JP2000115518A - Picture processor, picture processing method and a computer-readable recording medium storing program for allowing computer to execute the method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently burying a digital watermark in picture data while suppressing visual influence by burying of information. SOLUTION: Based on the appearance frequency (statistical property) of a pixel pattern in the local area of picture data, a table 15 coordinating plural pixel patterns with a bit 1 or a bit 0 is prepared in advance. Then, embedding of embedded information (binary data string) to picture data by an encoding processing part 16 and restoration of embedded information by a decoding processing part 17 are executed by using this table 15.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, a method and a recording medium for embedding an information sequence composed of a plurality of information elements into image data and extracting the information sequence from the image data. The present invention relates to an image processing apparatus, an image processing method, and a computer-readable recording medium storing a program for causing a computer to execute the method, in which watermark information can be efficiently embedded in image data while suppressing a visual effect.

[0002]

2. Description of the Related Art In recent years, with the development of a network infrastructure such as the Internet or the spread of an image database represented by an electronic library, a digital watermark technique for digital image data is required. Here, since image data can be easily digitally copied between various media, it is extremely difficult for a copyright holder of a digital image to claim the copyright. In particular, once such digital images once reach the hands of a third party due to the development of network infrastructure in recent years, there is a possibility that the images will be spread over a wide area in a short time.

[0003] For this reason, there is known a technique in which desired digital watermark information such as a signature is embedded in a digital image so that a copyright holder can claim the copyright of the digital image. In addition, the presence of a digital watermark is not recognized by a third party, that is, the original image data is not altered as much as possible at the time of embedding, and even if the presence of the digital watermark is recognized, It is required that it cannot be modified or decrypted.

For example, Japanese Unexamined Patent Publication No. 9-191394 discloses that data to be watermarked is decomposed by Fourier transform or the like, and a watermark is inserted into a perceptually important component of the decomposed data. There is disclosed an electronic watermark insertion method configured to inversely transform the decomposed data subjected to the above processing by inverse Fourier transform or the like to generate watermarked data.

That is, in this conventional technique, data is spectrum-decomposed using Fourier transform, discrete cosine transform, Hadamard transform, and the like, and digital watermark information is mixed into the decomposed data.

[0006]

However, the Fourier transform, discrete cosine transform, Hadamard transform, and the like employed in the prior art generally require a large amount of processing time, so that they cannot be practically applied to large-sized images. There was a problem.

Further, when processing using a frequency space such as Fourier transform is involved, visual characteristics at the time of inverse conversion from the frequency space to the real space are not guaranteed, so that embedding with little effect on vision can be performed. Not exclusively. In particular, when such processing is performed on a binary image, a topological change often occurs in the image.

[0008] For this reason, in the digital watermarking technique, how to embed the watermark information in the image data efficiently while suppressing the visual influence by embedding the information in the image, that is, the existence of the digital watermark is not recognized by a third party. Is an extremely important issue.

The present invention has been made in view of the above-mentioned problem (problem), and has as its object to efficiently embed watermark information into image data while suppressing the visual influence of embedding information.

[0010]

In order to achieve the above object, an image processing apparatus according to the first aspect of the present invention embeds an information sequence including a plurality of information elements into image data,
In the image processing apparatus for extracting the information sequence from the image data, one or a plurality of pixels may be included in data that can be taken by each information element of the information sequence based on a statistical property of a pixel pattern of each local region forming the image data. A table in which patterns are associated, encoding means for sequentially embedding each information element of the information sequence in the image data based on the table, and image data in which the encoding means embeds the information sequence based on the table. Decoding means for sequentially extracting each information element forming the information sequence.

According to the first aspect of the present invention, one or a plurality of pixel patterns correspond to data that can be taken by each information element of the information sequence based on the statistical properties of the pixel patterns of each local area forming the image data. A prepared table is prepared, and based on this table, each information element of the information sequence is sequentially embedded in image data to perform encoding, and the information sequence is formed from the image data in which the information sequence is embedded based on this table. Since the information elements to be decoded are sequentially extracted, the watermark information can be efficiently embedded in the image data while suppressing the visual influence of the information embedding.

According to a second aspect of the present invention, there is provided an image processing apparatus, comprising: calculating means for calculating an appearance frequency of a pixel pattern in each local region forming the image data; Table creation means for automatically creating the table in which a plurality of pixel patterns are assigned to data that can be taken by each information element of the information sequence.

According to the second aspect of the present invention, the frequency of appearance of the pixel pattern in each local area forming the image data is calculated, and one or a plurality of pixel patterns are assigned to each information element of the information sequence based on the calculation result. Automatically creates a table assigned to data that can be taken, and performs encoding and decoding based on the created table. Therefore, the appearance frequency (statistical property) of the pixel pattern of each local region forming the image data changes each time. Even if different, information can be embedded according to the statistical properties of the image data.

