8.8076
The watermark information was successfully detected.

Next, the composition will be described. Two watermarked images (Lena, sailboat, both
From the image obtained by combining power = 図) at a ratio of 1: 1 (FIG. 120), detection of watermark information was attempted using respective key information. The offset response peak value is
Lena = 9.549749, sailboat = 9.883106, and each symbol sponse peak value is Lena = 8.102563-6.
325033, sailboat = 6.661891 to 7.285404, and two different pieces of watermark information could be detected from one composite image.

Next, the detection processing time will be described. The offset search space of the proposed method of the present invention is N ² ,
Although it can be said that the full search is performed, when a simple full search in which the block is resampled for each offset candidate and the discrete Fourier transform is performed is compared, the amount of calculation per search is reduced. Also, the calculation of e- ^ix performed during the search process can be speeded up by having a table in advance. The processing time ratio between the proposed method and the above-described simple full search by computer simulation was 1:65. Note that the discrete Fourier transform was performed on a 128 × 128 pixel block, and a fast Fourier transform (FFT) was used.

[Twenty-second Embodiment] FIG. 121 shows a configuration of a local complexity creating section for an image according to a twenty-second embodiment of the present invention. The image local complexity creating unit 3010 shown in FIG.
An edge / texture component value image creating unit 3110;
The local complexity creating section 3010 for this image is used in a digital watermark embedding device.

FIG. 122 is a flowchart showing the operation of the local complexity creating section for images according to the twenty-second embodiment of the present invention. The image local complexity creating unit 3010 receives the input image 10 (step 22110), and
An edge texture component value image is generated from 0 (step 22120), and the edge texture component value image is output as the local complexity of the image (step 2213).
0).

More specifically, edge / texture component value image creating section 3110 creates an edge / texture component value image in the following procedure. FIG. 123 is a flowchart showing the operation of the edge / texture component value image creating unit according to the twenty-second embodiment of the present invention. Step 22121) The edge / texture component value image creating unit 3110 converts the input image 10 into {p _ij } (0 ≦ i <
Width, 0 ≦ j <Hight). 0 ≦ i <Width, 0
Step 22122 and subsequent steps are repeated for ≦ j <Hight.

Step 22122) The edge / texture component value image creating section 3110 clears the difference counter count to zero. Step 22123) For -1 ≦ k ≦ 1 and −1 ≦ h ≦ 1 If | p _ij −p _{i + k, j + h} |> thres, the process proceeds to step 22124, and if not, (k , H) = (− 1, −1) to (k, h).

Step 22124) The edge / texture component value image creating section 3110 increments the difference counter (count: = count + 1). Where thr
es is a predetermined threshold value. Step 22125) The value of the difference counter after repeating the above steps 22123 and 22124 is set as the value of the (i, j) coordinate of the edge / texture component value image. When expressed by an expression, e _ij : = count where {e _ij }: an edge / texture component value image (0 ≦ i <Wi
dth, 0 ≦ j <Hight) According to the above-described twenty-second embodiment of the present invention, the input image 10
It is possible to generate a local complexity of an image having a small value in a flat portion of the image and a large value in an edge or texture region.

As shown in FIG. 124, by using a threshold value and a difference counter, a texture having a gradual pattern and a local complexity of an image having the same value at a steep edge are generated. It is possible. [Twenty-third Embodiment] Next, an image local complexity creating section according to a twenty-third embodiment of the present invention will be described.

FIG. 125 shows the configuration of the local complexity creating section for an image according to the twenty-third embodiment of the present invention. The image local complexity creating unit 3200 shown in FIG. 8 includes an edge / texture component value image creating unit 3210 and a texture index creating unit 3200.
1 and an upsampler 3230, as shown in FIG.
In this configuration, a texture degree index creating unit 3220 and an upsampler 3230 are added to the configuration of the twenty-second embodiment shown in FIG.

FIG. 126 is a flowchart showing the operation of the image local complexity creating section according to the twenty-third embodiment of the present invention. When the input image 10 is input (Step 23210), the image local complexity creating unit 3200 creates an edge / texture component value image 3201 from the input image 10 in the edge / texture component value image creating unit 3210 (Step 23210). 23220).

Next, the texture degree index creating section 3220 creates a texture degree index 3202 from the edge / texture component value image 3201, and the upsampler 3230 upsamples the texture degree index 3202 (step 23230). , Input image 1
The size is made the same as 0, and this is output as the local complexity of the image (step 23240).

[0365] The operation of the texture degree index creating unit 3220 will be described. FIG. 127 is a flowchart showing the operation of the texture index generation unit according to the twenty-third embodiment of the present invention. Step 23231) Texture index creation unit 322
0 indicates that the edge / texture component value image 3201 is n × N
It is subdivided into pixel blocks (assuming that as a result of the subdivision, the edge-texture component value image is composed of P × Q blocks). Pixels in the (p, q) -th block are represented by b _ij
^{(p, q)} (0 ≦ i <n, 0 ≦ j <n).

Step 23232) The (p, q) -th texture degree index index _pq is cleared to 0. Step 23233) index _pq : = index _pq + b _ij
^{(p, q),} and (i, j) = (0,0) to (n−1, n−
This process is repeated until 1), and when it is completed, (p, q)
= (0,0) to (P-1, Q-1) Step 232
The process after 32 is repeated.

That is, {b _ij ^(pq) }: block pixel values (0 ≦ i <n, 0 ≦ j <n) of the (p, q) th edge-texture component image, and (p, q) Block (0 ≦ p <P, 0
The texture index index _pq of ≦ q <Q) is obtained by the following equation.

(Equation 26)

Next, the upsampler 3 in this embodiment is used.
The operation of 230 will be described. FIG. 128 is a view for explaining the operation of the upsampler according to the twenty-third embodiment of the present invention. The upsampler 3230 receives the texture degree index 3202 (P × Q size) as input, and
By multiplying, the local complexity (Width × Height size) of an image having the same size as the input image 10 is generated and output.

As described above, according to the present embodiment, the texture degree index is large in a block including many large values in a block group obtained by subdividing an edge / texture component value image into n × n pixels. If the value takes a large value and includes many small values, the texture degree index takes a small value. FIG. 129 shows an edge / texture component value image and a texture degree index according to the twenty-third embodiment of the present invention. As shown in the figure, in the simple edge part (upper right part of the figure) in the edge / texture component value image 3201, the number of points having a large value included in the block is small. Takes a small value. Thus, by using this embodiment, it is possible to generate the local complexity of an image having a larger value in the texture region than in the twenty-second embodiment described above.

[Twenty-fourth Embodiment] Next, a twenty-fourth embodiment of the present invention will be described.
An example will be described. Except for the parts described below, the configuration is the same as that of the above-described twenty-third embodiment. First, the texture degree index creation unit 3220 of the digital watermark embedding device according to the present embodiment will be described.

FIG. 130 is a flowchart showing the operation of the local complexity creating section for images according to the twenty-fourth embodiment of the present invention. When the input image 10 is input (Step 24310), the image / local complexity creating unit 3200 creates an edge / texture component value image 3201 from the input image 10 in the edge / texture component value image creating unit 3210 (Step 24310). 24320). Next, the texture degree index creation unit 3220 creates a texture degree index 3202 from the edge / texture component value image 3201, and the upsampler 3230 upsamples the texture degree index 3202 (step 2433).
0), the same size as the input image 10 is output as the local complexity of the image (step 24340).

FIG. 131 is a flow chart showing the operation of the texture index generation unit according to the twenty-fourth embodiment of the present invention. Steps 24331 to 24333) The texture index generation unit 3220 performs the same processing as the texture index generation unit 3220 described in the twenty-third embodiment.
Find the texture index index _pq .

[0373] Step 24334) The texture degree index creating unit 3220 determines the obtained texture degree index index.
A value obtained by mapping _pq using a predetermined function f (x) is newly used as a texture degree index. When this is represented by an expression, index _pq : = f (index _pq ) (0 ≦ p <P, 0 ≦ q <Q).

According to the twenty-fourth embodiment, by mapping the texture degree index 3202 obtained in the twenty-third embodiment by a function, it is possible to obtain a value more consistent with visual characteristics. It is. [Twenty-fifth Embodiment] Next, a digital watermark embedding apparatus according to a twenty-fifth embodiment of the present invention will be described.

FIG. 132 is a configuration diagram of an image local complexity creating section according to the twenty-fifth embodiment of the present invention. The image local complexity creating unit 3400 shown in FIG. 18 includes an edge / texture component value image creating unit 3410, a texture degree index creating unit 3420, and a texture component value image creating unit 3430. Texture component value image creation unit 3430
Creates and outputs a texture component value image 3403 from the edge / texture component value image 3401 and the texture degree index 3402.

FIG. 133 is a flowchart showing the operation of the local complexity creating section for an image according to the twenty-fifth embodiment of the present invention. When the input image 10 is input (step 25410), the image local complexity creating unit 3400 creates the edge / texture component value image 3401 from the input image 10 (step 25410). 25420).

Next, a texture degree index creating unit 3420
Creates the texture degree index 3402 from the edge / texture component value image 3401 (the creation method may be any of the twenty-third embodiment and the twenty-fourth embodiment) (step 2)
5430). Then, the texture component value image creation unit 34
30 is a texture component value image 34 based on the edge / texture component value image 3401 and the texture degree index 3402.
03 (step 25440), and outputs this as the local complexity of the image (step 25450).

Next, the texture component value image creating section 3420 will be described. FIG. 134 shows the second embodiment of the present invention.
15 is a flowchart illustrating the operation of a texture component value image creation unit according to the fifth embodiment. Step 25441) Texture component value image creation unit 3
Reference numeral 430 denotes the edge / texture component value image 3401 as n
It is subdivided into blocks of × n pixels (assuming that the edge / texture component value image is composed of P × Q blocks). Step 254 for each block
Perform 42.

[0379] Step 25442) The (p, q) -th block (0
The processing for ≦ p <P, 0 ≦ q <Q) will be described. The texture component value image creation unit 3430 calculates the {b _ij ^(pq) }:
Assuming that block pixel values (0 ≦ i <n, 0 ≦ j <n) of the (p, q) th edge-texture component value image,
(P, q) th texture component value block {b ' _ij
^(pq) ｝, (0 ≦ i <n, 0 ≦ j <n) is obtained by the following equation.

B ′ _ij ^(pq) : = index _pq · b _ij ^(pq) Step 25443) After all the blocks have been processed, the texture component value block is made the same as the corresponding edge / texture component value image 3401. The image constituted by arranging at the position is referred to as a texture component value image 340.
Output as 3.

The image processing in this embodiment is described with reference to FIG.
This will be described with reference to FIG. The input image 10 shown in FIG. 1 includes a background (uniform pixel values), the sun (having edges, the inside is uniform), a trunk (having a weak texture), and leaves (having a strong texture). In the edge / texture component value image 3401 (brighter values are larger), edges and textures are treated similarly. There is no significant difference in the edge texture component values between the stem portion, the leaf portion, and the edge portion of the sun. This is because a difference counter is used.

The value of the texture degree index 3402 is small in a block having a small number of edge / texture components due to the way of making the index. Therefore, the texture degree index is small in the sun portion having only the edge. Conversely, in the leaf and stem portions, the edge / texture component value in the block is large, so that the texture degree index 3402 has a large value. In the texture component value image 3403 (the brighter the value, the larger the value), the edge is weak and the texture is strong, which matches the purpose. In addition, the block distortion is improved as compared with the output of the twenty-third embodiment, and it is possible to generate a more accurate local complexity of an image.

[Twenty-Sixth Embodiment] Next, a digital watermark embedding apparatus according to a twenty-sixth embodiment will be described. FIG.
FIG. 6 shows the configuration of a digital watermark embedding device according to a twenty-sixth embodiment of the present invention. Digital watermark embedding device 3500 shown in FIG.
Is composed of an image local complexity creating unit 3510, a watermark pattern creating unit 3520, and an image adding unit 3530.

The digital watermark embedding device 3500 receives an input image 51, a digital watermark 52, key information 53, and an intensity parameter. FIG. 137 is a flowchart showing the operation of the digital watermark embedding device according to the twenty-sixth embodiment of the present invention. Step 26510) First, the image local complexity creating unit 3510 creates the image local complexity (e _ij ) from the input image 51 (p _ij ). The local complexity of the image is
Ideally, the value indicates the amount of change of the target pixel value when the visual stimulus caused by the change of the pixel value is equal to or less than a certain value. For example, any of the digital watermark embedding apparatuses described in the twenty-second to twenty-fifth embodiments is used.

Step 26520) Create watermark pattern
The digital watermark 52 and the key information 53 are
Original watermark pattern 3502w_ijCreate Step 26530) In the image adding unit 3530,
The following formula is used to adaptively adjust the watermark pattern according to the complexity of the image.
The turn 3502 is added to the input image 51. p '_ij: = P_ij + Power ・ e_ij・ W_ij However, 'p'_ij｝: Digital watermark embedded image (0 ≦ i <Widt
h, 0 ≦ j <Height) ｛p_ij｝: Input image (0 ≦ i <Width, 0 ≦ j <Heigh
t) ｛w_ij｝: Watermark pattern (0 ≦ i <Width, 0 ≦ j <
Height) power: intensity parameter Step 26540) The image adding unit 3530 determines whether
The embedded image p _ijIs output.

Next, creation of the watermark pattern 3502 in the above step 26530 will be described. An N × N size complex matrix F (u, v) is prepared, and the initial value is set to a zero matrix. The k-th embedding coefficient position information L _k = (x _k ,
y _k , z _k ) (generated from key information)
Change the element of (u, v).

・ Z_kIs a value representing a real component: F
(X_k, Y_kM)_k(Key information and digital watermark
Generated from). In addition, the Fourier transform coefficient
F (N−x_k, N-y _kReal number
M_kAdd. In the formula, F (x_k, Y_k): = F (x_k, Y_k) + M_k・ Α F (N-x_k, N-y_k): = F (N-x_k, N-
y_k) + M_k・ Α However, α is a parameter that gives the strength of the pattern of the basic watermark turn.
Meter.

