WO2007049479A1

WO2007049479A1 - Information processing apparatus, information processing method, and program

Info

Publication number: WO2007049479A1
Application number: PCT/JP2006/320578
Authority: WO
Inventors: Nozomi Ishihara; Koki Abe
Original assignee: The University Of Electro-Communications
Priority date: 2005-10-25
Filing date: 2006-10-16
Publication date: 2007-05-03
Also published as: JPWO2007049479A1; JP4834844B2

Abstract

An information processing apparatus, an information processing method and a program wherein digital watermarking information can be effectively embedded. An input part (11) receives a target image (F), a digital watermarking image (W) and a factor selection key (C) required for embedding the digital watermarking image (W). The input part (11) supplies the received target image (F) to a converting part (12) and also supplies the received digital watermarking image (W) and factor selection key (C) to a digital watermark embedding part (13). The converting part (12) subjects the target image (F) supplied from the input part (11) to a wavelet conversion and supplies a factor of the resultant frequency-converted target image (F) to the embedding part (13). The embedding part (13) determines a factor of the lowest subband related to an embedding factor given from the factor selection key (C). The embedding part (13) modulates, based on the determined factor of the lowest subband, the pixels of the digital watermarking image (W) to embed the digital watermarking image (W) in accordance with the intensity of a quantized parameter included in the factor selection key (C).

Description

Specification

INFORMATION PROCESSING APPARATUS, INFORMATION PROCESSING METHOD, AND PROGRAM

Technical field

The present invention relates to an information processing apparatus, an information processing method, and a program, and in particular, distinguishes between tampering and malicious modification such as compression and noise, and restores digital watermark information accurately. The present invention relates to an information processing apparatus, an information processing method, and a program that enable accurate detection of tampering.

Background art

In recent years, as various types of data such as character data, image data, audio data, etc. are handled within computer networks, copyright information is protected to prevent unauthorized use (falsification) of the data. A technology has been developed for embedding user information as digital watermark information.

[0003] Embedding electronic penetration information changes the data to be embedded, so if the embedding strength is increased, the quality of the content represented by the data may be degraded. On the other hand, in order to accurately detect tampering, it is necessary to make the digital watermark information more robust. Therefore, as a technique for strengthening the resistance to attacks on the electronic permeability information that does not cause deterioration in content quality, that is, to increase the embedding strength, a plurality of predetermined coefficients of the wavelet-transformed coefficients are conventionally used. On the other hand, there is known an invention in which a plurality of pixels constituting digital watermark information are embedded, and when extracting the digital watermark information, each pixel value is determined by majority decision (for example, Patent Document 1). .

Patent Document 1: Japanese Patent Application Laid-Open No. 2000-106624

Disclosure of the invention

Problem that invention tries to solve

By the way, in the falsification detection, the electronic watermarking system is correctly restored, and the data is edited, compressed, expanded, noise and so on without malicious modification and the falsification as a malicious attack. It is necessary to distinguish and detect. Here, as characteristics of coefficients in the wavelet domain of image data, it is possible to have a strong resistance to compression, noise, etc. as the level gets. In addition, since the amount of change in coefficient is not so large in compression and noise, etc., the damage of the embedded electronic watermark is in a dispersion tendency, while the damage tends to be concentrated in tampering.

However, in the invention described in Patent Document 1, in the case where strong compression or noise is applied to image data, the present invention described in Patent Document 1 makes uniform majority judgment without considering these features. There has been a problem that the electronic transparency and the damage to the information increase, the majority decision does not work, the electronic transparency can not be restored, and malicious tampering can not be detected.

The present invention has been made in view of such a situation, and distinguishes between malicious tampering and non-malicious change such as compression, and accurately restores electronic transparency information and tampering with it. Can be accurately detected.

Means to solve the problem

An information processing apparatus according to a first aspect of the present invention is an information processing apparatus for detecting falsification of target information in which digital watermark information is embedded by wavelet conversion, and performs conversion such as wavelet conversion of the target information. Means, embedding coefficient detection means for detecting a plurality of embedding coefficients in which a plurality of identical pixels constituting the digital watermark information are respectively embedded, and a plurality of the embedding coefficients detected by the embedding coefficient detection unit V, And counting means for counting according to the sub-band to which each of the embedding coefficients belongs.

In the information processing apparatus according to the first aspect, the determination means for determining the embedding coefficient value which has won the majority as the pixel value constituting the digital watermark information based on the counting result by the counting means is further added. It can be provided.

In the information processing apparatus according to the first aspect, based on the counting result by the counting unit, the embedded coefficient value which loses the majority is regarded as a breaking coefficient, and the wavelet transform unit performs wavelet transform by the wavelet transform unit. In the target information, calculation means for calculating the number of embedding coefficients and the number of breaking coefficients for each group of related coefficients across subbands, the number of embedding coefficients calculated by the calculating means, and the number of breaking coefficients Are weighted according to the embedding factor and the level of the sub-band to which the corruption factor belongs. Determining whether or not the coefficient group has been tampered with, based on a ratio of the weighting unit, the number of the breaking coefficients weighted by the weighting unit, and the embedding coefficient weighted by the weighting unit Means may be further provided, and operation means for operating the coefficient value of the coefficient group based on the determination result by the determination means.

When the ratio is equal to or less than a predetermined threshold value, the operation means erases the breakage coefficient included in the coefficient group, and when the ratio is larger than a predetermined threshold value, the operation unit deletes the broken coefficient group. The embedding factor included can be changed to a breaking factor.

[0013] A plurality of identical pixels constituting the digital watermark information is a second subband associated with an embedding coefficient randomly determined in advance from the coefficient of the first subband in the wavelet-transformed object information. The modulation coefficient may be modulated based on the related coefficient of and embedded in the embedding coefficient.

The first sub-band may be a sub-band other than the lowest sub-band, and the second sub-band may be the lowest sub-band.

[0015] A plurality of stages of wavelet transforms are performed as needed, and each time a stage of wavelet transform is performed, the coefficient power of the first sub-band The second sub-band of the second sub-band associated with the embedding coefficient determined in advance randomly. Related coefficients can be detected.

The first sub-band may be a sub-band other than the highest sub-band, and the second sub-band may be the highest sub-band.

The information processing apparatus according to the first aspect further comprises demodulation means for performing demodulation based on the related coefficient, and the counting means includes the embedding coefficient demodulated by the demodulation means. Each of the embedding coefficients can be weighted and counted according to the sub-band to which it belongs.