According to a third aspect of the present invention, in the image processing apparatus, the encoding unit sequentially acquires information elements to be embedded from the information sequence,
Applying a window to the image data to obtain a window pattern of a local region where the window is located; and a window pattern obtained by the pattern obtaining unit is other than an information element obtained by the information element obtaining unit. Only if it matches any of the pixel patterns registered in the table corresponding to the information element of
Changing means for changing the pixel pattern of the local area to which the window is applied to a pixel pattern registered in the table corresponding to the information element obtained by the information element obtaining means, and applying the pattern obtaining means to the image data Window control means for controlling the position and order of the windows.

According to the third aspect of the present invention, the information elements to be embedded are sequentially obtained from the information sequence,
Applying a window to the image data to obtain a window pattern of a local area where the window is located, and obtaining the window pattern with one of the pixel patterns registered in a table corresponding to an information element other than the obtained information element. Only when there is a match, the pixel pattern of the local area to which the window is applied is changed to the pixel pattern registered in the table corresponding to the acquired information element. Thus, the change of the image data can be reduced.

The image processing apparatus according to claim 4, wherein the decoding means applies a window to the image data to obtain a window pattern of a local area where the window is located, When the window pattern acquired by the pattern acquisition unit matches any of the pixel patterns registered in the table, a restoration unit that restores data corresponding to the pixel pattern as an information element; and Window control means for controlling the position and order of windows applied to image data.

According to the fourth aspect of the present invention, a window is applied to the image data to acquire a window pattern of a local area where the window is located, and the acquired window pattern is assigned to any of the pixels registered in the table. When the pattern matches the pattern, the data corresponding to the pixel pattern is restored as an information element. Therefore, each information element embedded in the image data can be quickly and uniquely extracted to restore an information sequence.

The image processing apparatus according to a fifth aspect of the present invention is characterized in that the window control means controls the movement of windows applied to the image data by the pattern acquisition means at equal intervals.

According to the fifth aspect of the present invention, since the position of the window applied to the image data is moved and controlled at equal intervals, the conflicting trade-off relationship between the amount of information to be embedded and the degree of inconspicuousness can be easily adjusted. can do.

According to a sixth aspect of the present invention, in the image processing apparatus, the window control means controls the movement of the position of the window applied to the image data by the pattern acquisition means based on a reproducible pseudo-random number sequence. It is characterized by the following.

According to the sixth aspect of the invention, since the movement of the position of the window applied to the image data is controlled based on the pseudo-random number sequence having reproducibility, the position of the window which is the digital watermark information embedded in the image data is controlled. Decryption and falsification by three parties can be prevented, and the validity of the digital watermark information can be secured.

According to a seventh aspect of the present invention, in the image processing method of embedding an information sequence including a plurality of information elements into image data and extracting the information sequence from the image data, Each information element of the information sequence is based on a table in which one or more pixel patterns are associated with data that can be taken by each information element of the information sequence based on a statistical property of a pixel pattern of each local region to be formed. A first step of sequentially embedding the information sequence in image data; and a second step of sequentially extracting each information element forming the information sequence based on the table from the image data in which the information sequence is embedded.

According to the seventh aspect of the present invention, one or a plurality of pixel patterns correspond to data that can be taken by each information element of the information sequence based on the statistical properties of the pixel patterns of each local area forming the image data. A prepared table is prepared, and based on this table, each information element of the information sequence is sequentially embedded in image data to perform encoding, and the information sequence is formed from the image data in which the information sequence is embedded based on this table. Since the information elements to be decoded are sequentially extracted, the watermark information can be efficiently embedded in the image data while suppressing the visual influence of the information embedding.

[0024] In the image processing method according to the present invention, the first step acquires an information element to be embedded from the information sequence and applies a window to the image data. Obtains the window pattern of the local area where is located, and applies the window only when the obtained window pattern matches any of the pixel patterns registered in the table corresponding to the information element other than the obtained information element. The pixel pattern of the obtained local area is changed to the pixel pattern registered in the table corresponding to the acquired information element, the position of the window applied to the image data is updated, and at least all information forming the information sequence The above process is repeated until an element is embedded.

According to the invention of claim 8, information elements to be embedded are sequentially obtained from the information sequence,
Applying a window to the image data to obtain a window pattern of a local area where the window is located, and obtaining the window pattern with one of the pixel patterns registered in a table corresponding to an information element other than the obtained information element. Only when there is a match, the pixel pattern of the local area to which the window is applied is changed to the pixel pattern registered in the table corresponding to the acquired information element. Thus, the change of the image data can be reduced.