When z _k is a value representing an imaginary component: F
(X _k, y _k) is added m _k to the imaginary component of the. Also, F (N−x _k , N
Subtracting m _k from the imaginary component of -y _k). To write the _{formula, F (x k, y k} ): = F (x k, y k) + m k · α · i F (N-x k, N-y k): = F (N-x k, N-
y _k ) −m _k · α · i (i is an imaginary unit) This is sequentially performed for k = 0... n−1, and the obtained coefficient matrix F (u, v) is subjected to inverse discrete Fourier transform to obtain a basic watermark. Get the pattern.

According to the twenty-sixth embodiment, the watermark pattern is weighted according to the local complexity of the image and added to the input image to obtain the image with the digital watermark embedded therein.
This makes it possible to suppress visual degradation in image quality due to embedding, and improves the same level of resistance to image degradation as compared to the conventional technology. [Twenty-Seventh Embodiment] Next, a digital watermark embedding apparatus according to a twenty-seventh embodiment of the present invention will be described.

FIG. 138 shows the structure of a digital watermark embedding device according to the twenty-seventh embodiment of the present invention. The digital watermark embedding device 3600 shown in the figure includes a block dividing unit 3610, an image local complexity creating unit 3620, an adaptive watermark pattern creating unit 3630, a basic watermark pattern creating unit 3640,
It comprises an image adder 3650. The digital watermark embedding device 3600 includes an input image 61, a digital watermark 6
2. Input key information 63 and strength parameter 64.

FIG. 139 is a flowchart showing the operation of the digital watermark embedding device according to the twenty-seventh embodiment of the present invention. Step 27610) The block dividing unit 3610 converts the input image 61 (consisting of Width × Height pixels) into N × N
It is divided into size blocks 3601. Step 27620) Using the divided block 3601, the image local complexity creating unit 3620 uses
Create the local complexity of a block-sized image. As in the case of the twenty-fifth embodiment, the local complexity of the image ideally represents the amount of change of the symmetric pixel value when the visual stimulus due to the change of the pixel value is kept below a certain level. . For example, any of the digital watermark embedding devices described in the twenty-second to twenty-fifth embodiments is used.

(Step 27630) In the basic watermark pattern creating section 3640, the digital watermark 62 and the key information 6
3 is created based on the basic watermark pattern 3604. It is assumed that the basic watermark pattern 3604 is the same as that described as the “watermark pattern” in the twenty-sixth embodiment. This is specifically expressed as follows. An N × N size complex matrix F (u, v) is prepared, and the initial value is set to a zero matrix.

The k-th embedding coefficient position information L_k=
(X_k, Y_k, Z_k) (Generated from key information)
To change the element of F (u, v).・ Z_kIs a value representing a real component: F (x_k, Y_k) Fruit
M for several components_k(Generated from key information and digital watermark)
Add. Also, keep the Fourier transform coefficient symmetric.
And F (N-x_k, N-y _kM)_kAdd
I can. In the formula, F (x_k, Y_k): = F (x_k, Y_k) + M_k・ Α F (N-x_k, N-y_k): = F (N-x_k, N-
y_k) + M_k・ Α However, α is a parameter that gives the strength of the pattern of the basic watermark turn.
Meter.

When z _k is a value representing an imaginary component: F
(X _k, y _k) is added m _k to the imaginary component of the. Also, F (N−x _k , N
Subtracting m _k from the imaginary component of -y _k). To write the _{formula, F (x k, y k} ): = F (x k, y k) + m k · α · i F (N-x k, N-y k): = F (N-x k, N-
y _k ) −m _k · α · i (i is an imaginary unit) This is sequentially performed for k = 0... n−1, and the obtained coefficient matrix F (u, v) is subjected to inverse discrete Fourier transform to obtain a basic watermark. Get the pattern.

(Step 27650) Further, adaptive watermark pattern creating section 3630 selects adaptive watermark pattern 36
02 is created. The method of obtaining the adaptive watermark pattern 3602 is represented by the following equation.

[Equation 27]

Here, ｛p ′ _ij :｝: digital watermark embedded block (0 ≦ i
<N, 0 ≦ j <N) ｛p _ij :｝: block (0 ≦ i <N, 0 ≦ j <N) ｛e _ij :｝: local complexity of image (0 ≦ i <N, 0 ≦ j
<N) ｛a _ij :｝: Adaptive watermark pattern (0 ≦ i <N, 0 ≦ j
<N) {b _ij :}: Basic watermark pattern (0 ≦ i <N, 0 ≦ j
<N) power: Intensity parameter e _max : Maximum value of the range according to the definition formula of the local complexity of the image e _min : Minimum value of the range according to the definition formula of the local complexity of the image β: Enlargement magnification of the adaptive watermark pattern 27660) After creating the digital watermark embedded block 3605 for all the blocks, the digital watermark embedded block 3605 is arranged at the position of the original block to obtain the digital watermark embedded image 65.

Next, the adaptive watermark pattern creating section 3630 in this embodiment will be described. The adaptive watermark pattern creation unit 3630 emphasizes the orthogonal transform coefficient corresponding to the basic watermark pattern 3604 according to the amplitude value of the orthogonal transform coefficient of the block currently being referenced.

FIG. 140 is a flowchart showing the operation of the adaptive watermark pattern creating section according to the twenty-seventh embodiment of the present invention. Step 27641) Applicable watermark pattern creation unit 36
30 performs a discrete Fourier transform on the blocks divided by the block dividing unit 3610 to obtain a block coefficient matrix G (u,
Obtain v).

Step 27642) The basic watermark pattern is subjected to a discrete Fourier transform to generate a watermark coefficient matrix F (u, v) (0 ≦ u <N, 0 ≦ v <N) which is a complex matrix. Step 27643) (u, v) = (0,0) to (N
For −1, N−1), if | G (u, v) | <α, the process proceeds to step 27644; otherwise, the process proceeds to step 27645.

Step 27644) | G (u, v) |
When <α, γ: = α, and the flow shifts to step 27646. If | G (u, v) | ≧ α, γ: = | G (u, v) |. Step 2
Α in 7644 and 27645 is a threshold value set in advance.

Step 27646) F '(u, v):
= F (u, v) · γ Step 27647) F ′ (u, v) is subjected to discrete Fourier transform to obtain and output an adaptive watermark pattern 3602. As described above, according to the present embodiment, as shown in FIG. 141, it is possible to embed a digital watermark using a watermark pattern that well reflects a local pattern of an input image. That is, the pattern of the original block 3601 (the oblique line in the left front sector) and the basic watermark pattern 3604 have no correlation, but the adaptive watermark pattern 3602 reflects the pattern of the block 3601 well. When weighted and added using the local complexity of these images, information can be embedded with strong power in a form that reflects the original pattern of block 3601 well. As a result, the logarithm is further improved as compared with the twenty-fifth embodiment.

Further, the configuration and operation of the above embodiment are constructed as a program, and a disk device connected to a computer used as a digital watermark embedding device and a digital watermark detecting device, a floppy disk, a CD-ROM, etc. The present invention can be easily realized by storing it in a portable storage medium and installing it when implementing the present invention.

[Twenty-eighth Embodiment] In this embodiment, the digital watermark embedding device and the digital watermark detection device described in the first to twenty-seventh embodiments are integrated circuits as shown in FIG. It can also be realized by The integrated circuit shown in FIG. 142 includes a memory unit 1, a microprocessor unit 2, and an interface unit 3 that manages an interface with the outside. FIG. 1 shows the main part, and may include other circuits. The microprocessor unit 2 executes the program stored in the memory unit 1. Various other configurations are possible for the configuration of the integrated circuit.

Further, by incorporating this integrated circuit into various devices, for example, a video camera or a reproducing device, it is possible to perform the digital watermarking process of the present invention. Also,
A disk device connected to a computer used as the digital watermark embedding device and the digital watermark detection device, and a portable storage medium such as a floppy disk or a CD-ROM, by constructing the configuration and operation in the above embodiment as a program. The present invention can be easily realized by storing the information in a storage device and installing it when implementing the present invention. It should be noted that the present invention is not limited to the above-described embodiment, but can be variously modified and applied within the scope of the claims.

[0405]

As described above, according to the present invention, sub-information is embedded by adding a watermark pattern in a tile form over the entire input image. Since it is not necessary to divide the input image into blocks and perform discrete Fourier transform and discrete inverse Fourier transform for each block, the embedding process can be sped up as compared with the conventional technology.

At the time of detection, by finding the start position of the watermark pattern, it is possible to realize a digital watermarking method which is resistant to arbitrary translation. Also, if the cutout is a certain size (about the block size) or more,
The embedded information can be read from an image obtained by cutting out a part of the digital watermark embedded image.

Further, it is possible to divide the detection target image into blocks having the same size as the watermark pattern, and to improve the detection accuracy by using the detection results from all the blocks. Further, by using the detection from the added block obtained by adding all the blocks, it is possible to improve the detection accuracy and speed up the detection process. In the eleventh to sixteenth embodiments, the discrete Fourier transform only needs to be performed once before searching for the start position of the watermark pattern, so that the search time can be greatly reduced.

Further, as described in the nineteenth to twenty-first embodiments, when the detection result is significant using a predetermined threshold value, the detection processing is performed normally, and the detection result is significant. If it is not, the detection failure is improved, and the reliability of the detection result is improved. In particular, the configuration is such that the problem of how to set the threshold when the accuracy is specifically determined can be mathematically shown, and is highly practical.

In the present invention, since information is embedded by changing the Fourier transform coefficient value, it is possible to simultaneously withstand the conventional parallel movement and cutoff. Also, an orthogonal transform other than the discrete Fourier transform may be used.
Furthermore, by applying the present invention to each frame of a moving image, it is possible to embed a digital watermark in the moving image and to detect a digital watermark from the moving image.

Further, according to the present invention, it is possible to create local complexity of an image having a relatively large value in a texture region and a relatively small value in a flat region or an edge portion. Become. As a result, digital watermark embedding adaptive to visual characteristics can be realized as compared with the related art, and the resistance of the digital watermark is relatively increased. In addition, since information is embedded in the image by mainly changing a complicated part of the image, it is possible to relatively improve the tolerance compared with the conventional technology.

Further, at the time of embedding, since a pattern reflecting the pattern of the original image is added, it is possible to further improve the image quality, that is, further improve the relative tolerance. Further, as the information detecting device, a conventional device can be used as it is, and the transition cost can be reduced. Furthermore, by applying the present invention to each frame of a moving image, it is possible to embed a digital watermark in the moving image and to detect a digital watermark from the moving image.

[0412] Thus, since the purpose of the copyright holder is to protect the copyright of the information content, the digital watermark information needs to be resistant to various editing by the user. Therefore, it is necessary to be able to use the information content without being conscious of the digital watermark. Therefore, the degradation must not be perceived in the image subjected to the digital watermark processing. According to the present invention, the image quality /
By improving both the resilience and the copyright protection system,
It functions more effectively and allows the author to provide information on the content with peace of mind, thereby promoting information distribution through a network.

[0413] The present invention relates to a content distribution system for uniquely identifying a content.
By applying D to the content management system by embedding it in the content and linking it to the copyright owner information of the content and usage conditions, correct use of the content is promoted. You can be assured that the content right holder is yourself.

[0414] Further, the present invention can suppress unauthorized copying by embedding user information at the time of content delivery. Further, by embedding a copy management signal used in a DVD or the like, stronger copy protection can be realized.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining a first principle of the present invention.

FIG. 2 is a first principle configuration diagram of the present invention.

FIG. 3 is a diagram for explaining a second principle of the present invention.

FIG. 4 is a third principle configuration diagram of the present invention.

FIG. 5 is a configuration diagram of a digital watermark embedding device according to the first embodiment of the present invention.

FIG. 6 is a flowchart showing the operation of the digital watermark embedding device according to the first embodiment of the present invention.

FIG. 7 is a configuration diagram of an embedding target component designation section according to the first embodiment of the present invention.

FIG. 8 is a flowchart illustrating an operation of an embedding target component specifying unit according to the first embodiment of this invention.

FIG. 9 is a configuration diagram of an embedded sequence generation unit according to the first example of the present invention.

FIG. 10 is a flowchart illustrating an operation of an embedded sequence generation unit according to the first exemplary embodiment of the present invention.

FIG. 11 is a flowchart illustrating an operation of a component value changing unit according to the first embodiment of this invention.

FIG. 12 is a diagram illustrating the operation of the discrete inverse Fourier transform unit according to the first embodiment of the present invention.

FIG. 13 is a diagram for explaining the operation of the watermark pattern adding unit according to the first embodiment of the present invention.

FIG. 14 is a configuration diagram of a digital watermark detection device according to a second embodiment of the present invention.

FIG. 15 is a flowchart showing an operation of the digital watermark detection device according to the second embodiment of the present invention.

FIG. 16 is a configuration diagram of a detection target component designation unit according to the second embodiment of the present invention.

FIG. 17 is a flowchart illustrating an operation of a detection target component designation unit according to the second embodiment of this invention.

FIG. 18 is a configuration diagram of a position marker sequence generation unit according to a second example of the present invention.

FIG. 19 is a flowchart illustrating an operation of a position marker sequence generation unit according to the second example of the present invention.

FIG. 20 is a configuration diagram of a position marker detection unit according to a second example of the present invention.

FIG. 21 is a flowchart illustrating an operation of a position marker detection unit according to the second example of the present invention.

FIG. 22 is a diagram for explaining an operation of the block generation unit according to the second example of the present invention.

FIG. 23 is a flowchart illustrating an operation of a block generation unit according to the second embodiment of this invention.

FIG. 24 is a flowchart illustrating an operation of a detection target sequence generation unit according to the second embodiment of this invention.

FIG. 25 is a configuration diagram of a digital watermark detection unit according to a second example of the present invention.

FIG. 26 is a flowchart illustrating an operation of a digital watermark detection unit according to the second embodiment of this invention.