[0018] A first information processing method or program according to the present invention is an information processing method for detecting falsification of target information in which electronic wave penetration information is embedded by wavelet conversion. In a program that causes a computer to execute processing for detecting falsification of target information in which penetration information is embedded, a conversion step of performing wavelet conversion on the target information, and a plurality of identical pixels constituting the digital watermark information An embedding coefficient detection step of detecting a plurality of embedding coefficients to be embedded respectively and the plurality of embedding coefficients detected in the processing of the embedding coefficient detection step, weighting according to a sub-band to which each of the embedding coefficients belongs And counting steps.

In the information processing apparatus, the information processing method, or the program according to the first aspect as described above, the target information is wavelet-transformed, and a plurality of identical pixels constituting the digital watermark information are respectively embedded. The embedded coefficients are detected and weighted and counted according to the sub-bands to which the embedded coefficients belong to the plurality of detected embedded coefficients.

Effect of the invention

According to the first aspect of the present invention, digital watermark information can be embedded or extracted efficiently and safely.

Brief description of the drawings

FIG. 1 is a block diagram showing a configuration example of an image processing apparatus to which the present invention is applied.

FIG. 2 is a diagram for explaining wavelet transform.

FIG. 3 is a flowchart for explaining the digital watermark embedding process of embedding unit 13 in FIG. 1.

FIG. 4 is a diagram for explaining the process of step S4 in FIG. 3;

[FIG. 5] Another view for explaining the processing of step S 4 in FIG. 3.

FIG. 6 is another diagram for explaining the process of step S4 in FIG. 3.

7 is a view for explaining the process of step S5 and step S6 of FIG. 3;

FIG. 8 is a block diagram showing a configuration example of an image processing apparatus 41 to which the present invention is applied.

FIG. 9 is a flowchart illustrating digital watermark extraction processing of the tampering detection unit 53 of FIG. 8;

FIG. 10 is a diagram for explaining the processing of steps S23 to S26 in FIG. 9;

FIG. 11 is a view for explaining the process of step S27 and step S28 of FIG. 9;

FIG. 12 is a view showing an example of a tampering detection image.

FIG. 13 is a diagram showing a breakage rate according to weights.

FIG. 14 is a flowchart for describing image processing in the image processing unit 54 of FIG. 8; 15 is a view for explaining the process of step S52 in FIG. 14;

[FIG. 16] Another view for explaining the process of step S 52 in FIG.

FIG. 17 is a view for explaining the process of step S55 in FIG. 14;

FIG. 18A is a diagram showing an example of a target image.

FIG. 18B is a diagram showing an example of a target image.

FIG. 19 is a block diagram showing another configuration example of the image processing apparatus 1 to which the present invention is applied.

FIG. 20 is a flow chart for explaining the electron permeability and embedding process in the embedding section 101 of FIG. 19;

21 is a view for explaining the process of step S84 and step S85 of FIG. 20. FIG.

FIG. 22 is another view showing the breakage rate according to the weight.

FIG. 23 is a view for explaining a method of embedding an electron permeability image.

[FIG. 24A] Another view for explaining the method of embedding an electron permeability image.

FIG. 24B is another view illustrating the method of embedding an electron permeability image.

FIG. 25 is a block diagram showing a configuration example of a computer.

Explanation of sign

1 image processing device, 12 wavelet transform units, 13 digital watermark embedding units, 14 inverse wavelet transform units, 52 wavelet transform units, 53 tamper detection units, 101 digital watermark embedding units

BEST MODE FOR CARRYING OUT THE INVENTION

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

FIG. 1 shows a configuration example of an image processing apparatus 1 to which the present invention is applied. The image processing apparatus 1 embeds a digital watermark image W of a binary image as digital watermark information into a target image F (for example, an image captured by a digital camera) to which the digital watermark image W is to be embedded. Can.

The input unit 11 receives a target image F, a digital watermark image W, and a coefficient selection key C necessary for embedding the digital watermark image W. The coefficient selection key C indicates the embedding number (replication number) of one pixel constituting the digital watermark image W, the coefficient of the target image F to which the digital watermark image W is to be embedded (hereinafter referred to as embedding coefficient), and the embedding strength. Quantization parameter Data is included.

[0026] The embedding coefficients are also predetermined at random in advance of the subbands LL1, HH1, LH1, HL2, HH2, LH2, HL3, HH3, or LH3 excluding the subband LL3. For example, in the example of FIG. 4, the subbands HL1 and LH1 of level 1 to the coefficient P1 and the coefficient P2 power to the subband HH2 of level 2 to the coefficient P3 force and the subband LH3 to level 3 of the coefficient 3 are respectively Determined as the embedding coefficient.

The input unit 11 supplies the input target image F to the wavelet transform unit 12, and supplies the input electronic watermark image W and the coefficient selection key C to the digital watermark embedding unit 13.

The wavelet transform unit 12 performs wavelet transform on the target image F supplied from the input unit 11, and the resulting coefficient of the frequency-converted target image F is converted into an electronic watermark embedding unit 13. Supply to

[0029] In the wavelet transformer 12, the two-dimensional level distribution of the image is frequency-transformed into data that can be easily compressed.

Specifically, the target image F is decomposed into a plurality of frequency bands by repeating the process of dividing the pixel data of the target image F into high frequency components and low frequency components.

For example, in the case of performing three steps of dividing the high frequency component and the low frequency component into two steps, for example, the process is performed on the pixel data of the target image F in A of FIG. As shown in B of FIG. 2, sub-band LL1 of "low-pass I low-pass", sub-band L H1 of "low-pass I high-pass", sub-band HL1 of "high-pass I low-pass", and It is decomposed into the sub-band ΉΗ 1 of “high range I high range”. The subbands LL1, LH1, HL1, or HH1 obtained by this processing are referred to as level 1 subbands as appropriate.

Next, the second process is performed on the lower band subband LL1, and the lower band LL1 is the “lower band I lower band” subband LL2, as shown in C of FIG. It is decomposed into the low band I high band sub-band LH2, the high band I low band sub-band HL2, and the high band I high band sub-band HH2. The subbands LL2, LH2, HL2, or HH2 obtained by this processing are appropriately referred to as level 2 subbands.