According to a ninth aspect of the present invention, in the image processing method, the second step applies a window to the image data to acquire a window pattern of a local area where the window is located, and acquires the acquired window pattern. If the pattern matches any of the pixel patterns registered in the table, the data corresponding to the pixel pattern is restored as an information element, the position of the window applied to the image data is updated, and the above processing is repeated. It is characterized by.

According to the ninth aspect of the present invention, a window is applied to the image data to obtain a window pattern of a local area where the window is located, and the obtained window pattern is assigned to any of the pixels registered in the table. When the pattern matches the pattern, the data corresponding to the pixel pattern is restored as an information element. Therefore, each information element embedded in the image data can be quickly and uniquely extracted to restore an information sequence.

According to a tenth aspect of the present invention, in the image processing method for embedding an information sequence including a plurality of information elements into image data and extracting the information sequence from the image data, Calculate the appearance frequency of the pixel pattern of each local region to be formed, automatically create a table in which one or more pixel patterns are assigned to data that can be taken by each information element of the information sequence based on the calculated result, Each information element of the information sequence is sequentially embedded in the image data based on the created table, and each information element forming the information sequence is sequentially extracted from the image data in which the information sequence is embedded based on the table. And

According to the tenth aspect, the frequency of appearance of the pixel pattern in each local region forming the image data is calculated, and one or a plurality of pixel patterns are determined based on the calculation result. Automatically creates a table assigned to data that can be taken, and performs encoding and decoding based on the created table. Therefore, the appearance frequency (statistical property) of the pixel pattern of each local region forming the image data changes each time. Even if different, information can be embedded according to the statistical properties of the image data.

[0030] A storage medium according to an eleventh aspect of the present invention records a program for causing a computer to execute the method according to any one of the seventh to tenth aspects, thereby making the program machine-readable. Thereby, the operation of any one of claims 7 to 10 can be realized by a computer.

[0031]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of an image processing apparatus, an image processing method, and a computer readable recording medium according to the present invention are described below with reference to the accompanying drawings. The form will be described in detail. In the following, a case will be mainly described in which a binary data sequence (bit sequence) is embedded in the binary image data.

(Embodiment 1) FIG. 1 is a functional block diagram showing a configuration of an image processing apparatus according to Embodiment 1.
The image processing apparatus 10 shown in FIG.
Creates a table 15 in which a plurality of pixel patterns are associated with bit 1 or bit 0 based on the frequency of appearance of pixel patterns in a local region of the average image data processed by The embedding information (binary data string) is embedded and restored in the data.

Specifically, a window is applied to the image data to apply a pixel pattern to a local region of the image data (hereinafter referred to as a pixel pattern).
A pixel pattern obtained by a window is called a “window pattern”. ) Is acquired, and when this window pattern matches any of the pixel patterns registered in the table 15, the constituent bits of the binary data string indicating the embedding information are embedded in the image data.

For example, when a table in which four types of pixel patterns shown in FIG. 2A are associated with bit 1 and four types of pixel patterns shown in FIG. 2B are associated with bit 0 is used, It is checked whether the window pattern acquired from the image data matches any of these eight types of pixel patterns. If there is no matching window pattern, the window is moved without embedding.

On the other hand, when embedding bit 1 in image data, if this window pattern matches any of the four types of pixel patterns shown in FIG. Complete the embedding and move the window. That is, in this case, the existence of the pixel pattern of the image data is determined by the bit 1
Will mean.

On the other hand, when embedding bit 1 in image data, if this window pattern matches any of the four types of pixel patterns shown in FIG. 2B, the pixel pattern of the local area where the window is located Figure 2
After changing to the pixel pattern shown in (a) and actually embedding bit 1, the window is moved. That is, in this case, the bit 1 is embedded by changing the pixel values of two pixels.

As described above, since the image processing apparatus 10 is configured to embed the embedding information in the image data by using the statistical property of the pixel pattern of the image data, it is necessary to change the pixel value. Information can be efficiently embedded in image data while minimizing it.

Next, a specific configuration of the image processing apparatus 10 shown in FIG. 1 will be described. As shown in FIG. 1, the image processing apparatus 10 includes an image input unit 11, an embedded information input unit 12, a control unit 13, a storage unit 14, a table 1
5, an encoding unit 16 and a decoding unit 17.

The image input unit 11 is an input device such as a scanner for inputting binary image data. Specifically, the image input unit 11 optically reads a document placed on a document table (not shown),
The read image data is binarized using a discriminant analysis method or the like, and binary image data is input. The image input unit 11 may be configured by a magnetic disk device or the like, and may be configured to read image data stored on a magnetic disk.

The embedding information input unit 12 is an input device such as a keyboard for inputting information (binary data string) to be embedded in image data. The information input from the embedding information input unit 12 is stored in the storage unit 14. You. It is also possible to input a character from the embedded information input unit 12 and store a bit string in which the ASCII code of the character is expressed in a binary number in the storage unit 14.