FIG. 27 is a configuration diagram of a symbol detection unit according to a second embodiment of the present invention.

FIG. 28 is a flowchart illustrating an operation of the symbol detection unit according to the second example of the present invention.

FIG. 29 is a configuration diagram of a symbol sequence generation unit according to the second embodiment of the present invention.

FIG. 30 is a flowchart illustrating an operation of the symbol sequence generation unit according to the second example of the present invention.

FIG. 31 is a configuration diagram of a detection result generation unit according to the second example of the present invention.

FIG. 32 is a flowchart illustrating an operation of a detection result generation unit according to the second example of the present invention.

FIG. 33 is a flowchart illustrating an operation of a block generation unit according to the third embodiment of this invention.

FIG. 34 is a diagram for explaining the operation of the block generation unit according to the third example of the present invention.

FIG. 35 is a diagram for explaining the operation of the block generation unit according to the fourth embodiment of the present invention.

FIG. 36 is a flowchart illustrating an operation of a block generation unit of a position marker detection unit according to the fourth embodiment of the present invention.

FIG. 37 is a configuration diagram of a digital watermark detection unit according to a fifth embodiment of the present invention.

FIG. 38 is a flowchart illustrating an operation of the digital watermark detection unit according to the fifth embodiment of the present invention.

FIG. 39 is a diagram for explaining the operation of the block division unit according to the fifth embodiment of the present invention.

FIG. 40 is a configuration diagram of a symbol detection unit according to a fifth embodiment of the present invention.

FIG. 41 is a flowchart showing the operation of the symbol detector of the fifth embodiment of the present invention.

FIG. 42 is a configuration diagram of a symbol sequence generation unit according to a fifth embodiment of the present invention.

FIG. 43 is a flowchart illustrating an operation of the symbol sequence generation unit according to the fifth example of the present invention.

FIG. 44 is a configuration diagram of a detection result generation unit according to the fifth embodiment of the present invention.

FIG. 45 is a flowchart illustrating an operation of a detection result generation unit according to the fifth example of the present invention.

FIG. 46 is a configuration diagram of a digital watermark detection unit according to a sixth embodiment of the present invention.

FIG. 47 is a flowchart showing the operation of the digital watermark detection unit according to the sixth embodiment of the present invention.

FIG. 48 is a diagram for explaining the operation of the block division unit according to the sixth embodiment of the present invention.

FIG. 49 is a configuration diagram of a symbol detection unit according to a sixth embodiment of the present invention.

FIG. 50 is a flowchart showing the operation of the symbol detector of the sixth embodiment of the present invention.

FIG. 51 is a configuration diagram of a symbol sequence generation unit according to a sixth embodiment of the present invention.

FIG. 52 is a flowchart illustrating an operation of the symbol sequence generation unit according to the sixth example of the present invention.

FIG. 53 is a configuration diagram of a detection result generator according to the sixth embodiment of the present invention.

FIG. 54 is a flowchart illustrating an operation of a detection result generation unit according to the sixth example of the present invention.

FIG. 55 is a configuration diagram of a digital watermark detection unit according to a seventh embodiment of the present invention.

FIG. 56 is a flowchart showing the operation of the digital watermark detection unit according to the seventh embodiment of the present invention.

FIG. 57 is a diagram illustrating a block division unit according to a seventh embodiment of the present invention.

FIG. 58 is a configuration diagram of a symbol detection unit according to a seventh embodiment of the present invention.

FIG. 59 is a flowchart showing the operation of the symbol detector of the seventh embodiment of the present invention.

FIG. 60 is a configuration diagram of a symbol sequence generation unit according to a seventh embodiment of the present invention.

FIG. 61 is a flowchart illustrating an operation of the symbol sequence generation unit according to the seventh example of the present invention.

FIG. 62 is a configuration diagram of a detection result generator according to a seventh embodiment of the present invention.

FIG. 63 is a flowchart illustrating an operation of a detection result generation unit according to the seventh example of the present invention.

FIG. 64 is a configuration diagram of a block generation unit of a position marker detection unit according to an eighth embodiment of the present invention.

FIG. 65 is a flowchart showing the operation of the block generation unit of the position marker detection unit according to the eighth embodiment of the present invention.

FIG. 66 is a diagram for explaining the operation of the block generation unit according to the eighth embodiment of the present invention.

FIG. 67 is a configuration diagram of a block generation unit of a position marker detection unit according to a ninth embodiment of the present invention.

FIG. 68 is a flowchart illustrating an operation of a block generation unit of the position marker detection unit according to the ninth embodiment of the present invention.

FIG. 69 is a diagram for describing block division by the block generation unit according to the ninth embodiment of the present invention.

FIG. 70 is a diagram for explaining a cyclic shift operation of the block generation unit according to the ninth embodiment of the present invention.

FIG. 71 is a configuration diagram of a block generation unit of a position marker detection unit according to a tenth embodiment of the present invention.

FIG. 72 is a flowchart illustrating an operation of a block generation unit according to the tenth embodiment of the present invention.

FIG. 73 is a diagram for explaining the operation of the block generation unit according to the tenth embodiment of the present invention.

FIG. 74 is a diagram for describing division processing of the block generation unit according to the tenth embodiment of the present invention.

FIG. 75 is a view for explaining the operation of the cyclic shift section according to the tenth embodiment of the present invention.

FIG. 76 is a configuration diagram of a digital watermark detection device according to an eleventh embodiment of the present invention.

FIG. 77 is a flowchart showing the operation of the digital watermark detection device according to the eleventh embodiment of the present invention.

FIG. 78 is a configuration diagram of a detection target component designation unit according to the eleventh embodiment of the present invention.

FIG. 79 is a flow chart showing the operation of the detection target component designation unit according to the eleventh embodiment of the present invention.

FIG. 80 is a configuration diagram of a position marker sequence generation unit according to the eleventh embodiment of the present invention.

FIG. 81 is a flowchart showing the operation of the position marker sequence generation unit according to the eleventh embodiment of the present invention.

FIG. 82 is a configuration diagram of a position marker detection unit according to the eleventh embodiment of the present invention.

FIG. 83 is a flowchart showing the operation of the position marker detection unit according to the eleventh embodiment of the present invention.

FIG. 84 is a flowchart showing the operation of the detection target sequence generation unit of the position marker detection unit according to the eleventh embodiment of the present invention.

FIG. 85 is a configuration diagram of a digital watermark detection unit according to an eleventh embodiment of the present invention.

FIG. 86 is a flowchart showing the operation of the digital watermark detection unit according to the eleventh embodiment of the present invention.

FIG. 87 is a configuration diagram of a symbol detection unit according to an eleventh embodiment of the present invention.

FIG. 88 is a flowchart showing the operation of the symbol detector of the eleventh embodiment of the present invention.

FIG. 89 is a configuration diagram of a symbol generation unit according to an eleventh embodiment of the present invention.

FIG. 90 is a flowchart showing the operation of the symbol generation unit according to the eleventh embodiment of the present invention.

FIG. 91 is a configuration diagram of a detection result generation unit according to the eleventh embodiment of the present invention.

FIG. 92 is a flowchart showing the operation of the detection result generation unit according to the eleventh embodiment of the present invention.

FIG. 93 is a diagram for explaining the operation of the position marker detection unit according to the twelfth embodiment of the present invention.

FIG. 94 is a configuration diagram of a digital watermark detection unit according to a thirteenth embodiment of the present invention.

FIG. 95 is a flowchart showing the operation of the digital watermark detection unit according to the thirteenth embodiment of the present invention.

FIG. 96 is a diagram for explaining the operation of the block division unit according to the thirteenth embodiment of the present invention.

FIG. 97 is a configuration diagram of a symbol detector of a thirteenth embodiment of the present invention.

FIG. 98 is a flowchart showing the operation of the symbol detector of the thirteenth embodiment of the present invention.

FIG. 99 is a configuration diagram of a symbol sequence generation unit according to a thirteenth embodiment of the present invention.

FIG. 100 is a flowchart showing an operation of the symbol sequence generation unit according to the thirteenth embodiment of the present invention.

FIG. 101 is a configuration diagram of a detection result generation unit according to a thirteenth embodiment of the present invention.

FIG. 102 is a flowchart showing the operation of the detection result generation unit according to the thirteenth embodiment of the present invention.

FIG. 103 is a diagram for explaining the operation of the block division unit of the digital watermark detection unit according to the fourteenth embodiment of the present invention.

FIG. 104 is a configuration diagram of a block generation unit of a position marker detection unit according to a fifteenth embodiment of the present invention.

FIG. 105 is a flowchart illustrating an operation of a block generation unit of the position marker detection unit according to the fifteenth embodiment of the present invention.

FIG. 106 is a diagram for explaining the operation of the block generation unit according to the fifteenth embodiment of the present invention.

FIG. 107 is a configuration diagram of a block generation unit of a position marker detection unit according to a sixteenth embodiment of the present invention.

FIG. 108 is a flowchart illustrating an operation of a block generation unit of the position marker detection unit according to the sixteenth embodiment of the present invention.

FIG. 109 is a diagram for explaining the operation of the block generation unit according to the sixteenth embodiment of the present invention.

FIG. 110 is a flowchart showing the operation of the offset information generation unit of the position marker detection unit according to the nineteenth embodiment of the present invention.

FIG. 111 is a flowchart showing the operation of the detection result generation unit of the digital watermark detection unit according to the twentieth embodiment of the present invention.

FIG. 112 is a graph showing a value of corr and a theoretical value by computer simulation according to a twenty-first embodiment of the present invention.

FIG. 113 is a graph of symbol detection information by computer simulation according to the twenty-first embodiment of the present invention.

FIG. 114 PSN when embedding strength of the present invention is changed
It is a figure showing R.

FIG. 115 is a diagram showing a result of detecting watermark information from a watermarked image by computer simulation of the present invention.

FIG. 116 is a diagram showing detection results from partial images by computer simulation according to the present invention.

FIG. 117 is a diagram showing detection failure and success partial images by computer simulation of the present invention.

FIG. 118 shows J obtained by computer simulation according to the present invention.
FIG. 7 is a diagram illustrating a detection result from a PEG compressed image.

FIG. 119 is a diagram showing a gradation conversion image (binary image) by computer simulation of the present invention.

FIG. 120 is an example of a composite image of a computer simulation of the present invention.

FIG. 121 is a configuration diagram of an image local complexity creating unit according to a twenty-second embodiment of the present invention.

FIG. 122 is a flowchart showing an operation of an image local complexity creating unit according to the twenty-second embodiment of the present invention.

FIG. 123 is a flowchart showing the operation of an edge / texture component value image creating unit according to the twenty-second embodiment of the present invention.

FIG. 124 is an example of an input image and an edge / texture component value image according to the twenty-second embodiment of the present invention.

FIG. 125 is a configuration diagram of an image local complexity creating unit according to a twenty-third embodiment of the present invention.

FIG. 126 is a flowchart showing an operation of an image local complexity creating unit according to the twenty-third embodiment of the present invention.

FIG. 127 is a flowchart showing the operation of the texture index generation unit according to the twenty-third embodiment of the present invention.

FIG. 128 is a view for explaining the operation of the upsampler according to the twenty-third embodiment of the present invention.

FIG. 129 is a diagram showing an edge / texture component value image and a texture degree index according to the twenty-third embodiment of the present invention.

FIG. 130 is a flowchart showing the operation of the local complexity creating unit for images according to the twenty-fourth embodiment of the present invention;

FIG. 131 is a flowchart showing an operation of the texture degree index creating unit according to the twenty-fourth embodiment of the present invention.

FIG. 132 is a configuration diagram of an image local complexity creating unit according to a twenty-fifth embodiment of the present invention.

FIG. 133 is a flowchart showing the operation of an image local complexity creating unit according to the fourth embodiment of this invention.

FIG. 134 is a flowchart illustrating an operation of the texture component value image creating unit according to the twenty-fifth embodiment of the present invention.

FIG. 135 is a diagram illustrating input / output images at each unit according to the twenty-fifth embodiment of the present invention.

FIG. 136 is a configuration diagram of a digital watermark embedding device according to a twenty-sixth embodiment of the present invention.

FIG. 137 is a flowchart showing the operation of the digital watermark embedding device according to the twenty-sixth embodiment of the present invention.

FIG. 138 is a configuration diagram of a digital watermark embedding device according to a twenty-seventh embodiment of the present invention.

FIG. 139 is a flowchart showing the operation of the digital watermark embedding device according to the twenty-seventh embodiment of the present invention.

FIG. 140 is a flowchart showing the operation of the adaptive watermark pattern creating unit according to the twenty-seventh embodiment of the present invention.

FIG. 141 is a diagram showing blocks and patterns in a digital watermark embedding process according to a twenty-seventh embodiment of the present invention.