Then, the third process is performed on the lower band subband LL2, and the lower band LL2 power as shown in D of FIG. 2 indicates that the “lower band I lower band” subband LL3, “lower band Sub-bar of It is decomposed into the band LH3, the high band I low band sub-band HL3, and the high band I high band sub-band HH 3. The subbands LL3, LH3, HL3 or HH3 obtained by this processing are appropriately referred to as level 3 subbands.

Returning to FIG. 1, the electron permeability embedded unit (hereinafter referred to as embedded unit) 13 supplies the coefficients for one frame of the target image F supplied from the wavelet conversion unit 12 from the input unit 11. Similarly, the supplied digital watermark image W is embedded according to the coefficient selection key C supplied from the input unit 11, and the resulting coefficient is supplied to the inverse wavelet transform unit 14. The embedding unit 13 also supplies, to the output unit 15, a coefficient selection key C necessary for extracting the electronically transparent image W.

[0035] The inverse wavelet transform unit 14 performs inverse inverse one-Bretlet transform on the coefficients supplied from the embedding unit 13 (performs the inverse process to the process shown in FIG. 2), a digital watermark A target image F (target image F ′) in which the image W is embedded is generated and supplied to the output unit 15.

The output unit 15 outputs the target image F ′ supplied from the inverse wavelet transform unit 14 and the coefficient selection key C supplied from the embedding unit 13 to an external device (for example, the image processing device 41 of FIG. 8). Supply to

Next, the digital watermark embedding process in the embedding unit 13 of the image processing apparatus 1 will be described with reference to the flowchart of FIG.

In step S1, the embedding unit 13 inputs the coefficient (D in FIG. 2) of the target image F for one frame from the wavelet transformation unit 12.

In step S 2, the embedding unit 13 selects one pixel of the digital watermark image W supplied from the input unit 11.

Next, in step S3, the embedding unit 13 calculates the pixel value of the pixel of the digital watermark image W selected in step S2 (hereinafter, referred to as a transparency bit) from the coefficient supplied from the input unit 11. Replicate as many copies as indicated in the selection key C.

Next, in step S4, the embedding unit 13 detects a coefficient (related coefficient) on the subband LL3 associated with the embedding coefficient given by the coefficient selection key C.

[0042] In the wavelet transform coefficients, the coefficient of coordinates (x, y) on the subband of level (r = 1, 2) and the coordinate of the subband on level r + 1 (xZ2, yZ2) And the nature of the factor Is known to have (ie, high spatial correlation) (hereinafter this relationship is referred to as a parent-child relationship).

Therefore, in the present invention, using this parent-child relationship, as shown in FIG. 5, from the embedding coefficient of the coordinates (xl, yl) of the level 1 subband, the coordinates (xlZ4, Coefficient force of ylZ4) From the embedding coefficient of the level 2 sub-band (x2, y2), the coefficient of the sub-band LL3 coordinate (x2Z2, y2Z2) and the level 3 sub-band (except for the sub-band LL3) From the embedding coefficients of the coordinates (x3, y3) of (1), the coefficient forces of the coordinates (x3, y3) of the subband LL3 are detected as the coefficients of the subband LL3 associated with the embedding coefficients.

This parent-child relationship can be shown in FIG.

Further, in the example of FIG. 4, the coefficient Q1 is from the coefficient P1, the coefficient Q2 is from the coefficient P2, the coefficient Q is from the coefficient P3, and the coefficient Q4 is from the coefficient P4. It shall be detected as a factor of 3.

Next, in step S5, the embedding unit 13 detects, for example, three coefficients adjacent to the coefficient for each of the coefficients of the subband LL 3 detected in step S4, and The average value of the coefficient values of four coefficients (hereinafter, these four coefficients are appropriately referred to as corresponding coefficients) of the coefficients and the coefficients of the subband LL3 detected in step S4 is calculated. Quantize and convert to a binary value (1 or 0).

For the coefficient Q1 in FIG. 4 (for example, coordinates (xql, yql)), as shown in FIG. 7, coordinates (xql, yql + l) adjacent to the coefficient Q1, (xql + 1, yql), And three coefficients of (xql + 1, yql + 1) are detected, and the average value of the coefficient value of the coefficient Q1 and its three coefficients is calculated, and it is quantized to obtain a binary value rl. Be For coefficient Q2 (for example, coordinates (xq2, yq2)), three coefficients of coordinates (xq2, yq2 + 1), (xq2 + 1, yq2), and (xq2 + 1, yq2 + 1) are detected , The average value of coefficient values of coefficient Q2 and its three coefficients is calculated, and it is quantized to obtain binary value r2.

In addition, three coefficients, (xq3, yq3 + l), (xq3 + 1, yq3), and (xq3 + l, yq3 + 1), for the coefficient Q3 (for example, coordinates (xq3, yq3)) The coefficient of is detected, the average value of coefficient values of coefficient Q3 and its three coefficients is calculated, and it is quantized to obtain binary value r3. Factor Q4 ( For example, for the coordinate (xq4, yq4), three coefficients of the coordinates (xq4, yq4 + 1), (xq4 + l, yq4), and (xq4 + l, yq4l) are detected, and the coefficient Q4 and its 3 The average of the coefficient values of the coefficients is calculated and quantized to obtain the binary value r4.

Next, in step S6, the embedding unit 13 performs an exclusive OR operation (XOR) of each of the watermark bits copied in step S3 with each of the binary values of the average value obtained in step S5. (Ie, modulate the transmission bit with the binary value of the average value obtained in step S5), and include the resulting value (hereinafter referred to as the modulation bit) in the coefficient selection key C. Embed according to the strength of the quantization step.

In the example of FIG. 7, the modulation bit wal obtained as a result of exclusive OR operation (XOR) of the watermark bit wa and the value rl is taken as the coefficient value of the coefficient P1, and the watermark bit wa and the value r2 The modulation bit wa2 obtained as a result of XOR is taken as the coefficient value of the coefficient P2, and the modulation bit wa3 obtained as a result of XOR of the watermark bit wa and the value r3 is taken as the coefficient value of the coefficient P3. The modulation bit wa4 obtained as a result of XOR of wa and the value r4 is taken as the coefficient value of the coefficient P4.

In step S 7, embedding unit 13 determines whether or not all the pixels of digital watermark image W have been selected, and if it is determined that there are still unselected pixels, step 13 is performed. The process returns to step S2 to select the next pixel, and the processes after step S3 are similarly performed.