The control unit 13 is a control unit that performs overall control of the image processing apparatus 10.
a and a window pattern acquisition unit 13b. Specifically, when the image data is received from the image input unit 11, the image data is stored in the storage unit 14. When the embedding information is received from the embedding information input unit 12, the embedding information is stored in the storage unit 14. It is stored in the storage unit 14 in the form of a value data string.

When the embedding information is embedded (encoded) in the image data, the embedded data extracting unit 13a sequentially extracts the bits to be embedded in the image data and outputs the extracted bits to the encoding processing unit 16; 13b outputs the acquired window pattern of the image data to the encoding processing unit 16.

Further, when restoring (decoding) embedded information from image data, the window pattern acquisition unit 13b outputs the window pattern of the image data acquired to the encoding processing unit 16.

The embedded data extraction unit 13 a
Embedded information 14b consisting of a binary data string stored in the storage unit 4
, The bits to be embedded next are sequentially extracted and output to the encoding processing unit 16.

The window pattern acquisition section 13b has 3 ×
A window having a size of 3 is set in the image data, a window pattern indicating a pixel pattern of a local region of the image data is obtained, and output to the encoding processing unit 16 or the decoding processing unit 17. Once the window pattern has been acquired, this window is set at a predetermined interval (3
(Pixel unit). By changing the interval, it is possible to easily adjust the conflicting relationship between the amount of information to be embedded and the degree of inconsistency.

The storage section 14 is a storage device for storing image data (binary image data) input from the image input section 11 and embedded information (binary data string) input from the embedded information input section 12.

The table 15 stores a plurality of pixel patterns in bit 1 based on the frequency of appearance of pixel patterns in a local region of average image data processed by the image processing apparatus 10.
Alternatively, it is a table corresponding to bit 0 and is created in advance.

The encoding unit 16 stores the table 1
5 is a processing unit for embedding the embedding information 14b in the image data by using No. 5. Specifically, when receiving the bit to be embedded in the image data and the window pattern, the pixel pattern corresponding to the window pattern is stored in the table 1.
It is determined whether the pixel pattern matches any one of the pixel patterns registered in No. 5 and, if not, outputs to the control unit 13 that no embedding is performed.

The control unit 13 that has received this output
Moves the window by three pixels, acquires a new window pattern, and outputs this window pattern to the encoding processing unit 16.

On the other hand, when the window pattern is registered in the table 15 as a pixel pattern corresponding to the bit to be embedded, the fact that the embedding is completed is output to the control unit 13.

For example, when embedding bit 1 in image data, this window pattern is
If it matches with any of the pixel patterns, it outputs to the control unit 13 that the embedding of the bit 1 is completed.

The control unit 13 that has received this output
Moves the window by three pixels to acquire a new window pattern, acquires the next bit to be embedded, and outputs the window pattern and the bit to be embedded to the encoding processing unit 16.

On the other hand, when the window pattern is registered in the table 15 as a pixel pattern corresponding to bits other than the bits to be embedded, the pixel pattern corresponding to the bits to be embedded in the image data registered in the table 15 is stored in the control unit 13. Output to

For example, when embedding bit 1 in image data, this window pattern is
If any of the pixel patterns match, FIG.
The pixel pattern shown in FIG.

After receiving this output, the control unit 13 changes the pixel pattern on the image data 14a where the window is located to the pixel pattern received from the encoding processing unit 16, and then moves the window by three pixels to create a new window. While acquiring the pattern, the next bit to be embedded is acquired, and the window pattern and the bit to be embedded are output to the encoding processing unit 16.

In this embodiment, even if the embedding of all the bits forming the embedding information is once terminated, each bit forming the embedding information is repeatedly embedded.

The decoding processing section 17 is a processing section for sequentially extracting bits embedded in image data using the table 15 and restoring a binary data string forming embedded information.

More specifically, when a window pattern is received from the control unit 13, it is determined whether or not a pixel pattern corresponding to the window pattern is registered in the table 15, and if the pixel pattern is not registered in the table 15, ,
A request for the next window pattern is made to the control unit 15, and if the next window pattern is registered in the table 15, the bit corresponding to this pixel pattern is stored in a shift register internally held, and then the next window pattern is sent to the control unit 15. Request to. The configuration of the image processing apparatus 10 used in the present embodiment has been described above.

Next, the encoding processing procedure of the image processing apparatus 10 will be described. FIG. 3 is a flowchart showing an encoding processing procedure of the image processing apparatus 10 shown in FIG. Here, it is assumed that the embedded information 14b has already been stored in the storage unit 14.