FIG. 142 is a configuration diagram of an integrated circuit according to a twenty-eighth embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 memory 2 microprocessor 3 interface unit 1A changing unit 2A watermark pattern generating unit 3A adding unit 4A embedded image generating unit 5A detection target component specifying unit 6A position marker sequence generating unit 7A position marker detecting unit 8A electronic watermark detecting unit 1B local Complexity creating means 2B Watermark pattern creating means 3B Image adding means 10 Input image 20 Local complexity of image 51 Input image 52 Digital watermark 53 Key information 54 Strength parameter 55 Digital watermark embedded image 61 Input image 62 Digital watermark 63 key Information 64 Intensity parameter 65 Digital watermark embedded image 100 Information embedding device 101 Input image 102 Intensity parameter 103 Digital watermark 104 Key information 105 Embedded image 110 Watermark pattern adding unit 120 Embedded sequence generation unit 121 Embedded system 122 Diffusion sequence generator 130 Embedding target component specifying unit 131 Embedding target component position information 132 Random number generator 140 Component value changing unit 141 Watermark coefficient matrix 150 Discrete inverse Fourier transform unit 151 Watermark pattern 200 Digital watermark detection device 201 Detection target image 202 Key information 203 Detection result 204 Offset candidate information 205 Pixel block 206 Detection target coefficient matrix 207 Detection target sequence 208 Position marker detection information 210 Detection target component designation unit 211 Detection target component position information 212 Random number generator 220 Position marker sequence generation unit 221 Position marker sequence 222 Spread sequence generator 230 Position marker detector 231 Offset information 232 Offset candidate information generator 233 Block generator 234 Discrete Fourier transform 235 Detection target sequence generator 236 Marker detection information generation unit 237 Offset information generation unit 240 Digital watermark detection unit 241 Block generation unit 242 Discrete Fourier transform unit 243 Detection target sequence generation unit 244 Symbol detection unit 245 Detection result generation unit 246 Pixel block 247 Detection target coefficient matrix 248 Detection target Sequence 249 Symbol detection information 251 Detection result 500 Digital watermark detection unit 510 Block division unit 511 Pixel block 520 Discrete Fourier transform unit 521 Detection target sequence matrix 530 Detection target sequence generation unit 531 Detection target sequence 540 Symbol detection unit 541 Symbol candidate generation unit 542 Symbol sequence generation unit 543 Symbol detection information generation unit 544 Symbol candidate 545 Symbol sequence 546 Symbol detection information 550 Detection result generation unit 551 Sequence generation unit for each symbol position 55 2 Detected symbol generator 553 Inverted symbol converter 554 Sequence for each symbol position 555 Detected symbol 556 Detection result 600 Digital watermark detector 601 Pixel block 602 Detected sequence matrix 603 Detected sequence 604 Symbol detection information 610 Block dividing unit 620 Discrete Fourier Conversion section 630 Detection target sequence generation section 640 Symbol detection section 641 Symbol candidate generation section 642 Symbol candidate 643 Symbol sequence generation section 644 Symbol sequence 645 Symbol detection information generation section 646 Symbol detection information 650 Detection result generation section 651 Sequence generation for each symbol position Unit 652 sequence for each symbol position 653 detection symbol generation unit 654 detection symbol 655 inverse symbol conversion unit 656 detection result 663 detection target sequence 700 digital watermark detection unit 701 Raw block 702 Detection target coefficient matrix 703 Detection target sequence 704 Symbol detection information 705 Detection result 710 Block division unit 720 Discrete Fourier transform unit 730 Detection target sequence generation unit 740 Symbol detection unit 741 Symbol candidate generation unit 742 Symbol candidate 743 Symbol sequence generation unit 744 Symbol sequence 745 Symbol detection information generator 746 Symbol detection information 750 Detection result generator 751 Each symbol position sequence generator 752 Each symbol position sequence 753 Detection symbol generator 754 Detection symbol 755 Inverse symbol converter 756 Detection result 800 blocks Generation unit 801 t blocks 802 addition block 803 pixel block 810 block division unit 820 block addition unit 900 block generation unit 901 t blocks 902 Addition block 903 Pixel block 910 Block division unit 920 Block addition unit 930 Cyclic shift unit 1000 Block generation unit 1010 Block division unit 1011 t blocks 1012 Addition block 1013 Pixel block 1020 Block addition unit 1030 Cyclic shift unit 1100 Digital watermark detection Device 1101 Offset information 1102 Position marker sequence 1103 Detection target component position information 1104 Pixel block 1105 Detection target coefficient matrix 1106 Offset candidate information 1107 Detection target sequence 1108 Position marker detection information 1110 Position marker detection unit 1111 Block generation unit 1113 Discrete Fourier transform unit 1115 Detection target sequence generation unit 1117 Offset candidate information generation unit 1118 Position marker detection information generation unit 1 19 offset information generation unit 1120 position marker sequence generation unit 1121 diffusion sequence generator 1130 detection target component specification unit 1131 random number generator 1140 digital watermark detection unit 1141 block generation unit 1142 pixel block 1143 discrete Fourier transform unit 1144 detection target coefficient matrix 1145 detection Target sequence generation unit 1146 Detection target sequence 1147 Symbol detection unit 1148 Symbol detection information 1149 Detection result generation unit 2000 Digital watermark detection unit 2001 Pixel block 2002 Detection target coefficient matrix 2003 Detection target sequence 2004 Symbol detection information 2100 Block division unit 2200 Discrete Fourier transform Unit 2300 detection target sequence generation unit 2400 symbol detection unit 2401 symbol candidate 2402 symbol sequence 2403 symbol detection information 2410 Symbol candidate generation unit 2420 Symbol sequence generation unit 2421 Spread sequence generator 2430 Symbol detection information generation unit 2500 Detection result generation unit 2501 Sequence for each symbol position 2502 Detection symbol 2510 Sequence generation unit for each symbol position 2520 Detection symbol generation unit 2530 Inverse symbol Conversion unit 2441 Symbol candidate generation unit 2442 Symbol sequence generation unit 2443 Symbol detection information generation unit 2444 Symbol candidate 2445 Symbol sequence 2446 Symbol detection information 2447 Spread sequence generator 2451 Sequence generation unit for each symbol position 2452 Sequence for each symbol position 2453 Inverse symbol Conversion section 2454 Sequence for each symbol position 2455 Detection symbol 3000 Block generation section 3001 t blocks 3002 addition block 3100 blocks Section 3200 block adding section 3400 image local complexity creating section 3401 edge / texture component value image 3402 texture degree index 3403 texture component value image 3410 edge / texture component value image creating section 3420 texture index index creating section 3430 texture component value image Creation unit 4000 Block generation unit 4001 t blocks 4002 addition block 4100 block division unit 4200 block addition unit 5421 spreading sequence generator 6431 spreading sequence generator 7431 spreading sequence generator 11491 symbol position sequence generation unit 11492 sequence for each symbol position 11493 Detected symbol generator 11494 Detected symbol 11495 Inverse symbol converter 11471 Symbol candidate generator 11472 Symbol candidate 11473 symbol Column generation unit 11474 Symbol sequence 11475 Symbol detection information generation unit 11476 Symbol detection information 114731 Spread sequence generator 3000 Block generation unit 3001 block 3002 Addition block 3010 Image local complexity generation unit 3020 Image local complexity generation unit 3100 block Division unit 3110 Edge / texture component value image creation unit 3111 Edge / texture component value image 3200 Block addition unit 3201 Edge / texture component value image 3202 Texture degree index 3210 Edge / texture component value image creation unit 3220 Texture degree index creation unit 3230 Up Sampler 3400 Image processing device 3401 Edge / texture component value image 3402 Texture degree index 3403 Texture component value image 3410 Edge Texture component value image creation unit 3420 Texture degree index creation unit 3430 Texture component value image creation unit 3500 Digital watermark embedding device 3501 Image local complexity 3502 Watermark pattern 3510 Image local complexity creation unit 3520 Watermark pattern creation unit 3530 Image adder 3600 Digital watermark embedding device 3601 Block 3602 Adaptive watermark pattern 3604 Basic watermark pattern 3605 Digital watermark embedded block 3610 Block divider 3620 Image local complexity generator 3630 Adaptive watermark pattern generator 3640 Basic watermark pattern generator Unit 3650 image adding unit

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Junki Tomioka 2-3-1 Otemachi, Chiyoda-ku, Tokyo Inside Nippon Telegraph and Telephone Corporation (72) Inventor Yoichi Takashima 2-3-1 Otemachi, Chiyoda-ku, Tokyo No.1 Nippon Telegraph and Telephone Corporation F-term (reference) 5B057 AA11 CA08 CA12 CA16 CB08 CB12 CB16 CC02 CE08 CH08 CH20 DB02 DB09 DC16 DC30 5C076 AA14 AA36 BA06 5J104 AA14 NA13 NA27 (54) [Title of Invention] Watermarking method, digital watermark detecting method, digital watermark embedding device, digital watermark detecting device, storage medium storing digital watermark embedding program, storage medium storing digital watermark detecting program, digital watermark system and integration circuit

Claims (176)