If it is determined in step S7 that all the pixels have been selected, the embedding unit 13 ends the digital watermark embedding process.

The above-described processing is performed each time a target image F in which the digital watermark image W is to be embedded is input.

As described above, when embedding digital watermark information, embedding coefficients are determined randomly (step S3), it becomes difficult for a third party to specify the embedding coefficients. For example, in the method of Patent Document 1, since the coefficient for embedding the electronic penetration information is the nth largest absolute value and limited to the coefficient, it is easy to identify the coefficient in which the digital watermark information is embedded,

Further, as described above, since the embedding process is performed on the embedding coefficients determined at random (steps S4 to S6), the embedding process can be performed efficiently. For example, in the method of Patent Document 1, in order to detect the nth coefficient whose absolute value is n, all coefficients are Because it needs to be scanned, it takes time for processing.

In addition, based on the relationship between the electronic transparency report and the coefficient in which the digital watermark information is embedded, the coefficient value is manipulated, and the electronic force is embedded in the heat force, and the information is embedded, In the case of so-called “collage attack,” the coefficient value is manipulated based on the relationship between the digital watermark information and the coefficient in which the digital watermark information is embedded, and falsification is performed. It is effective to associate the coefficient in which the digital watermark information is embedded and the coefficient at another position in any relation. For example, check the magnitude of the coefficient into which the electron permeability information is embedded and the coefficient next to it, and if the coefficient into which the electron permeability information is embedded is larger, 0 is set, and if the next coefficient is large, 1 is set. It is a method to embed values.

[0057] Thus, as described above, the embedding coefficient and the lower band subband (in this example, subband LL3) associated with it are correlated to determine the value (modulation bit) to be embedded (step Steps S4 to S6) can protect the Collage Attack.

Next, referring to FIG. 8, image processing for extracting the digital watermark image W from the target image F (target image F ′) in which the electronic permeability image W is embedded by the image processing apparatus 1 of FIG. The apparatus 41 will be described.

The target image F ′ and the coefficient selection key C for extracting electronic transparency information output from the image processing apparatus 1 of FIG. 1 are input to the input unit 51.

The input unit 51 supplies the input target image F ′ to the wavelet transform unit 52, and supplies the input coefficient selection key C to the tampering detection unit 53 and the image processing unit 54.

The wavelet transform unit 52 subjects the target image F ′ supplied from the input unit 51 to the same wavelet transform as the wavelet transform unit 12 shown in FIG. 1 (FIG. 2), and one frame obtained as a result The coefficients of the target image F ′ for the minute are supplied to the tampering detection unit 53 and the image processing unit 54.

The tampering detection unit 53 extracts the electronically transmitted image W based on the coefficient selection key C from the coefficients of the target image F ′ supplied from the wavelet transform unit 52, and is obtained at that time. The falsification information is supplied to the image processing unit 54.

The image processing unit 54 compares the coefficient of the target image F ′ supplied from the wavelet transformation unit 52 with the coefficient. Then, the coefficient operation is performed according to the coefficient selection key C supplied from the input unit 51 and the tampering information supplied from the tampering detection unit 53.

Next, extraction processing of the digital watermark image and the tampering detection image in the tampering detection unit 53 will be described with reference to the flowchart in FIG.

In step S21, the tampering detection unit 53 inputs the coefficient of the target image F ′ for one frame from the wavelet transform unit 52.

Next, in step S22, the tampering detection unit 53 reads the coefficient selection key for one pixel of the digital watermark image W from the coefficient selection key supplied from the input unit 51.

In step S23, the tampering detection unit 53 determines, based on the read coefficient selection key, the embedding coefficient (one pixel selected in step S22) from the coefficients for one frame input in step S21. Detection of embedded coefficients in which a bit is embedded.

For example, in the case of extracting an electronic watermark image W from a target image F in which a digital watermark image W is embedded as shown in FIG. 7, level 1 as in the case of FIG. 7 as shown in FIG. The coefficient P1 of subband HL1 and the coefficient P2 of subband LH1 are detected as the embedding coefficients, respectively, the coefficient 2 of level 2 to the coefficient P3 and the coefficient of level 3 subband LH3 to the coefficient P4. Ru.

Next, in step S24, the tampering detection unit 53 detects the coefficient of the sub-band LL3 related to the detected embedding coefficient based on the parent-child relationship of the wavelet transform.

In the example of FIG. 10, as in FIG. 7, coefficients P1 to Q1 are related, coefficients P2 to Q2 are related, coefficients P3 to Q3 are related, and coefficients P4 to Q4 are related, respectively. It is detected as a coefficient of subband LL3.

In step S 25, for each of the coefficients of subband LL 3 detected in step S 24, tampering detection unit 53 detects, for example, three coefficients adjacent to the coefficient, and the three coefficients thereof. And the coefficients of the sub-band LL3 detected in step S24, the average value of the coefficient values of four coefficients (corresponding coefficients) is calculated, and the average value is quantized and converted into a binary value.

In the case of the example of FIG. 10, as in the case of FIG. 7, the coordinates (xql, yql + l), (xql + 1, yql) for the coefficient Q1 (coordinates (xql, yql)), And three coefficients of (xql + 1, yql + 1) are detected, and the average value of the coefficient value of the coefficient Q1 and its three coefficients is calculated and quantized. The binary value rl is obtained. For coefficient Q2 (coordinate (xq2, yq2)), three coefficients of coordinates (xq2, yq2 + 1), (xq2 + l, yq2), and (xq2 + l, yq2 + 1) are detected, The mean value of the coefficient value of the coefficient Q2 and its three coefficients is calculated and quantized to obtain the binary value r2 '.

In addition, three coefficients of coordinates (xq3, yq3 + 1), (xq3 + 1, yq3), and (xq3 + l, yq3 + l) for the coefficient Q3 (coordinates (xq3, yq3)) Is detected, and the average value of coefficient Q3 and the coefficient of its three coefficients is calculated, and it is quantized to obtain binary value _r3 '. For coefficient Q4 (coordinate (xq4, yq4)), three coefficients of coordinates (xq4, yq4 + l), (xq4 + l, yq4) and (xq4 + l, yq4 + 1) are detected, and the coefficient Q4 is obtained. The mean value of the coefficient values of the three coefficients is calculated and quantized to obtain a binary value r4 '.