As shown in the flowchart of FIG. 3, after the image data is input from the image input unit 11 and stored in the storage unit 14 (step S301), the embedding data extraction unit 13a sets the embedding information comprising a binary data string. The first bit of 14b is extracted and output to the encoding processing unit 16 (step S302). The window pattern acquisition unit 13b acquires a window pattern by setting a window at the start point of the image data (steps S303 to S303).
304), and outputs the result to the encoding processing unit 16.

After that, the encoding processing unit 16 determines that the window pattern received from the window acquiring unit 13b is the bit 1 and the bit 0 registered in the table 15.
Is determined (step S305), and if it does not match any of the pixel patterns (No at step S305), the process proceeds to step S311.

On the other hand, if this window pattern matches any of the pixel patterns registered in the table 15 (Yes at step S305), it is checked whether the bit to be embedded is 1 (step S30).
6).

As a result, if the bit to be embedded is 1 and the window pattern does not correspond to bit 1 (No at step S307), the encoding processing unit 16 instructs the control unit 13 to correspond to bit 1. The control section 13 outputs a pixel pattern to be processed.
The upper window pattern is changed to a pixel pattern corresponding to bit 1 (step S308).

If the bit to be embedded is 0 and the window pattern does not correspond to bit 0 (No at step S 309), the encoding processor 16 instructs the controller 13 to correspond to bit 0. The pixel pattern is output, and the control unit 13 changes the window pattern on the image data 14a to a pixel pattern corresponding to bit 0 (step S310).

The bit to be embedded is 1, and
If the window pattern corresponds to bit 1 or if the bit to be embedded is 0 and the window pattern corresponds to bit 0, the bit is not changed without changing the image data 14a. Can be embedded.

If the scanning by the window has not been completed up to the end point of the image data 14a, the window is moved by three pixels in the raster direction (step S).
311 to 312), and determines whether or not the bit is the last bit of the embedded information 14b (step S31).
3).

If it is the last bit (Yes at Step S313), the leading bit of the embedded information 14b is extracted (Step S314), and if it is not the last bit (Step S313 No), the next bit is extracted. (Step S315), Step S304
Move to The window scanning is repeated until the end point of the image data 14a by repeating the processes of steps S304 to 315, and the process ends when the window scanning is completed (Yes at step S311).

Next, the decoding processing procedure of the image processing apparatus 10 shown in FIG. 1 will be described. FIG. 4 is a flowchart showing a decoding processing procedure of the image processing apparatus 10 shown in FIG. As shown in FIG. 4, after the image data is input from the image input unit 11 and stored in the storage unit 14 (step S401), the window pattern acquisition unit 13b sets the window at the start point of the image data and sets the window pattern. Acquisition is performed (steps S402 to S403), and output to the decoding processing unit 17.

Thereafter, the decoding processing unit 17 determines that the window pattern received from the window obtaining unit 13b is
It is determined whether the pixel pattern matches the pixel pattern corresponding to any of bit 1 and bit 0 registered in table 15 (step S404). If the pixel pattern matches the pixel pattern corresponding to bit 1 (step S405: YES), Adds 1 to the bit string held therein (step S40).
6) If the pixel pattern matches the pixel pattern corresponding to bit 0 (No at Step S405), 0 is added to the internally held bit string (Step S407).

Thereafter, the window is moved by three pixels in the raster direction (step S409), and the same processing is repeated. When the window has been scanned to the end point of the image (step S408: YES), the processing ends.

As described above, in the first embodiment,
Based on the appearance frequency of the pixel pattern in the local region of the image data, a table 15 in which a plurality of pixel patterns are associated with bit 1 or bit 0 is created, and the table 15 is used to embed information ( Since the embedding and restoring of the binary data sequence are performed, (1) the change of the image data can be minimized, and (2) the visual influence of the embedding information can be suppressed. (3) It is possible to obtain an effect that watermark information can be efficiently embedded in image data by simple processing.

(Embodiment 2) By the way, in Embodiment 1 described above, based on the appearance frequency of the pixel pattern in the local area of the average image data processed by the image processing apparatus 10,
The case where the table 15 in which a plurality of pixel patterns are associated with the bit 1 or the bit 0 is created in advance is shown. However, when various image data are handled in a wide range, the statistical properties of the average image data do not correspond. Can not do it.

Thus, a second embodiment in which the statistical properties are automatically calculated for each image data and the table 15 is automatically created will be described below. FIG. 5 is a functional block diagram showing a configuration of the image processing device 50 used in the second embodiment. However, functional units having the same functions as those of the image processing apparatus 10 shown in FIG. 1 are given the same numbers, and detailed descriptions thereof will be omitted.

As shown in FIG. 5, this image processing device 50
Is provided in the image processing apparatus 10 shown in FIG.
Is added. The table creation unit 51 automatically creates a table 15 in which a plurality of pixel patterns are associated with bit 1 or bit 0 based on the frequency of appearance of pixel patterns in a local region of image data input from the image input unit 11. .