[Claims] 1. A digital watermark embedding method for embedding a digital watermark in a digital image so as not to be perceived by human perception, wherein when the digital watermark is embedded in a digital original image, key information and digital information are embedded in the digital image. A watermark component is generated by independently changing a real component and an imaginary component of each coefficient value of a watermark coefficient matrix, which is a complex matrix, from a digital watermark to be embedded, performing a discrete inverse Fourier transform on the changed watermark sequence matrix, and generating a watermark pattern. Is added to the original image in a tile form,
A digital watermark embedding method, which generates an embedded image. 2. Generating embedding target component position information which is a sequence of a length for embedding the digital watermark from the key information, and using the key information, spreading position marker information and the digital watermark to the embedding sequence; Using the embedding target component position information and the embedding sequence, change the component value of the watermark coefficient matrix whose initial value is a complex matrix of zero matrix, perform discrete inverse Fourier transform of the watermark coefficient matrix, and generate the watermark pattern. 2. The electronic watermark embedding method according to claim 1, wherein the watermark pattern is enhanced by an intensity parameter, and an embedded image obtained by adding the enhanced watermark pattern in a tiled manner to an input image is generated. 3. The electronic device according to claim 2, wherein, when generating the position information of the component to be embedded, a random number sequence having the key information as an initial value is generated, and the position information of the component to be embedded is generated from the random number sequence. Watermark embedding method. 4. When the position marker information and the digital watermark are spread into the embedded sequence, a 0th spread sequence with the key information as an initial value is generated, and the spread sequence represents position marker information. A spreading sequence that converts the digital watermark into a symbol expression, obtains the total number of symbols and each symbol value, and sets the key information plus the symbol value as an initial value Generating, as a sequence representing a symbol, an average value of the stored values in addition to the stored terms of the spread sequence, and subtracting the average value from each term as an embedded sequence. 2. The digital watermark embedding method according to 2. 5. The digital watermark embedding method according to claim 4, wherein, when generating the spread sequence, a code that takes one of +1 and −1 is used as the spread sequence. 6. When the position marker information and the digital watermark are spread into an embedded sequence, a spread sequence is generated using the key information as an initial value, and the spread sequence is multiplied by a predetermined parameter to store the spread sequence. The result obtained by converting the digital watermark into a symbol expression, adding a spreading sequence corresponding to the symbol of the symbol expression to each item of the storage unit, and subtracting an average value from each item of the storage unit 3. The digital watermark embedding method according to claim 2, wherein 7. The digital watermarking method according to claim 6, wherein, when generating the spread sequence, a code that takes one of +1 and −1 is used as the spread sequence. 8. A digital watermark detection method for detecting a digital watermark from a detection target image in which a digital image is embedded so as not to be perceived by human perception, wherein the digital watermark is embedded in the detection target image. When detecting the digital watermark, a block is cut out from an arbitrary position in the detection target image, the block is subjected to discrete Fourier transform, a coefficient matrix is obtained, and detection target component position information specified by key information is generated. Then, each term of the detection target component position information is multiplied by the phase difference of the coefficient according to the amount of parallel movement to generate a position marker sequence, and the start point of the embedded watermark pattern and the pixel block cut out from the detection target image. The offset information, which is the amount of parallel movement when the start point matches, is extracted, and the pixel cut out based on the offset information is extracted. A digital watermark detection method, wherein the digital watermark embedded from a block is detected. 9. When generating the detection target component position information from the key information, generating a random number sequence with the key information as an initial value, and generating the detection target component position information from the random number sequence. 9. The digital watermark detection method according to item 9. 10. The digital watermark according to claim 8, wherein, when generating the position marker sequence, a 0th spreading sequence with the key information as an initial value is generated, and the spreading sequence is output as the position marker sequence. Detection method. 11. When extracting the offset information, generate offset candidate information, cut out a pixel block from the position of the offset candidate information of the detection target image, perform a discrete Fourier transform of the pixel block, and perform a detection target coefficient matrix. Generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, and combining the position marker sequence, the correlation value of the detection target sequence with the offset candidate information, and detecting the position marker. 9. The digital watermark detection method according to claim 8, wherein information is generated, and offset candidate information when a correlation value between the position marker sequence and the detection target sequence is maximum is extracted as offset information. 12. When detecting the digital watermark, a pixel block of N × N pixels is cut out from a position translated from the detection target image by the offset candidate information, and a discrete Fourier transform is performed on the pixel block. Generating a target coefficient matrix, generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, generating symbol detection information from the detection target sequence and the key information, and detecting a symbol from the symbol detection information. 9. The digital watermark detection method according to claim 8, wherein a detection result is detected and output as a digital watermark for each of the detected symbols. 13. When detecting the digital watermark, a sequence for each symbol position is generated for each symbol position from the symbol detection information, and a symbol candidate for taking the maximum value in the sequence for each symbol position is determined. The digital watermark detection method according to claim 12, wherein the detection symbol is used as the detection symbol, the detection symbol is subjected to inverse symbol conversion, and the detection result is obtained. 14. The digital watermark detection method according to claim 8, wherein, when detecting the digital watermark, a pixel block is output by cyclically shifting the cut-out pixel block based on the offset information. 15. When detecting the digital watermark, if the detection target image is equal to or larger than N × N size,
A pixel block of N × N pixels is cut out from the detection target image, and when the detection target image is smaller than N × N size,
An average pixel value of a portion of the detection target image included in the N × N size of the detection target image is obtained, and a portion less than the N × N size is filled with the average pixel value to generate a pixel block. 9. The digital watermark detection method according to claim 8, wherein the pixel block is output by cyclically shifting the pixel block. 16. When detecting the digital watermark, a pixel block is divided from an offset position of the detection target image, a discrete Fourier transform is performed on the divided pixel block, a detection target coefficient matrix is generated, and the detection is performed. The method according to claim 8, wherein a detection target sequence is generated from the target component position information and the detection target coefficient matrix, symbol detection information is generated from the detection target sequence and the key information, and a pixel block is output from the symbol detection information. Digital watermark detection method. 17. When detecting the digital watermark, the detection target image is divided into pixel blocks, the pixel blocks are cyclically shifted according to offset information, and the cyclically shifted pixel blocks are subjected to a discrete Fourier transform. Generating a detection target sequence matrix; generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; generating symbol detection information from the detection target sequence and the key information; The digital watermark detection method according to claim 8, wherein a detection symbol is generated for the digital watermark, and a digital watermark is detected from the detection symbol. 18. When detecting the digital watermark, the detection target image is divided into pixel blocks, and the pixel blocks are cyclically shifted by offset. At this time, N × For the portion less than the N size, the missing portion is filled with an average pixel value, the pixel block is subjected to discrete Fourier transform, a detection target coefficient matrix is generated, and the detection target component position information and the detection target coefficient matrix are detected. The digital watermark according to claim 8, further comprising: generating a target sequence; generating symbol detection information from the detection target sequence and the key information; generating each detection symbol from the symbol detection information; and detecting a digital watermark from the detection symbol. Detection method. 19. The method according to claim 8, wherein, when generating the position marker sequence, the detection target image is divided into blocks based on offset information, and an added block to which the pixel values of the pixel blocks are added is output as a pixel block. Digital watermark detection method. 20. When generating the position marker sequence, the detection target image is divided into blocks, and a block obtained by adding pixel values of the divided blocks is set as an addition block. 9. The digital watermark detection method according to claim 8, wherein a pixel block is generated by performing a cyclic shift based on the offset information. 21. When the position marker sequence is generated, the detection target image is divided into blocks. At this time, a pixel area less than one block around an end of the detection target image is divided into blocks. An average value corresponding to the detection target image is obtained and filled with the average value. For a part that is still insufficient, an average value of the pixel values corresponding to the detection target image is obtained, and the average value is filled to form a block. 20. The digital watermark detection method according to claim 19, wherein a block obtained by adding pixel values of the block is set as an addition block, and the addition block is cyclically shifted based on the offset information to generate a pixel block. 22. When generating the position marker sequence, after generating the detection target sequence, the average of the detection target sequence is set to 0.
The digital watermark detection method according to claim 11, wherein the method is changed to: 23. The digital watermark detection method according to claim 12, wherein when detecting the digital watermark, after generating the detection target sequence, an average of the detection target sequence is changed to 0. 24. When extracting the offset information at the time of detecting the position marker sequence, a maximum value of all the position marker detection information is obtained, and when the maximum value is smaller than a predetermined threshold, the offset value is detected. 9. The digital watermark detection method according to claim 8, wherein a signal indicating a failure is output, and when the maximum value is equal to or more than the predetermined threshold, offset information for obtaining the maximum value is output. 25. When detecting the digital watermark, symbol detection information is divided for each symbol position, a sequence for each symbol position is generated, a maximum value of the sequence for each symbol position is obtained, and the maximum value is a predetermined value. When the maximum value is smaller than the predetermined threshold, a signal indicating that detection of the digital watermark embedded in the detection target image has failed is output. The digital watermark detection method according to claim 8, wherein a symbol candidate at the time of taking is set as a detected symbol, and the detected symbol is converted into a digital watermark expression. 26. The digital watermarking method according to claim 25, wherein a threshold determined based on a distribution function value of a normal distribution having an average of 0 and a variance of 1 is applied as the threshold. 27. A digital watermark detection method for detecting a digital watermark from a detection target image in which a digital watermark is embedded so that the digital image is not perceived by human perception, wherein the digital watermark is detected from the detection target image In this case, the detection target component position information is generated from the key information, a position marker sequence is generated from the key information, and the starting point of the watermark pattern at the time of embedding and the detection target image are cut out based on the detected position marker detection information. Detecting offset information indicating a deviation from the start point of the pixel block, and detecting a digital watermark embedded in the pixel block using the offset information. 28. When generating the detection target component position information, a random number sequence having the key information as an initial value is generated, and the same length as the length of the embedded sequence embedded based on the random number sequence is generated. 28. The digital watermark detection method according to claim 27, wherein the detection target component position information is generated. 29. The digital watermark detection method according to claim 27, wherein when generating the position marker sequence, a first spread sequence with the key information as an initial value is generated, and the spread sequence is used as a position marker sequence. 30. When detecting the position marker detection information, an N × N size pixel block is cut out from the detection target image using the detection target image, the detection target component position information, and the position marker sequence. A discrete Fourier transform of the pixel block to generate a detection target coefficient matrix, a detection target sequence from the offset information, the detection target coefficient matrix, and the detection target component position information, and the position marker sequence and the detection 28. The digital watermark detection method according to claim 27, wherein a set of correlation values with a target sequence is used as position marker detection information. 31. When generating the position marker sequence, the detection target image is divided into N × N size blocks, and an addition block obtained by adding pixel values of all blocks is set as a pixel block. Electronic watermark detection method described in the above. 32. When detecting the digital watermark, a pixel block is cut out from the detection target image using the detection target image, the detection target component position information, the offset information, and the key information. The block is subjected to discrete Fourier transform to generate a detection target coefficient matrix, a detection target sequence is generated from the offset information, the detection target coefficient matrix, and the detection target component position information, and the detection target sequence is generated from the key information and the symbol candidate information. A symbol detection result is generated from a correlation value between the target sequence and the detection target sequence, a symbol candidate value having a maximum correlation value at each symbol position of the symbol detection result is used as a detection symbol, and the detection symbol is used in a digital watermark format. 28. The digital watermark detection method according to claim 27, wherein a result of the conversion is output. 33. When generating the position marker sequence, when a block of N × N size is cut out from the detection target image, if the block cannot be cut out to N × N size from the detection target image, N × N Only the portion corresponding to the detection target image included when trying to cut out by the size is cut out, and the portion that is not enough in the N × N size is obtained by averaging the pixel values corresponding to the detection target image, and the portion that is not enough in the average is obtained. 28. The digital watermark detection method according to claim 27, wherein the electronic watermark is embedded. 34. When detecting the digital watermark, dividing a pixel block from the detection target image, performing a discrete Fourier transform on the pixel block, generating a detection target coefficient matrix, the offset information and the detection target component 33. The digital watermark detection method according to claim 32, wherein a detection target sequence is generated from position information and the detection target coefficient matrix, a detection symbol is generated from the detection target sequence and the key information, and a sub-image is detected from the detection symbol. 35. When dividing a pixel block from the detection target image, when a block of N × N size is cut out from the detection target image, if the block cannot be cut out to N × N size from the detection target image, Only the portion corresponding to the detection target image included when trying to cut out in the N × N size is cut out, and the portion not enough in the N × N size is obtained by averaging the pixel values corresponding to the detection target image. Claim 3 to fill in the missing parts
4. The digital watermark detection method according to 4. 36. When the position marker sequence is detected and when the digital watermark is detected, the detection target image is divided into N × N size blocks, and the pixel values of all blocks are added. The digital watermark detection method according to claim 27, wherein the block is a pixel block. 37. When detecting the digital watermark, when an N × N size block is cut out from the detection target image, if the N × N size block cannot be cut out from the detection target image, an N × N size block is set. Only the portion corresponding to the detection target image included when trying to cut out is cut out, and the portion where the N × N size is insufficient is obtained by averaging the pixel values corresponding to the detection target image, and the portion where the average value is insufficient Claim 2
7. The digital watermark detection method according to 7. 38. The digital watermark detection method according to claim 31, wherein when detecting the position marker sequence, after generating the detection target sequence, a process of changing an average of the detection target sequence to 0 is performed. 39. The digital watermark detection method according to claim 32, wherein upon detecting the digital watermark, a process of changing the average of the detection target sequence to 0 after generating the detection target sequence. 40. The digital watermark detection method according to claim 34, wherein when detecting the digital watermark, a process of changing the average of the detection target sequence to 0 after generating the detection target sequence. 41. When detecting the offset information, a maximum value of the position marker detection information is obtained. If the maximum value is smaller than a predetermined threshold, a signal indicating that the detection of the offset information has failed is provided. 28. The digital watermark detection method according to claim 27, wherein, when the maximum value is equal to or more than a predetermined threshold value, offset information at the time of taking the maximum value is output. 42. When detecting the digital watermark, the symbol detection result is divided for each symbol position, a maximum value of a sequence for each symbol position for each symbol position is obtained, and the maximum value is smaller than a predetermined threshold value. In this case, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output. If the maximum value is equal to or greater than a predetermined threshold, a symbol candidate for taking the maximum value is output. 33. The electronic watermark detection method according to claim 32, wherein 43. The digital watermark detection method according to claim 42, wherein the threshold value uses a threshold value determined based on a value of a distribution function of a normal distribution having an average of 0 and a variance of 1. 44. A digital watermark embedding device for embedding a digital watermark in a digital image so as not to be perceived by human perception, wherein the watermark coefficient is a complex matrix based on the key information and the digital watermark embedded in the digital image. Means for independently changing the real and imaginary components of each coefficient value of the matrix; means for performing a discrete inverse Fourier transform on the changed watermark sequence matrix to generate a watermark pattern; and tiles of the watermark pattern on the original image. And add
Means for generating an embedded image. 45. An embedding target component designating means for generating embedding target component position information which is a sequence of lengths for embedding the digital watermark from the key information, and using the key information to divide position marker information and the digital watermark. An embedding sequence generating unit that spreads into an embedding sequence; and, using the embedding target component position information and the embedding sequence, a component value changing unit that changes a component value of a watermark coefficient matrix whose initial value is a complex matrix of a zero matrix. Discrete inverse Fourier transform of the watermark coefficient matrix, discrete inverse Fourier transform means for generating the watermark pattern, emphasizing the watermark pattern by an intensity parameter,
The digital watermark embedding device according to claim 44, further comprising watermark pattern adding means for generating an embedded image obtained by adding the emphasized watermark pattern to the input image in a tile shape. 46. The embedding target component designating unit includes a random number generating unit that generates a random number sequence with the key information as an initial value, and generates the embedding target component position information from the random number sequence. Digital watermark embedding device. 47. The embedded sequence generating means generates a 0th spread sequence with the key information as an initial value,