In step S26, the tampering detection unit 53 calculates each of the embedded coefficient coefficient values (modulation bit (step S6 in FIG. 3)) detected in step S23 and the corresponding average value obtained in step S25. XOR with each of the binary values (that is, demodulate the modulation bit with the binary value of the average value obtained in step S25), and obtain the resulting value (hereinafter referred to as demodulation bit) as In step S22, it is set as a candidate for the pixel value (penetration bit) of the pixel of the electron-permeable image W selected.

In the case of the example of FIG. 10, the demodulation bit wbl obtained as a result of XOR of the modulation bit wal, and the value rl, the demodulation bit wb2 obtained as a result of XOR of the modulation bit wa2 and the value r2. Demodulated bit wb3 obtained as a result of XOR of modulation bit wa3 and value r3, and force bit candidate for demodulation bit wb4 obtained as a result of XOR of modulation bit wa4 and value r4 It is assumed.

Next, in step S 27, the tampering detection unit 53 counts and counts the demodulation bits as the watermark bit candidate obtained in step S 26 according to the level of the sub-band to which the demodulation bits belong, and Based on the results, a majority is taken to determine the winning demodulation bits (hereinafter referred to as winning bits) and the losing demodulation bits (hereinafter referred to as losing bits).

For example, it is assumed that the weight of the demodulation bits of the subbands belonging to level 1 is 1, the weight of the demodulation bits of the subbands belonging to level 2 is 6, and the demodulation bits of the subbands belonging to level 3 Assuming that each weight is 8 and the number of demodulation bits multiplied by each weight is prepared, among them, many bits and few bits are detected. In the case of the example of FIG. 10, when the demodulation bits wbl, wb4, wb3, wb2 are 0, 0, 0, 1 as shown in FIG. 11, the value 0 as the level 1 demodulation bit wbl 1, level 3 demodulation bit wb4 value 0 0, level 2 demodulation bit wb3 value 0 6, level 1 demodulation bit wb2 value 1 1 or 15 There are 0 values and 1 value.

Therefore, in the case of this example, since the value 0 is 15 and the value 1 is 1, the value 0 is detected as a winning bit and the value 1 is detected as a losing bit.

Next, in step S28, as shown in FIG. 11, the tampering detection unit 53 selects the winning bit (0) detected in step S27 as the pixel value of the pixel of the image W which has been subjected to electron transparency selected in step S22. And the loss bit (1) is a coefficient corresponding to the coefficient P2 of the target image F 'of a white image (hereinafter referred to as an image for detecting falsification) M' prepared in advance with three stages of wavelet transform processing. Let it be a coefficient value of P2 '.

In step S29, the tampering detection unit 53 determines in step S22 whether the coefficient selection key for all the pixels of the digital watermark image W has been read or not, and there is a coefficient selection key that has not been read yet. If it is determined that there is, the process returns to step S22, the coefficient selection key of the next pixel is read out, and the processes after step S23 are performed.

If it is determined in step S29 that the coefficient selection keys of all the pixels have been read out, the process proceeds to step S30, the digital watermark image W is restored, and a tampering detection image M ′ is generated. The tampering detection unit 53 supplies the image processing unit 54 with the tampering detection image M ′. The tampering detection image M obtained by performing the inverse wavelet transform process on the tampering detection image M ′ is displayed as shown in FIG. 12, for example.

As described above, the electron penetration information is extracted.

As the features of the wavelet transform coefficients of the image, the deeper the level, the stronger the compression and noise!耐性 Because it is resistant, it is embedded in the deep, level electronic penetration information is strong in compression and noise V ヽ. Therefore, if the electronically transparent image W and the image for tampering detection M ′ embedded in the deep level are broken, it can be considered that it is due to tampering such as cutting and pasting processing that is not compressed or noise.

Therefore, as described above, the watermark information embedded in depth and level is shallow and level. The weight of the demodulation bit of the subband belonging to level 1 is 1 and the weight of the demodulation bit of the subband belonging to level 2 is 6 Since the weight of the demodulation bit of the sub-band belonging to level 3 is 8) and the influence of information embedded in the deep level is large, the detection of the transmission bit is performed. The electronic watermark can be detected more accurately.

[0086] FIG. 13 shows the weight applied to the level 2 watermark shown on the X axis and the weight applied to the level 3 watermark shown on the Y axis when the target image F is compressed with JPEG compression at a compression ratio of 60%. It shows the damage rate when used. In this example, when the weight of level 2 is 6 and the weight of level 3 is 8 to 10, the lowest damage rate of electron permeability information is obtained.

Next, the processing of the image processing unit 54 of the image processing apparatus 41 will be described with reference to the flowchart of FIG.

In step S 51, the image processing unit 54 sets each of the level 3 subbands LH 3 and HL 3 in the coefficients of the target image F for one frame supplied from the wavelet transform unit 52 to, for example, 4 × 4 pixels. Divide into blocks of

Next, in step S 52, the image processing unit 54 applies, for each of the blocks of subband LH 3 and subband HL 3 obtained in step S 51, subband LH 2 and subband HL 2 of level 2 corresponding thereto. , And blocks of level 1 subband LH1 and subband HL1 based on the parent-child relationship of wavelet transform

For example, a block of level 3 subband LH 3 or subband HL 3 can be identified by a predetermined block number, and level 2 subband LH 2 and subband HL 2 corresponding to the block, and level 1 sub The blocks of band LH1 and subband HL1 shall also be identified by the same block number as the corresponding blocks of subband LH3 or subband HL3.

That is, each block can be identified, for example, by a code bk (i) where level is k and a block number is i, and a block group further represented by the same code bk (i) is labeled as Bk It can be expressed as a block belonging to (i). For example, the block of level 3 subband LH3 and subband HL3 shown in FIG. 15 is block b3 (l), and the corresponding block of level 2 subband LH2 and subband HL 2 is block b2 ( l) block of level 1 sub-band LH1 and sub-band HL1 is block bl (l), and as shown in FIG. 16, block 1 of level 1 sub-band LH 1 and block HL of HL1 Let the block group represented be block group Bl (l), the block group represented by block 2 of level 2 subband LH2 and HL 2 block b2 (l) be block group B2 (1), and level 3 sub A block group represented by block b3 (l) of bands LH3 and HL3 is set as a block group B3 (l).