Specifically, the image data is sequentially scanned by a window, the frequency of appearance of the pixel pattern in the local area is calculated, a predetermined number of pixel patterns having a high calculated frequency are selected, and the selected pixel pattern is The table 15 is created by assigning 1 or 0.

FIG. 6 is a flowchart showing a table creation procedure by the table creation section 51. As shown in FIG. 6, image data is input from the image input unit 11,
When the image data is stored in the storage unit 14 (step S601), the window is set at the start point of the image, and a window pattern (pixel pattern) is obtained (steps S602 to S603).

After the frequency of the pixel pattern is incremented (step S604), the window is moved by a predetermined number of pixels (step S605), and the processes of steps S603 to S606 are repeated until the end point of the image is reached (step S605). Step S606). Thereafter, a predetermined number of pixel patterns having a high frequency are selected (step S607), and bit 1 or 0 is assigned to the selected pixel pattern to create a table 15 (step S608). By performing the above series of processing, the table 15 suitable for the image data can be automatically created.

As described above, in the second embodiment,
A table creation unit 51 is provided for automatically creating a table 15 in which a plurality of pixel patterns are associated with bits 1 or 0 based on the frequency of appearance of pixel patterns in a local region of image data input from the image input unit 11. With this configuration, even if the appearance frequency (statistical property) of the pixel pattern in each local region forming the image data is different each time, information can be embedded according to the statistical property of the image data.

(Embodiment 3) By the way, in Embodiments 1 and 2, the case where the window pattern acquisition unit 13b controls the movement of the window applied to the image data 14a at regular intervals for every predetermined number of pixels. However, the movement of such a window can be randomly controlled. Therefore, a third embodiment in which a window is randomly moved using a pseudo-random number sequence will be described below. FIG.
9 is a functional block diagram illustrating a configuration of an image processing device 70 used in Embodiment 3. FIG. However, functional units having the same functions as those of the image processing apparatus 10 shown in FIG. 1 are given the same numbers, and detailed descriptions thereof will be omitted.

As shown in FIG. 7, this image processing device 70
Has a key input unit 71 and a window pattern obtaining unit 7
2 generates a pseudo-random number sequence based on the input key from the key input unit 71, and controls the movement of the window according to the generated pseudo-random number sequence.

The key input section 71 is an input section for inputting a key for randomly controlling movement of a window applied to image data by the window pattern obtaining section 72 in accordance with a pseudo-random number sequence. For example, when the window pattern acquisition unit 72 uses a maximum long-period sequence (M sequence) or a two-dimensional maximum long-period sequence (M plane) as a pseudo-random number sequence, the primitive rule polynomial forming these maximum long-period sequences Alternatively, an initial value or the like can be used as a key.

The reason why the movement of the window is controlled based on the pseudo-random number sequence is to prevent information embedded in the image data from being decoded or falsified by a third party. Also, the reason why a pseudo-random number sequence is used instead of a physical random number is that the random number sequence used in the encoding process is needed again in decoding, so that reproducibility of the random number is required.

As described above, in the third embodiment,
A key input unit 71 for inputting a key of a pseudo-random number sequence is provided, and the window pattern acquisition unit 72 is configured to control movement of a window according to a pseudo-random number sequence generated based on the key. Can be prevented from being decrypted or tampered with by a third party.

In the third embodiment, the case where the movement of the window is controlled based on the pseudo-random number sequence has been described. However, the present invention is not limited to this, and the movement of the window is controlled in a predetermined order. It is also possible.

In the first to third embodiments, the case where the binary data string is embedded in the binary image data has been described. However, the present invention is not limited to this. Can be applied to the case where a binary data sequence or a multi-valued data sequence is embedded in a.

Further, in the first to third embodiments, information is embedded only when the window pattern matches the pixel pattern registered in the table 15, but the window pattern matches the pixel pattern registered in the table 15. When it is close, it is also possible to embed information.

[0087]

As described above in detail, according to the first aspect of the present invention, the data that can be taken by each information element of the information sequence based on the statistical properties of the pixel pattern of each local area forming the image data A table in which one or more pixel patterns are associated with each other is prepared, and based on this table, each information element of the information sequence is sequentially embedded in image data to perform encoding, and the information sequence is embedded based on this table. Since decoding is performed to sequentially retrieve each information element forming the information sequence from the image data, (1) the change of the image data can be minimized, and (2) the visual information provided by the embedded information can be obtained. (3) watermark information can be efficiently embedded in image data by simple processing, and (4)
By automatically embedding information strings such as the machine number and date of a copying machine in image data, an image processing apparatus capable of tracking an illegal copy of securities or the like ex post without being noticed by a user has been developed. The effect is obtained.