The spread sequence is a sequence representing position marker information, which is stored by multiplying by a predetermined number, the digital watermark is converted into a symbol expression, the total number of symbols and each symbol value are obtained, and the symbol value is added to the key information. Is generated as an initial value, and as a sequence representing a symbol, in addition to the respective items of the stored spreading sequence, an average value of the stored values is obtained, and the average value is subtracted from each item. 46. The electronic watermark embedding device according to claim 45, further comprising means for converting the set value into an embedding sequence. 48. The digital watermark embedding device according to claim 47, wherein said embedded sequence generating means uses one of +1 and -1 as said spread sequence. 49. The embedding sequence generating unit generates a spreading sequence using the key information as an initial value, multiplies the spreading sequence by a predetermined parameter, stores the multiplication in a storage unit, and stores the digital watermark in a symbol representation. And adding a spread sequence corresponding to the symbol of the symbol representation to each term of the storage means, and subtracting an average value from each term of the storage means as the embedded sequence. Item 45. The digital watermark embedding device according to Item 45. 50. The digital watermark embedding apparatus according to claim 49, wherein said embedded sequence generating means uses one of +1 and -1 as said spread sequence. 51. A digital watermark detection device for detecting a digital watermark from a detection target image in which a digital watermark is embedded so that the digital image is not perceived by human perception. Detection target component designating means for generating position information, position marker sequence generating means for generating a position marker sequence from the key information, detecting the position marker sequence, and using a watermark pattern start point and a detection target image when embedding. Position marker detection means for extracting offset information when the start point of the cut-out pixel block matches; digital watermark detection means for detecting the digital watermark embedded from the cut-out pixel block based on the offset information; An electronic watermark detection device comprising: 52. The digital watermark according to claim 51, wherein said detection target component designating means includes means for generating a random number sequence with said key information as an initial value, and generates said detection target component position information from said random number sequence. Detection device. 53. The position marker sequence generating means generates a 0th spread sequence with the key information as an initial value,
52. The digital watermark detection device according to claim 51, further comprising means for outputting the spread sequence as the position marker sequence. 54. The position marker detecting means, which generates offset candidate information, cuts out a pixel block from the position of the offset candidate information of the detection target image, and performs a discrete Fourier transform of the pixel block to obtain a detection target coefficient matrix. Means for generating, and a means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, a set of the position marker sequence and the correlation value of the detection target sequence and the offset candidate information, 52. The digital watermark detection device according to claim 51, further comprising: means for generating position marker detection information; and means for extracting, as offset information, offset candidate information when the correlation value between the position marker sequence and the detection target sequence is maximum. 55. The digital watermark detecting means, a means for cutting out a pixel block of N × N pixels from a position translated from the detection target image by the offset candidate information, and a discrete Fourier transform of the pixel block, Means for generating a target coefficient matrix; means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; means for generating symbol detection information from the detection target sequence and the key information; The digital watermark detection device according to claim 51, further comprising: means for generating each detection symbol from the symbol detection information; and means for outputting a detection result from each of the detection symbols as a digital watermark. 56. The digital watermark detection unit, comprising: a unit for generating a sequence for each symbol position for each symbol position from the symbol detection information; and a symbol candidate for obtaining the maximum value in the sequence for each symbol position. 55. A digital watermark detection apparatus according to claim 55, further comprising: means for converting the detected symbol into an inverse symbol to obtain the detection result. 57. The digital watermark detection device according to claim 51, wherein said digital watermark detection means includes means for outputting a pixel block by cyclically shifting said cut-out pixel block by said offset. 58. The digital watermark detecting means, wherein when the detection target image is equal to or larger than N × N size, the digital watermark detection means cuts out a pixel block of N × N pixels from the detection target image; If the size is smaller than the N size, an average pixel value of the detection target image portion included in the N × N size of the detection target image is obtained, and a portion less than the N × N size is filled with the average pixel value. 52. The digital watermark detection device according to claim 51, further comprising: means for generating a pixel block; and means for outputting a pixel block by cyclically shifting the cut-out pixel block. 59. The digital watermark detecting means, means for dividing a pixel block from an offset position of a detection target image, means for performing a discrete Fourier transform on the divided pixel block, and generating a detection target coefficient matrix, Means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; means for generating symbol detection information from the detection target sequence and the key information; and outputting a pixel block from the symbol detection information. 52. The digital watermark detection device according to claim 51, comprising: 60. The digital watermark detecting unit divides the detection target image into pixel blocks, cyclically shifts the pixel blocks with an offset, and performs a discrete Fourier transform on the cyclically shifted pixel blocks. Means for generating a detection target sequence matrix, means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, and means for generating symbol detection information from the detection target sequence and the key information, The digital watermark detection device according to claim 51, further comprising: means for generating each detected symbol from the symbol detection information; and means for detecting a digital watermark from the detected symbol. 61. The digital watermark detection unit divides the detection target image into pixel blocks, and cyclically shifts the pixel blocks by offsets.
For a portion less than the N × N size around the edge of the detection target image, a means for filling the missing portion with an average pixel value, a means for performing a discrete Fourier transform on the pixel block, and a detection target coefficient matrix, Means for generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; means for generating symbol detection information from the detection target sequence and the key information; and generating each detection symbol from the symbol detection information. 52. The digital watermark detection device according to claim 51, further comprising: a unit that detects a digital watermark from the detected symbols. 62. The position marker series generating unit includes: a unit that divides the detection target image into pixel blocks by offset; and a unit that outputs, as a pixel block, an addition block obtained by adding the pixel values of the pixel block. The digital watermark detection device according to claim 51. 63. The position marker series generating means, means for dividing the detection target image into blocks, a block obtained by adding pixel values of the divided blocks as an addition block, and the adding block 52. The digital watermark detection device according to claim 51, further comprising means for generating a pixel block by performing a cyclic shift by the offset. 64. The position marker series generation unit divides the detection target image into blocks. At this time, for a pixel area less than one block around an edge of the detection target image, the detection target image at the time of division The average value of the corresponding portion is obtained, the average value of the pixel values corresponding to the image to be detected is obtained for the portion that is not enough, and the average value is filled into the block, and the pixel value of the divided block is added. 63. A means for making the obtained block an addition block, and means for cyclically shifting the addition block by offset to generate a pixel block.
The electronic watermark detection device according to the above. 65. The position marker sequence generating means, after generating the detection target sequence, sets an average of the detection target sequence to 0.
55. The digital watermark detection device according to claim 54, further comprising: 66. The digital watermark detection device according to claim 55, wherein said digital watermark detection means includes means for changing the average of said detection target sequence to 0 after generating said detection target sequence. 67. The position marker detecting means obtains a maximum value of all the position marker detection information, and outputs a signal indicating that the offset detection has failed if the maximum value is smaller than a predetermined threshold value. 55. The digital watermark detection device according to claim 54, further comprising means for outputting an offset when the maximum value is obtained when the maximum value is equal to or greater than the predetermined threshold value. 68. A means for dividing the symbol detection information for each symbol position to generate a sequence for each symbol position, means for obtaining a maximum value of the sequence for each symbol position, Is smaller than a predetermined threshold, a signal indicating that detection of a digital watermark embedded in the detection target image has failed is output.If the maximum value is equal to or larger than the predetermined threshold, 52. The digital watermark detection device according to claim 51, further comprising: means for setting a symbol candidate at the time of taking a value as a detected symbol; and means for converting the detected symbol into a digital watermark expression. 69. The digital watermark detection device according to claim 68, wherein a threshold determined based on a distribution function value of a normal distribution having a mean of 0 and a variance of 1 is applied as the threshold. 70. A digital watermark detection device for detecting a digital watermark embedded in a digital image so as not to be perceived by human perception, comprising: a detection target component for generating detection target component position information from key information Designating means; position marker sequence generating means for generating a position marker sequence from the key information; start point of a watermark pattern at the time of embedding and start point of a pixel block cut out from a detection target image based on the detected position marker detection information Digital watermark detection, comprising: a position marker detecting unit that detects offset information indicating a deviation from the digital watermark; and a digital watermark detecting unit that detects a digital watermark embedded in the pixel block using the offset information. apparatus. 71. A detection target component designating means, comprising: a means for generating a random number sequence with the key information as an initial value; and detecting, based on the random number sequence, a length equal to the length of an embedded sequence to be embedded. 71. The digital watermark detection device according to claim 70, further comprising means for generating target component position information. 72. The position marker sequence generating means includes means for generating an initial spread sequence using the key information as an initial value, and using the spread sequence as a position marker sequence.
The electronic watermark detection device according to the above. 73. The position marker detecting unit, using the detection target image, the detection target component position information, and the position marker sequence, cutting out an N × N size pixel block from the detection target image, Means for generating a detection target coefficient matrix by performing a discrete Fourier transform of the pixel block; means for generating a detection target sequence from the offset information, the detection target coefficient matrix, and detection target component position information; and the position marker sequence. Means for setting a set of correlation values with the detection target sequence to obtain position marker detection information.
0. A digital watermark detection device according to item 0. 74. The electronic watermark detection unit, comprising: a unit that cuts out a pixel block from the detection target image using the detection target image, the detection target component position information, the offset information, and the key information; Means for generating a detection target coefficient matrix by performing a discrete Fourier transform on the block; means for generating a detection target sequence from the offset information, the detection target coefficient matrix, and the detection target component position information; and the key information and symbol candidates. Means for generating a symbol detection result from a correlation value between the sequence generated from the information and the detection target sequence; and, based on each symbol position of the symbol detection result, a symbol candidate value having a maximum correlation value as a detected symbol. 71. The electronic watermarking apparatus according to claim 70, further comprising: means for outputting the result of converting the detected symbol into a digital watermark format. Detector. 75. When a block of N × N size is cut out from the image to be detected when the block of N × N size cannot be cut out from the image to be detected, Means for cutting out only the portion equivalent to the detection target image included when trying to cut out, and calculating the average of the pixel values corresponding to the detection target image for the portion where the N × N size is insufficient, and filling in the portion missing in the average. 71. The digital watermark detection device according to claim 70, comprising: 76. The digital watermark detection means, means for dividing a pixel block from the detection target image, means for performing a discrete Fourier transform on the pixel block to generate a detection target coefficient matrix, the offset information and the detection 71. A means for generating a detection target sequence from the target component position information and the detection target coefficient matrix, a means for generating a detection symbol from the detection target sequence and the key information, and detecting a sub-image from the detection symbol. Digital watermark detection device. 77. The digital watermark detection means, when cutting out an N × N size block from the detection target image, if the block cannot be cut out into the N × N size from the detection target image, the digital watermark detection means Only the portion corresponding to the detection target image included when the image is to be extracted is cut out, and for the portion that is not enough in the N × N size, the average value of the pixel values corresponding to the detection target image is obtained, and the portion that is not sufficient is filled with the average value. 71. The electronic watermark detection device according to claim 70, comprising means. 78. The position marker series generating means and the digital watermark detecting means, wherein the detecting target image is divided into N × N size blocks, and an added block obtained by adding pixel values of all blocks is a pixel. 71. The digital watermark detection device according to claim 70, further comprising a block. 79. When the position marker sequence generation unit and the digital watermark detection unit cut out an N × N size block from the detection target image and cannot cut out the N × N size block from the detection target image, , Only the portion corresponding to the detection target image included when trying to cut out in the N × N size is cut out, and the missing portion in the N × N size is obtained by averaging the pixel values corresponding to the detection target image. 79. The digital watermark detection device according to claim 78, further comprising means for filling in a portion missing with the value. 80. The digital watermark detection device according to claim 73, wherein said position marker sequence generation means includes means for performing a process of changing an average of said detection target sequences to 0 after generating said detection target sequences. 81. The digital watermark detection device according to claim 74, wherein said digital watermark detection means includes means for performing a process of changing the average of said detection target series to 0 after generation of said detection target series. 82. A position marker detecting unit, comprising: a unit for obtaining a maximum value of the position marker detection information; and a signal indicating that the detection of the offset information has failed when the maximum value is smaller than a predetermined threshold. 71. The digital watermark detection apparatus according to claim 70, further comprising means for outputting, when the maximum value is equal to or greater than a predetermined threshold, offset information for obtaining the maximum value. 83. A means for dividing the digital watermark detection means into symbol detection information symbol positions generated from the detection target sequence and the key information, and a maximum value of a symbol position sequence for each symbol position. When the maximum value is smaller than a predetermined threshold, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output, and when the maximum value is equal to or larger than the predetermined threshold, 75. The digital watermark detection device according to claim 74, further comprising: means for outputting a symbol candidate when the maximum value is obtained. 84. The electronic watermark detection device according to claim 83, wherein the threshold value uses a threshold value determined based on a value of a distribution function of a normal distribution having an average of 0 and a variance of 1. 85. A storage medium storing a digital watermark embedding program for embedding a digital watermark in a digital image so as not to be perceived by human perception, comprising: a key information and a digital watermark embedded in the digital image. A process of independently changing a real component and an imaginary component of each coefficient value of a watermark coefficient matrix which is a matrix, a process of generating a watermark pattern by performing a discrete inverse Fourier transform of the changed watermark sequence matrix, and a process of generating the watermark pattern. Add to the original image in tiles,
A process for generating an embedded image. A storage medium storing a digital watermark embedding program. 86. An embedding target component designation process for generating, from the key information, an embedding target component position information, which is a sequence of lengths for embedding the digital watermark, using the key information to divide position marker information and the digital watermark An embedding sequence generation process for spreading to an embedding sequence, a component value changing process of changing a component value of a watermark coefficient matrix whose initial value is a complex matrix of zero matrix, using the embedding target component position information and the embedding sequence, A discrete inverse Fourier transform of the watermark coefficient matrix, and a discrete inverse Fourier transform process of generating the watermark pattern; emphasizing the watermark pattern with an intensity parameter;
86. A storage medium storing a digital watermark embedding program according to claim 85, further comprising: a watermark pattern adding process for generating an embedded image obtained by adding the emphasized watermark pattern in a tiled manner to an input image. 87. The embedding target component position information is generated from the random number sequence, and the embedding target component position information is generated from the random number sequence. A storage medium storing a digital watermark embedding program. 88. The embedded sequence generation process generates a 0th spread sequence with the key information as an initial value,
The spread sequence is a sequence representing position marker information, which is stored by multiplying by a predetermined number, the digital watermark is converted into a symbol expression, the total number of symbols and each symbol value are obtained, and the symbol value is added to the key information. Is generated as an initial value, and as a sequence representing a symbol, in addition to the respective items of the stored spreading sequence, an average value of the stored values is obtained, and the average value is subtracted from each item. 89. A storage medium storing a digital watermark embedding program according to claim 86, comprising a process of converting the set value into an embedding sequence. 89. The embedding sequence generation process generates a spread sequence using the key information as an initial value, multiplies the spread sequence by a predetermined parameter, stores the multiplied sequence in storage means, and symbolically expresses the digital watermark. And adding a spreading sequence corresponding to the symbol of the symbol representation to each term of the storage means, and subtracting an average value from each term of the storage means as the embedded sequence. A storage medium storing the digital watermark embedding program according to Item 86. 90. A storage medium storing a digital watermark embedding program according to claim 89, wherein said embedded sequence generating process uses one of +1 and −1 as said spread sequence. 91. A storage medium storing a digital watermark detection program for detecting a digital watermark from a detection target image in which a digital watermark is embedded so that the digital image is not perceived by human perception. A detection target component designation process for generating detection target component position information from information; a position marker sequence generation process for generating a position marker sequence from the key information; and a start point of a watermark pattern when the position marker sequence is detected and embedded. A position marker detection process for extracting offset information when a start point of a pixel block cut out from the detection target image matches, and detecting the digital watermark embedded from the pixel block cut out based on the offset information. Digital watermark detection program having a digital watermark detection process Storage medium storing the beam. 92. The electronic watermark according to claim 91, wherein the detection target component designation process includes a process of generating a random number sequence using the key information as an initial value, and the detection target component position information is generated from the random number sequence. A storage medium that stores a detection program. 93. The position marker sequence generation process includes: generating a 0th spread sequence with the key information as an initial value;