Next, in step S53, the image processing unit 54 selects one block number i, and in step S54, based on the coefficient selection key C, the block group BIG), B2 (i), B3 ( For each of the blocks belonging to i), calculate the number of embedding coefficients included in it. Further, the image processing unit 54 embeds the loss bit included in each of the blocks belonging to the block groups Bl (i), B2 (i) and B3 (i) based on the image M for tampering detection. Calculate the number of coefficients (hereinafter referred to as the breakage coefficients).

Next, in step S55, the image processing unit 54 sums up the number of embedding coefficients and the number of breaking coefficients of each block for each of the block groups Bl (i), B2 (i), B3 (i). And multiply the weight according to the level.

For example, as shown in FIG. 17, the sum of the number of embedding coefficients included in blocks belonging to block group B3 (i) and the sum of the number of corruption coefficients included in blocks belonging to block group B3G) Is multiplied by eight.

Block group A total of the number of embedding coefficients included in a block belonging to B2 (i), and a block group

Each of the sums of the number of breaking coefficients included in the block belonging to B2G) is multiplied by six.

Each of the sum of the number of embedding coefficients included in the block belonging to block group BIG and the total number of corruption coefficients included in the block B group 1 (block belonging to 0) is multiplied by one. .

In step S56, the image processing unit 54 calculates the sum of the breaking coefficients obtained by the weighting in step S55 as the embedding coefficient obtained by the weighting in step S55, as shown in equation (1). Divide by the sum. [0099] [Number 1]

Sum of the number of broken coefficients of B3 (i) x 8

Sum of the number of breaking coefficients of + B2 (i) x 6

+ Sum of the number of breaking coefficients of B1 (i) x 1

...)

Total number of embedding coefficients in B3 (i) x 8

+ Total number of embedding coefficients in B2 (i) x 6

+ Sum of the number of embedding coefficients in B1 (i) x 1

Next, in step S57, the image processing unit 54 determines whether or not the value obtained as a result of the calculation in step S56 (hereinafter referred to as the tampering rate) is equal to or less than a predetermined threshold, If it is determined that the following is true, it is determined that the block group of the block number i is not falsified, and the coefficient value (loss bit) of the corruption coefficient included in the target block is determined in step S58. Delete (hereinafter, this process is called "! /, Process").

On the other hand, if it is determined in step S57 that the block group is larger than the predetermined threshold value, the block group of the block number i is considered to be falsified, and the process proceeds to step S59 and the image processing unit 54 determines all coefficients belonging to the block Is changed to the failure factor (hereinafter this process is referred to as emphasizing process).

As described above, one threshold may be provided, or an upper threshold and a lower threshold may be provided. If the threshold is lower than the lower threshold, sieving is performed and the upper threshold is determined. In the above case, emphasis processing may be performed. In this case, if the tampering rate is between the upper threshold and the lower threshold, neither the! / ヽ nor the emphasizing is performed.

When the destruction coefficient is erased in step S58, or when the coefficient of the block group of block numbers is changed in step S59, the process proceeds to step S60, and the image processing unit 54 performs the block B in step S53. It is determined whether or not the number i is selected, and if it is determined that there is still! /,! Or! /! Block number, the process returns to step S53, the next block is selected, and step S54 or later is performed. Perform the same process.

If it is determined in step S60 that all block numbers i have been selected, the process ends.

Image processing is performed as described above. As described above, since the sieving process and the emphasizing process are performed based on the falsification rate, when the target image F is as shown in FIG. 18A, an image corresponding to the falsified block as shown in FIG. 18B. Can be highlighted.

[0107] Since compression and noise have a large amount of correction of pixel coefficients, damage to digital watermark information tends to be dispersed, but if there is tampering such as clipping or pasting, the image of the tampering The coefficients tend to be concentrated and corrupted. Therefore, by examining the density (alteration rate) of the corrupted electronic watermark information as described above and comparing it with a threshold value, it is possible to appropriately determine the damage due to compression, noise, or tampering. .

FIG. 19 shows another configuration example of the image processing apparatus 1. In this image processing apparatus, a digital watermark embedding unit 101 may be provided instead of the digital watermark embedding unit 13 of FIG. The other parts are the same as in FIG.

Although the embedding process in embedding section 13 in FIG. 1 associates the embedding coefficient with the coefficient of the lowest band (subband LL 3), embedding section 101 substitutes the embedding coefficient in the lowest band. The embedding process is performed in association with the coefficients of the same level sub-band 同一.

Digital watermark embedding processing in embedding section 101 will be described with reference to the flowchart in FIG.

In step S 81, the embedding unit 101 performs the first wavelet transformation on the target image F of one frame from the wavelet transformation unit 12, and the second wavelet transform is further performed. When, enter the resulting coefficient (C in Figure 2).

In step S 82, the embedding unit 101 selects one pixel of the digital watermark image (binary image) W.

Next, in step S 83, the embedding unit 101 uses the coefficient selection key supplied from the input unit 11 as the pixel value (watermark bit) of the pixel of the electronically transparent image W selected in step S 82.

Replicate as many copies as shown in C.

Next, in step S 84, embedding section 101 detects the coefficient of level 2 subband 2 associated with the embedding coefficient of level 2 subband HL 2 and subband LH 2 given by coefficient selection key C. . For example, when the coefficient Pll (coordinates (xqll, yqll)) of the subband LH2 of level 2 in FIG. 21 is the embedding coefficient, the coefficient Q11 of the coordinates (xqll, yqll) of the subband HH2 is the coefficient Pll. Detected as a coefficient of the sub-band HH2 associated with

Next, in step S85, the embedding unit 101 detects, for example, three coefficients adjacent to the coefficient of the sub-band HH2 detected in step S84, and the three coefficients, In step S84, the average value of the coefficient values of a total of four coefficients (corresponding coefficients) with the coefficients of the subband HH2 detected is calculated, and it is quantized and converted into a binary value (1 or 0).

For the coefficient Q11 (coordinate (xqll, yqll)) in FIG. 21, the coordinates (xqll, yqll + l), (xqll + 1, yqll), and (xqll + 1, yqll +) adjacent to the coefficient Q11 Three coefficients of 1) are detected, the average value of coefficient Q11 and coefficient values of the three coefficients is calculated, and it is quantized to obtain a binary value.