According to the second aspect of the present invention, the appearance frequency of a pixel pattern in each local region forming image data is calculated, and one or a plurality of pixel patterns are determined based on the calculation result. Since a table assigned to data that can be taken by elements is automatically created, and encoding and decoding are performed based on the created table, the appearance frequency (statistical property) of the pixel pattern of each local region forming image data ) Is different each time, an image processing apparatus capable of embedding information according to the statistical properties of the image data is obtained.

According to the third aspect of the present invention, information elements to be embedded are sequentially obtained from an information sequence, and a window is applied to image data to obtain a window pattern of a local area where the window is located. Only when the obtained window pattern matches any of the pixel patterns registered in the table corresponding to the information element other than the obtained information element, the pixel pattern of the local area to which the window is applied is added to the obtained information element. Since the pixel pattern is changed to the pixel pattern registered in the corresponding table, an image processing apparatus capable of reducing the change of the image data by using the pixel pattern originally included in the image data as an information element as much as possible is obtained. The effect is that it can be done.

According to the fourth aspect of the present invention, a window is applied to image data to acquire a window pattern of a local area where the window is located, and the acquired window pattern is registered in any of the tables. Since the data corresponding to the pixel pattern is restored as an information element when it matches the pixel pattern, each information element embedded in the image data can be quickly and uniquely extracted to restore the information sequence. This provides an effect that a simple image processing device can be obtained.

According to the fifth aspect of the present invention, since the position of the window applied to the image data is controlled to be moved at regular intervals, the conflicting trade-off relationship between the amount of information to be embedded and the degree of inconspicuousness. Is obtained, whereby an image processing apparatus that can easily adjust the image quality is obtained.

According to the sixth aspect of the present invention, since the movement of the position of the window applied to the image data is controlled based on a reproducible pseudo-random number sequence, the electronic watermark information is embedded in the image data. There is an effect that an image processing apparatus capable of preventing the information sequence from being decrypted and falsified by a third party and ensuring the validity of the digital watermark information can be obtained.

According to the seventh aspect of the present invention, one or a plurality of pixel patterns are assigned to data that can be taken by each information element of an information sequence based on the statistical properties of the pixel patterns of each local area forming image data. A corresponding table is prepared, and based on this table, each information element of the information sequence is sequentially embedded in the image data to perform encoding, and the information sequence is embedded from the image data in which the information sequence is embedded based on this table. Since decoding is performed to sequentially retrieve each information element to be formed, (1) the change of image data can be minimized, and (2) the visual influence of embedded information can be suppressed. so,
(3) It is possible to embed watermark information into image data efficiently with simple processing, and (4) automatically embed an information sequence such as a machine number and date of a copying machine into image data.
There is an effect that an image processing method capable of tracking an illegal copy of a securities or the like ex post fact without being noticed by the user is obtained.

According to the invention of claim 8, information elements to be embedded are sequentially obtained from the information sequence, and a window is applied to image data to obtain a window pattern of a local area where the window is located. Only when the obtained window pattern matches any of the pixel patterns registered in the table corresponding to the information element other than the obtained information element, the pixel pattern of the local area to which the window is applied is added to the obtained information element. Since the pixel pattern is changed to the pixel pattern registered in the corresponding table, an image processing method capable of reducing the change of the image data by using the pixel pattern originally included in the image data as an information element as much as possible is obtained. The effect is that it can be done.

According to the ninth aspect of the present invention, a window is applied to image data to acquire a window pattern of a local area where the window is located, and the acquired window pattern is registered in any of the tables. Since the data corresponding to the pixel pattern is restored as an information element when it matches the pixel pattern, each information element embedded in the image data can be quickly and uniquely extracted to restore the information sequence. This provides an effect that a simple image processing method can be obtained.

According to the tenth aspect of the present invention, the appearance frequency of a pixel pattern in each local area forming image data is calculated, and one or a plurality of pixel patterns are determined based on the calculation result. Since a table assigned to data that can be taken by elements is automatically created, and encoding and decoding are performed based on the created table, the appearance frequency (statistical property) of the pixel pattern of each local region forming image data ) Is different each time, an image processing method that can embed information according to the statistical property of the image data is obtained.

The storage medium according to the eleventh aspect of the present invention records a program for causing a computer to execute the method according to any one of the seventh to tenth aspects, thereby making the program readable by a machine. Item 7
It is possible to obtain a recording medium capable of realizing the operations (1) to (10) by a computer.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing a configuration of an image processing apparatus according to Embodiment 1 of the present invention.

FIG. 2 is an explanatory diagram showing an example of a pixel pattern associated with bit 1 or bit 0;

FIG. 3 is a flowchart illustrating an encoding processing procedure of the image processing apparatus according to the first embodiment;

FIG. 4 is a flowchart illustrating a decoding processing procedure of the image processing apparatus according to the first embodiment;

FIG. 5 is a functional block diagram showing a configuration of an image processing apparatus according to Embodiment 2 of the present invention.