The storage medium storing the digital watermark detection program according to claim 91, including a process of outputting the spread sequence as the position marker sequence. 94. The position marker detection process includes: generating offset candidate information; extracting a pixel block from a position of the offset candidate information of the detection target image; and performing a discrete Fourier transform of the pixel block to obtain a detection target coefficient matrix. And a process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, a set of the position marker sequence and the correlation value of the detection target sequence and the offset candidate information, 92. The digital watermark detection program according to claim 91, further comprising: a process of generating position marker detection information; and a process of extracting, as offset information, offset candidate information when the correlation value between the position marker sequence and the detection target sequence is maximum. Storage media. 95. The digital watermark detection process includes: a process of cutting out a pixel block of N × N pixels from a position translated from the detection target image by the offset candidate information; and performing a discrete Fourier transform of the pixel block to perform detection. A process of generating a target coefficient matrix; a process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix; a process of generating symbol detection information from the detection target sequence and the key information; The storage medium storing the digital watermark detection program according to claim 91, comprising: a process of generating each detected symbol from the symbol detection information; and a process of outputting a detection result from each detected symbol as a digital watermark. 96. The digital watermark detection process includes: a process of generating a sequence for each symbol position for each symbol position from the symbol detection information; and a symbol candidate for obtaining a maximum value in the sequence for each symbol position. 97. A storage medium storing a digital watermark detection program according to claim 95, further comprising: a process of converting the detected symbol into an inverse symbol and obtaining the detection result. 97. The storage medium according to claim 91, wherein said digital watermark detection process includes a process of outputting a pixel block by cyclically shifting said cut-out pixel block by said offset. 98. The electronic watermark detection process includes, when the image to be detected has a size of N × N or more, cutting out a pixel block of N × N pixels from the image to be detected; If the size is smaller than the N size, an average pixel value of the detection target image portion included in the N × N size of the detection target image is obtained, and a portion less than the N × N size is filled with the average pixel value. 92. The storage medium storing the digital watermark detection program according to claim 91, further comprising: a process of generating a digital watermark, and a process of outputting a pixel block by cyclically shifting the cut-out pixel block. 99. The digital watermark detection process includes: a process of dividing a pixel block from an offset position of a detection target image; a process of performing a discrete Fourier transform on the divided pixel block to generate a detection target coefficient matrix; A process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, a process of generating symbol detection information from the detection target sequence and the key information, and outputting a pixel block from the symbol detection information 92. A storage medium storing a digital watermark detection program according to claim 91, comprising a process. 100. The digital watermark detection process includes: dividing the detection target image into pixel blocks; cyclically shifting the pixel blocks with an offset; and performing a discrete Fourier transform on the cyclically shifted pixel blocks. A process of generating a detection target sequence matrix, a process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, and a process of generating symbol detection information from the detection target sequence and the key information, The storage medium storing the digital watermark detection program according to claim 91, comprising: a process of generating each detected symbol from the symbol detection information; and a process of detecting a digital watermark from the detected symbol. 101. The digital watermark detection process divides the image to be detected into pixel blocks, and cyclically shifts the pixel blocks with an offset.
For a portion less than the N × N size around the edge of the detection target image, a process of filling the missing portion with the average pixel value, a process of generating a detection target coefficient matrix by performing a discrete Fourier transform of the pixel block, A process of generating a detection target sequence from the detection target component position information and the detection target coefficient matrix, a process of generating symbol detection information from the detection target sequence and the key information, and generating each detection symbol from the symbol detection information 92. A storage medium storing a digital watermark detection program according to claim 91, further comprising: a process of performing a digital watermark detection process from the detected symbols. 102. The position marker sequence generation process includes a process of dividing the detection target image into blocks by offset, and a process of outputting an added block obtained by adding the pixel values of the pixel block as a pixel block. Item 9
A storage medium storing the digital watermark detection program according to 1. 103. The position marker sequence generation process includes a process of dividing the detection target image into blocks, a block obtained by adding pixel values of the divided blocks to an addition block, and the adding block The storage medium storing the digital watermark detection program according to claim 91, comprising a process of generating a pixel block by performing a cyclic shift by an offset. 104. The position marker series generation process divides the detection target image into blocks. At this time, a detection target image at the time of division is obtained for a pixel area less than one block around an edge of the detection target image. A process of obtaining an average value of a corresponding portion, obtaining an average value of pixel values corresponding to an image to be detected in a portion that is insufficient, filling the average value into a block, and adding the pixel values of the divided block are performed. 103. The storage medium storing the digital watermark detection program according to claim 102, comprising: a process of using the obtained block as an addition block; and a process of cyclically shifting the addition block by an offset to generate a pixel block. 105. The position marker sequence generation process may include, after generating the detection target sequence, setting an average of the detection target sequence to 0.
95. A storage medium storing the digital watermark detection program according to claim 94, including a process of changing the electronic watermark into a program. 106. A storage medium storing a digital watermark detection program according to claim 100, wherein said digital watermark detection process includes a process of changing the average of said detection target sequence to 0 after generating said detection target sequence. 107. The storage medium storing the digital watermark detection program according to claim 101, wherein the digital watermark detection process includes a process of changing the average of the detection target sequence to 0 after generating the detection target sequence. 108. The position marker detection process obtains a maximum value of all the position marker detection information, and outputs a signal indicating that the offset detection has failed if the maximum value is smaller than a predetermined threshold value. The storage medium storing the digital watermark detection program according to claim 94, further comprising a process of outputting an offset when the maximum value is equal to or greater than the predetermined threshold value. 109. The digital watermark detection process includes: dividing symbol detection information for each symbol position to generate a sequence for each symbol position; obtaining a maximum value of the sequence for each symbol position; Is smaller than a predetermined threshold, a signal indicating that detection of a digital watermark embedded in the detection target image has failed is output.If the maximum value is equal to or larger than the predetermined threshold, The storage medium storing the digital watermark detection program according to claim 95, comprising: a process of using a symbol candidate when a value is taken as a detection symbol; and a process of converting the detection symbol into a digital watermark expression. 110. A storage medium storing a digital watermark detection program according to claim 109, wherein a threshold determined based on a value of a normal distribution function having a mean of 0 and a variance of 1 is applied as said threshold. 111. A storage medium storing a digital watermark detection program for detecting a digital watermark embedded in a digital image so as not to be perceived by human perception. A detection target component specifying process to be generated, a position marker sequence generation process to generate a position marker sequence from the key information, and a watermark pattern starting point and a detection target image at the time of embedding are cut out based on the detected position marker detection information. A position marker detection process for detecting offset information indicating a deviation from a start point of the pixel block; and a digital watermark detection process for detecting a digital watermark embedded in the pixel block using the offset information. Storage medium storing a digital watermark detection program. 112. The detection target component designating process includes: a process of generating a random number sequence with the key information as an initial value; and detecting the same length as the length of an embedded sequence embedded based on the random number sequence. 112. A storage medium storing a digital watermark detection program according to claim 111, comprising a process of generating target component position information. 113. The digital watermark detection program according to claim 111, wherein the position marker sequence generation process includes a process of generating an initial spread sequence using the key information as an initial value and using the spread sequence as a position marker sequence. Storage medium storing. 114. The position marker detection process includes: using the detection target image, the detection target component position information, and the position marker sequence to cut out an N × N size pixel block from the detection target image; A process of generating a detection target coefficient matrix by performing a discrete Fourier transform on the pixel block, a process of generating a detection target sequence from the offset information, the detection target coefficient matrix, and the detection target component position information, and the position marker sequence. A storage medium storing the digital watermark detection program according to claim 111, further comprising: a process of combining the correlation value with the detection target sequence to generate position marker detection information. 115. The digital watermark detection process includes: a process of cutting out a pixel block from the detection target image using the detection target image, the detection target component position information, the offset information, and the key information; A process of generating a detection target coefficient matrix by performing a discrete Fourier transform on the block; a process of generating a detection target sequence from the offset information, the detection target coefficient matrix, and the detection target component position information; and the key information and symbol candidates. A process of generating a symbol detection result from a correlation value between the sequence generated from the information and the detection target sequence, and, based on each symbol position of the symbol detection result, a symbol candidate value having a maximum correlation value as a detected symbol. And outputting a result of converting the detected symbol into a digital watermark format. 112. A storage medium storing the digital watermark detection program according to claim 111. 116. The position marker sequence generation process includes the steps of: when cutting out an N × N size block from the detection target image, if the block cannot be cut out into the N × N size from the detection target image, A process of cutting out only the portion corresponding to the detection target image included when trying to cut out, and averaging the pixel values corresponding to the detection target image for the portion where the N × N size is insufficient, and filling in the portion missing by the average. 115. A storage medium storing the digital watermark detection program according to claim 114. 117. The digital watermark detection process includes a process of dividing a pixel block from the detection target image, a process of performing a discrete Fourier transform on the pixel block to generate a detection target coefficient matrix, the offset information and the detection. 111. A process of generating a detection target sequence from target component position information and the detection target coefficient matrix, a process of generating a detection symbol from the detection target sequence and the key information, and detecting a sub-image from the detection symbol. Storage medium storing the digital watermark detection program. 118. The digital watermark detection process comprises: when an N × N size block is cut out from the detection target image, if the block cannot be cut out to the N × N size from the detection target image, Only the portion corresponding to the detection target image included when the image is to be extracted is cut out, and for the portion that is not enough in the N × N size, the average value of the pixel values corresponding to the detection target image is obtained, and the portion that is not sufficient is filled with the average value. 112. A storage medium storing a digital watermark detection program according to claim 111, including a process. 119. The position marker sequence generation process includes a process of dividing the detection target image into N × N size blocks, and a process of setting an addition block obtained by adding pixel values of all blocks to a pixel block. A storage medium storing the digital watermark detection program according to claim 111. 120. The digital watermark detection process includes a process of dividing the detection target image into N × N size blocks, and a process of setting an addition block obtained by adding pixel values of all blocks to a pixel block. Item 111. A storage medium storing a digital watermark detection program according to item 111. 121. In the digital watermark detection process, when an N × N size block is cut out from the detection target image, if the block cannot be cut out to the N × N size from the detection target image, the digital watermark detection process Only the portion corresponding to the detection target image included when the image is to be extracted is cut out, and for the portion that is not enough in the N × N size, the average value of the pixel values corresponding to the detection target image is obtained, and the portion that is not sufficient is filled with the average value. 112. A storage medium storing a digital watermark detection program according to claim 111, including a process. 122. The position marker sequence generation process includes the steps of: when cutting out an N × N size block from the detection target image, if the block cannot be cut out into the N × N size from the detection target image, Only the portion corresponding to the detection target image included when the clipping is attempted is cut out, and the portion that is not sufficient in the N × N size is obtained by averaging the pixel values corresponding to the detection target image. 112. A storage medium storing a digital watermark detection program according to claim 111, comprising a filling process. 123. The position marker sequence generation process includes a process of, after generating the detection target sequence, performing a process of changing an average of the detection target sequence to 0.
A storage medium storing the digital watermark detection program according to Item 2. 124. The storage storing the digital watermark detection program according to claim 121, wherein the digital watermark detection process includes a process of changing a mean of the detection target sequence to 0 after the detection target sequence is generated. Medium. 125. The position marker detection process includes: a process of obtaining a maximum value of the position marker detection information; and a signal indicating that the detection of the offset information has failed if the maximum value is smaller than a predetermined threshold. 111. A process for outputting, when the maximum value is equal to or greater than a predetermined threshold value, outputting offset information when the maximum value is obtained.
A storage medium storing the described electronic watermark detection program. 126. The digital watermark detection process includes: a process of dividing the detection target sequence and a symbol detection information generated from the key information for each symbol position; and a process of obtaining a maximum value of a sequence for each symbol position for each symbol position. When the maximum value is smaller than a predetermined threshold, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output, and when the maximum value is equal to or larger than the predetermined threshold, 112. A storage medium storing a digital watermark detection program according to claim 111, further comprising: outputting a symbol candidate when the maximum value is obtained. 127. The digital watermark detection process is a process of dividing the detection target sequence and the symbol detection information generated from the key information into each symbol position, and a process of obtaining the maximum value of the symbol position sequence for each symbol position. When the maximum value is smaller than a predetermined threshold, a signal indicating that the detection of the digital watermark embedded in the detection target image has failed is output, and when the maximum value is equal to or larger than the predetermined threshold, A storage medium storing a digital watermark detection program according to claim 115, further comprising a process of outputting a symbol candidate when the maximum value is obtained. 128. The storage medium storing the digital watermark detection program according to claim 126, wherein the threshold value uses a threshold value determined based on a distribution function value of a normal distribution having a mean of 0 and a variance of 1. The storage medium storing the electronic watermark detection program according to claim 127, wherein the threshold value uses a threshold value determined based on a distribution function value of a normal distribution having an average of 0 and a variance of 1. 130. A digital watermark embedding device that embeds a digital watermark in a digital image so as not to be perceived by human perception, and a digital watermark detection device that detects the digital watermark, The digital watermark embedding device includes: an embedding target component designating unit that generates, from the key information for embedding, embedding target component position information that is a sequence having the same length as the length n of the embedding sequence of the digital watermark; An embedding sequence generating unit that spreads the position marker information and the digital watermark to the embedding sequence using key information; and a watermark whose initial value is a complex matrix of zero matrix using the embedding target component position information and the embedding sequence. A component value changing unit for changing a component value of a coefficient matrix, and a discrete inverse Fourier transform of the watermark coefficient matrix, the watermark pattern And inverse discrete Fourier transform means for generating for, highlights the watermark pattern by the intensity parameter,
Watermark pattern adding means for generating an embedded image obtained by adding the emphasized watermark pattern to the input image in the form of a tile, wherein the digital watermark detection device performs detection target component position information from key information. A position marker sequence generating unit that generates a position marker sequence from the key information, and detects the position marker sequence and cuts out the watermark image starting point and the detection target image at the time of embedding. Position marker detecting means for extracting offset information when the start point of the pixel block matches, and digital watermark detecting means for detecting the digital watermark embedded from the pixel block cut out based on the offset information An electronic watermarking system, characterized in that: 131. The digital watermarking system according to claim 130, wherein said digital watermark embedding device is constituted by an integrated circuit. 132. The digital watermark system according to claim 130, wherein said digital watermark detection device is configured by an integrated circuit. 133. A digital watermark embedding device for embedding a digital watermark in a digital image so as not to be perceived by human perception, and a digital watermark detecting device for detecting the digital watermark, The digital watermark embedding device includes an embedding target component designating unit that generates, from the key information for embedding, embedding target component position information that is a sequence having the same length as the length n of the embedding sequence of the digital watermark, An embedding sequence generating unit that spreads the position marker information and the digital watermark to the embedding sequence using the key information; and using the embedding target component position information and the embedding sequence, an initial value is a complex matrix of zero matrix. A component value changing means for changing a component value of the watermark coefficient matrix, and a discrete inverse Fourier transform of the watermark coefficient matrix to generate the watermark pattern And inverse discrete Fourier transform means for, highlights the watermark pattern by the intensity parameter,