Next, in step S 86, embedding section 101 performs an exclusive OR operation (XOR) of the watermark bit copied in step S 83 and the binary value of the average value obtained in step S 85 ( That is, the watermark bit is modulated with the binary value of the average value obtained in step S85), and the resulting value (modulation bit) is determined according to the strength of the quantization step included in the coefficient selection key C. To embed.

In step S 87, embedding unit 101 determines whether all the pixels of digital watermark image W have been selected or not, and if it is determined that there are any pixels not selected yet, step S 82. Return to step S83, select the next pixel, and execute the processing from step S83 onwards in the same way.

Note that since it is not necessary to embed all the watermark bits in the level 2 sub-band, when watermark bits not to be embedded in the level 2 sub-band are selected, the processing in steps S 83 to S 86 is performed as appropriate. It will be skipped.

If it is determined in step S87 that all the pixels have been selected, the embedding unit 101 proceeds to step S88, and determines whether all coefficients for one frame of the target image F have been input. , If it is determined that the whole has not been input, the process returns to step S81.

In the present case, the embedding unit 101 performs the wavelet transformation unit 12 in step S 81. When the third wavelet transform is performed on the target image F of one frame, input the coefficient (D in Fig. 2) obtained as a result. The processing in the case of inputting this coefficient will be described continuously.

That is, in step S82, one pixel of the digital watermark image W is selected.

Next, in step S83, the pixel value (watermark bit) of the pixel of the digital watermark image W selected in step S82 is replicated by the number of copies indicated by the coefficient selection key C supplied from the input unit 11. .

Next, in step S84, the coefficient of level 3 subband HH3 associated with the embedding coefficient of level 3 subband HL3 and subband LH3 given by coefficient selection key C is detected.

For example, if coefficient P21 (coordinates (xq21, yq21)) of subband 3 of level 3 in FIG. 21 is the embedding coefficient, coefficient Q21 of coordinates (xq21, yq21) of subband HH3 is detected. .

Next, in step S85, for example, with respect to the coefficients of subband HH3 detected in step S84, three coefficients adjacent to the coefficient are detected, and these three coefficients and step are also included. The average value of the coefficient values of a total of four coefficients (corresponding coefficients) with the coefficients of subband ΉΗ3 detected in S84 is calculated, and it is quantized and converted into a binary value (1 or 0).

For the coefficient Q21 (coordinate (xq21, yq21)) in FIG. 21, the coordinates (xq21, yq21 + l), (xq21 + l, yq21), and (xq21 + 1, yq21 +) adjacent to the coefficient Q21 are used. Three coefficients 1) are detected, the average value of coefficient Q21 and the coefficient values of the three coefficients is calculated, and it is quantized to obtain a binary value.

Next, in step S86, exclusive OR operation (XOR) is performed between the watermark bit copied in step S83 and the binary value of the average value obtained in step S85 (ie, transparency). Bit Force Modulated with the binary value of the average value obtained in step S85), and embeds the resulting value (modulation bit) according to the strength of the quantization step included in the coefficient selection key C.

In step S 87, it is determined whether all the pixels of the digital watermark image W have been selected. If it is determined that there are any pixels not selected yet, the process returns to step S 82. Are selected, and the processes after step S83 are similarly performed. Since it is not necessary to embed all the watermark bits in the level 3 sub-band, when watermark bits not to be embedded in the level 3 sub-band are selected, the processes in steps S 83 to S 86 are appropriately performed. It will be skipped.

If it is determined in step S87 that all the pixels have been selected, the process proceeds to step S88, in which it is determined that all the coefficients of one frame of the target image F have been input, and the process ends.

In embedding section 13 in FIG. 1, the embedding process is performed in such a manner that the embedding coefficient is associated with the lower band subband (in this example, subband LL3) associated therewith. The lowest subband is formed after all stages of wavelet transform processing have been performed, as shown in FIG. 2D.

That is, the embedding process in the embedding unit 13 of FIG. 1 had to be performed after the completion of the wavelet process in the wavelet transformation unit 12.

On the other hand, embedding section 101 of FIG. 19 substitutes the coefficients of the sub-bands 同一 at the same level as the embedding coefficients, instead of the lowest band sub-bands, as described above. Since embedding was performed when wavelet coefficients were obtained (since the wavelet transform and embedding were performed in parallel), it should be performed more quickly than the embedding in the case of FIG. Can.

[0136] Note that the sub-band ΉΗ used here is the weakest band against tampering, which leads to an increase in the breakage rate. For example, even if the embedding coefficient of the subband HL or LH is not broken, if the coefficient of the subband HH associated with it is greatly changed, it becomes broken. Therefore, in this case, Level 1 was not used for embedding, and Level 2 and Level 3 were embedded.

The operation of the image processing device 41 in the case of extracting the digital watermark image W from the target image F (target image F ′) in which the digital watermark image W is embedded by the image processing device 1 of FIG. The description is omitted because it is as shown in FIG.

FIG. 22 shows that the target image F is compressed when the digital watermark image W is extracted from the target image F (target image F ′) in which the digital watermark image W is embedded by the image processing device 1 of FIG. Indicates the level 2 watermarking weight shown on the X axis and the corruption rate using the level 3 watermarking weight shown on the Y axis when compressed with a 60% JPEG compression. There is. In this example, when the level 2 weight is 1 and the level 3 weight is 5 or the level 2 weight is 2 and the level 3 weight is 9 or 10, the lowest damage rate is obtained.

In the above, embedding of the digital watermark image W into the target image F is obtained, for example, as a result of exclusive OR operation (XOR) of the average value of the corresponding coefficients of the sub-band LL3 and the permeability bit. Although the modulation bits are embedded according to the strength of the quantization parameter (steps S4 to S6 in FIG. 3), they can be embedded as described below.

Here, it is assumed that a wavelet image of the target image F is represented by Expression (2). In equation (2), k is represented by equation (3), 1 is a level, and (m, n) represent the position (coordinates) of a pixel.

[Number 2]

f i, k n, n) '' '(2)

[Number 3]

k = {lh, hl, hh} (LH, HL, HH bands) ■ ■ ■ (3)

In embedding of the electron permeability image W, for example, f (m, n) (Formula (2)) is mapped to a binary value using the quantization function Q defined by the formula (4). Forces performed by The l, k described here

Embedding is also performed so that f (m, n) after embedding becomes the center value of the section represented by equation (5).

l, k

It is

Picture

Q (f) =

'0, if Γ Δ <f <(r + 1) A, r = 0, ± 2,' ",

1, if rA <f <(r + 1) A, r = ± 1, ± 3, ... 丄 ■ ■ ■ (4)

[Number 5]

mod F = | f i, k (m, n) I mod Δ ■ ■ ■ ■ (5)

In the equation (4), Δ is a quantization parameter. The i-th pixel of the digital watermark image w is expressed by equation (6).