FIG. 6 is a flowchart illustrating a table creation procedure by a table creation unit of the image processing apparatus according to the second embodiment.

FIG. 7 is a functional block diagram showing a configuration of an image processing apparatus according to Embodiment 3 of the present invention.

[Explanation of symbols]

 Reference Signs List 10 Image processing device 11 Image input unit 12 Embedded information input unit 13 Control unit 13a Embedded data extraction unit 13b Window pattern acquisition unit 14 Storage unit 14a Image data 14b Embedded information 15 Table 16 Encoding processing unit 17 Decoding processing unit 51 Table creation unit 71 Key input unit 72 Window pattern acquisition unit

Claims (11)

[Claims] 1. An image processing apparatus for embedding an information sequence including a plurality of information elements into image data and extracting the information sequence from the image data, wherein a statistical property of a pixel pattern of each local region forming the image data is provided. A table in which one or a plurality of pixel patterns are associated with data that can be taken by each information element of the information sequence based on the encoding unit that sequentially embeds each information element of the information sequence in the image data based on the table. An image processing apparatus, comprising: decoding means for sequentially extracting each information element forming the information sequence from the image data in which the information sequence is embedded, based on the table. 2. A calculating means for calculating the frequency of appearance of a pixel pattern in each local area forming the image data; and one or more pixel patterns based on the calculation result by the calculating means. The image processing apparatus according to claim 1, further comprising: a table creating unit that automatically creates the table assigned to data that can be taken by an information element. 3. An information element acquisition unit for sequentially acquiring information elements to be embedded from the information sequence, a window pattern in a local area where the window is located by applying a window to the image data. And the window pattern acquired by the pattern acquisition unit matches any of the pixel patterns registered in the table corresponding to information elements other than the information element acquired by the information element acquisition unit. Only, changing means for changing the pixel pattern of the local area to which the window is applied to a pixel pattern registered in the table corresponding to the information element obtained by the information element obtaining means; and Control means for controlling the position and order of windows applied to windows The image processing apparatus according to claim 1 or 2, further comprising a. 4. A decoding apparatus comprising: a decoding unit configured to apply a window to the image data to obtain a window pattern of a local area where the window is located; A restoring unit for restoring data corresponding to the pixel pattern as an information element when the pixel pattern matches any of the pixel patterns registered in the image data, and controlling a position and an order of windows applied to the image data by the pattern acquiring unit. The image processing apparatus according to any one of claims 1 to 3, further comprising: a window control unit configured to perform the operation. 5. The image processing apparatus according to claim 3, wherein the window control unit controls movement of a window applied to the image data by the pattern acquisition unit at equal intervals. 6. The window control unit according to claim 3, wherein the pattern acquisition unit controls movement of a window position applied to the image data based on a reproducible pseudo-random number sequence. Image processing device. 7. An image processing method for embedding an information sequence composed of a plurality of information elements into image data and extracting the information sequence from the image data, wherein a statistical property of a pixel pattern of each local region forming the image data is provided. A first step of sequentially embedding each information element of the information sequence in the image data based on a table in which one or more pixel patterns are associated with data that can be taken by each information element of the information sequence based on the information; A second step of sequentially extracting each information element forming the information sequence from the image data in which the sequence is embedded based on the table. 8. The first step is to acquire an information element to be embedded from the information sequence, apply a window to the image data, and acquire a window pattern of a local area where the window is located, Only when the acquired window pattern matches any of the pixel patterns registered in the table corresponding to the information element other than the acquired information element, the pixel pattern of the local area to which the window is applied is acquired as the acquired information element. Changing the pixel pattern registered in the table corresponding to the above, updating the position of the window applied to the image data, and repeating the above processing until at least all the information elements forming the information sequence are embedded. The image processing method according to claim 7. 9. The second step is to apply a window to the image data to obtain a window pattern of a local area where the window is located, and to obtain any one of the pixels registered in the table with the obtained window pattern. 9. The method according to claim 7, wherein when the pattern matches, the data corresponding to the pixel pattern is restored as an information element, the position of a window applied to the image data is updated, and the processing is repeated. Image processing method. 10. An image processing method for embedding an information sequence composed of a plurality of information elements into image data and extracting the information sequence from the image data, comprising: determining an appearance frequency of a pixel pattern of each local region forming the image data; Calculating, automatically creating a table in which one or a plurality of pixel patterns are assigned to data that can be taken by each information element of the information sequence based on the calculated result, and based on the created table, each information of the information sequence An image processing method comprising: sequentially embedding elements in the image data; and sequentially extracting information elements forming the information sequence from the image data in which the information sequence is embedded, based on the table. 11. A computer-readable recording medium on which a program for causing a computer to execute the method according to claim 7 is recorded.