Watermark pattern adding means for generating an embedded image obtained by adding the emphasized watermark pattern to the input image in the form of a tile, wherein the digital watermark detection device performs detection target component position information from key information. A position marker sequence generating unit that generates a position marker sequence from the key information, and cuts out a watermark pattern starting point and a detection target image at the time of embedding based on the detected position marker detection information. Position marker detecting means for detecting offset information indicating a deviation from the start point of the pixel block, and digital watermark detecting means for detecting a digital watermark embedded in the pixel block using the offset information. Digital watermarking system featuring 134. The digital watermarking system according to claim 133, wherein said digital watermark embedding device is constituted by an integrated circuit. 135. An electronic watermarking system according to claim 133, wherein said electronic watermark detection device is constituted by an integrated circuit. 136. A digital watermark embedding method for embedding a digital watermark in a digital image, wherein a local complexity of the image is obtained, and a complicated portion of the image is divided into the portions according to the local complexity of the image. Make more changes than less complex parts,
A digital watermark embedding method characterized by acquiring an embedded image. 137. When creating local complexity of an image having a large value in a texture region in the image and having a small value in a flat region or an edge portion, the input image and each pixel are adjacent to each other. 138. The digital watermark embedding method according to claim 136, wherein when the difference from the pixel is equal to or larger than a certain threshold value, the edge and the texture area are measured by a counter that is incremented. 138. An edge / texture component value image obtained by measuring the edge and the texture area is subdivided into minute blocks, and a sum of edge / texture component values in the block is obtained for each subdivided block. 138. The digital watermark embedding method according to claim 137, wherein the sum of the edge texture component values is used as a texture degree index of the block. 139. An edge / texture component value image obtained by measuring the edge and the texture region is subdivided into minute blocks, and a sum of edge / texture component values in the block is obtained for each subdivided block. 138. The digital watermark embedding method according to claim 137, wherein a value obtained by mapping the sum with a predetermined function is used as a texture degree index of the block. 140. The edge texture component value is divided into minute blocks based on the edge texture component value image and the texture degree index, and the edge texture component value corresponds to each edge texture component value in the divided block. 139. The digital watermark embedding method according to claim 138, wherein a texture component value image is obtained by multiplying the texture degree index. 141. The edge texture component value is divided into minute blocks using the edge texture component value image and the texture degree index, and the edge texture component value corresponds to each edge texture component value in the divided block. 140. The digital watermark embedding method according to claim 139, wherein a texture component value image is obtained by multiplying the texture degree index. 142. When changing an image according to the local complexity of the image, the local complexity of the image may be determined when a difference between an input image, each pixel, and an adjacent pixel is equal to or greater than a certain threshold. 136. An edge / texture component value image is obtained by measuring an edge and a texture area based on a value of a counter to be incremented, and using the edge / texture component value image as a local complexity of the image. Digital watermark embedding method. 143. According to a local complexity of the image,
When changing the image, as the local complexity of the image, an edge / texture component value image obtained by measuring the edge and the texture region is subdivided into minute blocks, and the block is divided into blocks each for the subdivided block. 138. The digital watermark embedding according to claim 136, wherein a sum of edge texture component values in the block is obtained, the sum is used as a texture index of the block, and an upsampled texture index is used as a local complexity of the image. Method. 144. According to a local complexity of the image,
When changing the image, as the local complexity of the image, the edge / texture component value image obtained by measuring the edge and the texture area is subdivided into minute blocks, and the block is divided into blocks each for the subdivided block. The sum of the edge / texture component values is obtained, a value obtained by mapping the sum with a predetermined function is used as the texture index of the block, and the upsampled texture index is used as the local complexity of the image. 138. The digital watermark embedding method according to claim 136. 145. Depending on the local complexity of the image,
When changing the image, the edge / texture component value is divided into minute blocks using the edge / texture component value image and the texture index as the local complexity of the image. 143. The electronic device according to claim 136, wherein a texture component value image is obtained by multiplying each edge / texture component value in the block by a texture degree index corresponding thereto, and the texture component value image is used as a local complexity of the image. Watermark embedding method. 146. A digital watermark embedding method for embedding a digital watermark in a digital image so that the digital image is not perceived by human perception. When embedding another digital watermark in the digital image so as not to be perceived, the digital image Generating the local complexity of the image from the input image, generating a watermark pattern from the digital watermark using key information necessary for embedding, and converting the watermark pattern into the input image according to the local complexity of the image. A digital watermark embedding method, wherein an information embedded image is generated. 147. When generating the local complexity of the image, when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, one is added to a difference counter, and 146. The digital watermark embedding method according to claim 146, wherein the value of the difference counter at the time of completion of the processing of is used as an edge / texture component value image, and the edge / texture component value image is used as the local complexity of the image. 148. When generating the local complexity of the image, if a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold, the difference counter is incremented by one, and The value of the difference counter at the time when the processing for is completed is defined as an edge / texture component value image, and the edge / texture component image is subdivided into blocks.
Take the sum of the edge and texture component values of the block,
146. The digital watermark embedding method according to claim 146, wherein a texture degree index of the block is used, and an up-sample of the texture degree index is used as a local complexity of the image. 149. When generating the local complexity of the image, when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold, the difference counter is incremented by one, and The value of the difference counter at the time when the processing for is completed is defined as an edge / texture component value image, and the edge / texture component image is subdivided into blocks.
Take the sum of the edge and texture component values of the block,
148. The digital watermark embedding according to claim 146, wherein the texture index is multiplied by an edge texture component image of each block, a texture component value image is obtained, and the texture complexity index is used as the local complexity of the image. Method. 150. A digital watermark embedding method for embedding a digital watermark in a digital image so as not to be perceived by human perception, wherein a digital watermark is embedded in the digital image so as not to be perceived. Is divided into blocks of a predetermined size, a local complexity of the image is generated, and a key watermark necessary for embedding is used to generate a basic watermark pattern which is a basic pattern of the digital watermark. Generating an adaptive watermark pattern emphasizing the corresponding orthogonal transform coefficient of the basic watermark pattern according to the amplitude value of the orthogonal variable coefficient of the block; and combining the basic pattern and the watermark pattern with local complexity and intensity parameters of the image. Digital watermark embedding to generate an image in which information is embedded. Method. 151. When generating the local complexity of the image, when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold, the difference counter is incremented by one, and 151. The digital watermark embedding method according to claim 150, wherein the value of the difference counter at the time of the completion of the processing for is used as an edge / texture component value image, and the edge / texture component value image is used as the local complexity of the image. 152. When generating the local complexity of the image, if a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, add 1 to a difference counter, and The value of the difference counter at the time when the processing for is completed is defined as an edge / texture component value image, and the edge / texture component image is subdivided into blocks.
Take the sum of the edge and texture component values of the block,
The digital watermark embedding method according to claim 150, wherein a texture degree index of the block is set, and an up-sample of the texture degree index is set as a local complexity of the image. 153. The digital watermarking method according to claim 152, wherein a value obtained by mapping the texture degree index created from the edge texture component value image using a predetermined function is used as a texture degree index. 154. When generating the local complexity of the image, when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold, the difference counter is incremented by one, and The value of the difference counter at the time when the processing for is completed is defined as an edge / texture component value image, and the edge / texture component image is subdivided into blocks.
Take the sum of the edge and texture component values of the block,
150. The digital watermark embedding according to claim 150, wherein the texture index is multiplied by an edge texture component image for each block, a texture component value image is obtained, and the texture complexity index is used as the local complexity of the image. Method. 155. An electronic watermark embedding device for embedding an electronic watermark in a digital image so as not to be perceived by human perception, wherein an input image to be input is input and a local complexity of the image is obtained. A local complexity creating unit, a watermark pattern creating unit that creates a pattern representing the digital watermark using key information necessary for embedding, a pattern created by the watermark pattern creating unit on the input image, Adding the local complexity of the image,
An electronic watermark embedding device, comprising: an image adding unit that generates an information embedded image. 156. The local complexity creating means adds 1 to a difference counter when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and performs processing for the adjacent pixel. 160. The electronic watermark embedding device according to claim 155, further comprising: means for setting the value of the difference counter at the time of completion to an edge texture component value image; and means for setting the edge texture component value image to local complexity of the image. . 157. The local complexity creating means adds 1 to a difference counter when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and performs processing on the adjacent pixel. Means for setting the value of the difference counter at the time of completion as an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
155. The digital watermark embedding device according to claim 155, further comprising: means for setting a texture degree index of the block; and means for setting an up-sample of the texture degree index as local complexity of an image. 158. The local complexity creating means: means for generating an edge texture component value image from the input image; means for creating a texture index from the edge texture component value image; 155. The digital watermark embedding device according to claim 155, further comprising means for mapping the index using a predetermined function to obtain the local complexity of the image. 159. The local complexity creating means adds 1 to a difference counter when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and performs processing for the adjacent pixel. Means for setting the value of the difference counter at the time of completion as an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
A means for setting a texture degree index of the block; and a means for multiplying the edge / texture component image of each block by the texture degree index to obtain a texture component value image and setting the local complexity of the image. Item 15
5. The digital watermark embedding device according to 5. 160. An electronic watermark embedding device for embedding an electronic watermark in a digital image so as not to be perceived by human perception, comprising: an input image to be input; Block dividing means for subdividing into blocks; basic watermark pattern generating means for creating a basic watermark pattern representing the digital watermark from key information used for embedding and a digital watermark to be embedded in the input image; Adaptive watermark pattern creating means for creating an adaptive pattern emphasizing the corresponding orthogonal transform coefficient of the basic watermark pattern according to the amplitude value of the orthogonal transform coefficient of the block thus obtained; and determining the local complexity of an image from the block. Means for creating local complexity, the basic watermark pattern and the adaptive watermark pattern for each block. And an adding means for adding an information embedded image to the block in accordance with an intensity parameter representing a local complexity of the image and an embedding strength to generate an information embedded image. Implanting device. 161. The local complexity creating means adds 1 to a difference counter when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and performs processing for the adjacent pixel. 160. The electronic watermark embedding device according to claim 160, further comprising: means for setting the value of the difference counter at the time of completion to an edge texture component value image; and means for setting the edge texture component value image to local complexity of the image. . 162. The local complexity creating means adds one to a difference counter when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and performs processing on the adjacent pixel. Means for setting the value of the difference counter at the time of completion as an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
163. The digital watermark embedding device according to claim 160, further comprising: means for setting a texture degree index of the block; and means for setting an up-sample of the texture degree index as local complexity of an image. 163. The means for generating local complexity comprises: means for generating an edge / texture component value image from the input image; means for generating a texture index from the edge / texture component value image; 163. The digital watermark embedding device according to claim 160, further comprising means for mapping the index using a predetermined function to obtain the local complexity of the image. 164. The local complexity creating means adds 1 to a difference counter when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and performs processing on the adjacent pixel. Means for setting the value of the difference counter at the time of completion as an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
A means for setting a texture degree index of the block; and a means for multiplying the edge / texture component image of each block by the texture degree index to obtain a texture component value image and setting the local complexity of the image. Item 16
Electronic watermark embedding device according to 0. 165. The electronic watermark embedding device according to claim 155, wherein the electronic watermark embedding device is configured by an integrated circuit. 166. The digital watermark embedding device according to claim 160, wherein said digital watermark embedding device is configured by an integrated circuit. 167. A storage medium storing a digital watermark embedding program for embedding a digital watermark in a digital image so as not to be perceived by human perception, wherein a local complexity of the image is calculated from an input image to be input. A local complexity creating process for obtaining a digital watermark, a watermark pattern creating process for creating a pattern representing the digital watermark, a pattern created in the watermark pattern creating process for the input image, and a local complexity of the image. A digital watermark embedding program, wherein the digital watermark embedding program generates an information embedded image. 168. The local complexity creating process, wherein a difference counter is incremented by 1 when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value. 168. The digital watermark embedding program according to claim 167, further comprising: a process of setting the value of the difference counter at the time of completion to an edge / texture component value image; and a process of setting the edge / texture component value image to local complexity of the image. Storage medium storing. 169. The local complexity creating process, wherein a difference counter is incremented by 1 when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value. A process of setting the value of the difference counter at the time of completion to an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
170. The storage medium storing the digital watermark embedding program according to claim 167, comprising: a process of setting a texture degree index of the block; and a process of setting an upsampled texture degree index to local complexity of an image. 170. The local complexity creating process includes: generating an edge texture component value image from the input image; creating a texture index from the edge texture component value image; 168. The storage medium storing the digital watermark embedding program according to claim 167, further comprising a process of mapping the index using a predetermined function to obtain a local complexity of the image. 171. The local complexity creating process, wherein a difference counter is incremented by 1 when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and processing for the adjacent pixel is performed. A process of setting the value of the difference counter at the time of completion to an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
A process of multiplying the texture index by the edge texture component image for each block, obtaining a texture component value image, and setting a local complexity of the image. 167. A storage medium storing the digital watermark embedding program according to item 167. 172. A storage medium storing a digital watermark embedding program for embedding a digital watermark in a digital image so that the digital image is not perceived by human perception. A basic watermark pattern generation process for creating a basic watermark pattern representing the digital watermark from key information used for embedding and a digital watermark to be embedded in the input image, An adaptive watermark pattern creating process for creating an adaptive pattern emphasizing the corresponding orthogonal transform coefficient of the basic watermark pattern according to the amplitude value of the orthogonal transform coefficient of the block divided in the block dividing process; and A local complexity generation process for determining the local complexity, and the block An adding process of adding the basic watermark pattern and the adaptive watermark pattern to the block according to an intensity parameter representing the local complexity of the image and the strength of embedding to generate an information-embedded image. A storage medium storing a digital watermark embedding program. 173. The local complexity creating process, wherein a difference counter is incremented by 1 when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value. 172. The digital watermark embedding program according to claim 172, comprising: a process of setting the value of the difference counter at the time of completion to an edge / texture component value image; and a process of setting the edge / texture component value image to local complexity of the image. Storage medium storing. 174. The local complexity creating process, wherein a difference counter is added by 1 when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value, and processing for the adjacent pixel is performed. A process of setting the value of the difference counter at the time of completion to an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
173. The storage medium storing the digital watermark embedding program according to claim 172, comprising: a process of setting the texture index of the block as a texture index; and a process of setting an upsampled texture index to a local complexity of an image. 175. The local complexity creating process comprising: generating an edge texture component value image from the input image; creating a texture index from the edge texture component value image; 173. The storage medium storing the digital watermark embedding program according to claim 172, further comprising a process of mapping the index using a predetermined function to obtain a local complexity of the image. 176. The local complexity creating process, wherein a difference counter is incremented by 1 when a difference between each pixel of the input image and an adjacent pixel is larger than a predetermined threshold value. A process of setting the value of the difference counter at the time of completion to an edge / texture component value image, and subdividing the edge / texture component image into blocks,
Take the sum of the edge and texture component values of the block,
A process of multiplying the texture index by the edge texture component image for each block, obtaining a texture component value image, and setting a local complexity of the image. 172. A storage medium storing the digital watermark embedding program according to Item 172.