[Number 6]

w (i), i = 0, Nw-1 ■ ■ ■ (6)

For example, for the quantization function Q (equation (4)), when the equation (7) holds, the following rules are used: , F (m, n) satisfying equation (5) is changed to the center of the interval.

l, k

[Number 7]

Q (fi, _k (m, n)) = w (i) ■ ■ ■ (7)

Specifically, when equation (8) holds, f (m, n) is obtained by equation (9).

l, k

[Number 8]

fi, k (m, n)> 0 ■ ■ ■ (8)

[Equation 9] f (m, n): = f i, k (m, η) + mod F) ■ ■ ■ (9)

The change of f (m, n) based on this rule is schematically shown in FIG.

l, k

Further, when equation (10) holds, f (m, n) can be obtained by equation (11).

l, k

[Number 10]

fi, k (m, n) <0 ■ ■ ■ (10)

[Equation 11] f (m, n): = f i, k (m, n)-(one mod F) ■ ■-(11)

Further, when the quantization function Q (Eq. (4)) and the equation (12) hold, f (m, n) becomes f (m, n) such that the equation (13) holds, using the following rules: Be changed.

l, k

[Number 12]

Q (fi, k (m, n)) w (i) ■ ■ ■ (12)

[Number 13]

Q (fi, _k (m, n)) = w (i) ■ ■ ■ (13)

Specifically, when equation (14) holds, f (m, n) can be obtained by equation (15).

l, k

[Number 14]

fi, k (m, n)> 0 ■ ■ ■ (14)

[Number 15] k (m, n): =

Further, when equation (16) holds, f (m, n) can be obtained by equation (17).

l, k

[Equation 16]

f i, k (m, n) <0 ■ ■ ■ (16)

[Number 17]

f i, k (m, n): =

In the case based on this rule, a change of f (m, n) when f (m, n)> 0 is schematically represented by l, k l, k

The results are as shown in FIGS. 24A and 24B. FIG. 24A shows the case where equation (18) and FIG. 24B shows the case where equation (19) holds.

[Equation 18] modi ≥ (18)

2

[Number 19] modF (19)

2

As in:, l, k such that f (m, n) (Eq. (2)) after embedding becomes the center value of the interval represented by

In addition, by embedding the digital watermark image W in the target image F, it is possible to reduce the influence on the image quality of the target image F due to the digital watermark image W being embedded, as well as to an attack such as compression. It can also be strong.

The series of processes described above may be executed by hardware, or software It can also be executed by When a series of processes are to be executed by software, it is possible to execute various functions by installing a computer incorporated in hardware dedicated to the program power constituting the software, or various programs. It is installed from a program recording medium to a possible personal computer, for example.

FIG. 23 is a block diagram showing an example of a configuration of a personal computer that executes the series of processes described above according to a program. A central processing unit (CPU) 201 executes various types of processing in accordance with a program stored in a read only memory (ROM) 202 or a storage unit 208. The RAM (Random Access Memory) 203 appropriately stores programs and data to be executed by the CPU 201. The CPU 201, the ROM 202, and the RAM 203 are connected to one another by a bus 204!

Program storage media for storing programs installed in a computer and made executable by the computer are, as shown in FIG. 23, a magnetic disk (including a flexible disk), an optical disk (CD-ROM (a flexible disk), and the like. Compact Disc-Read Only Memory (DV) (including Digital Versatile Disc (DV D)), magneto-optical disc, or removable media 211, which is a package medium consisting of semiconductor memory, etc., or the program temporarily or permanently stores The ROM 202 and the hard disk constituting the storage unit 208 are included. The program may be stored in the program storage medium via a wired or wireless communication medium such as a local area network, Internet, digital satellite broadcast, or the like via the communication unit 209 which is an interface such as a router or a modem, if necessary. It is done using.

Note that in the present specification, the steps of describing the program stored in the program recording medium are not limited to processing performed chronologically in the order described, but necessarily processing chronologically If not, it also includes processes executed in parallel or individually.

Further, the image processing apparatus 1 and the image processing apparatus 41 may be integrally configured by, for example, a personal computer, or the image processing apparatus 1 is mounted on a digital camera, and the image processing apparatus 41 is personal. Configure them separately, such as mounting on a computer May be In particular, when the image processing apparatus 1 is mounted on a digital camera, high speed processing of embedding processing of digital watermark information is required. From this point of view, the digital watermark image is simply embedded without embedding a plurality of identical pixels. Let's process wavelet transform and embedding of digital watermark information in parallel!

Claims

The scope of the claims

[1] In an information processing apparatus for detecting falsification of target information in which wavelet transformation is performed and electronic penetration information is embedded,

Transformation means for wavelet transforming the object information;

Embedding coefficient detection means for detecting a plurality of embedding coefficients in which a plurality of identical pixels constituting the electron penetration information are respectively embedded;

And counting means for weighting and counting the plurality of embedding coefficients detected by the embedding coefficient detection means according to a sub-band to which each of the embedding coefficients belongs.

An information processing apparatus comprising:

[2] A determination means for determining the embedding coefficient value that has won the majority as the pixel value forming the information of the electronic penetration based on the counting result by the counting means

The information processing apparatus according to claim 1, further comprising:

[3] Based on the counting result by the counting means, the embedding coefficient value which has lost the majority is regarded as a breaking factor, and in the target information wavelet-transformed by the wavelet transformation means, for each coefficient group related across subbands. Calculating means for calculating the number of embedding coefficients and the number of breaking coefficients;

Weighting means for weighting the number of embedding coefficients and the number of breaking coefficients calculated by the calculating means according to the embedding coefficient and the level of the sub-band to which the breaking coefficient belongs;

A determination unit that determines whether the coefficient group has been tampered with based on a ratio between the number of broken coefficients weighted by the weighting unit and the embedding coefficient weighted by the weighting unit;

Operation means for operating the coefficient value of the coefficient group based on the judgment result by the judgment means;

The information processing apparatus according to claim 1, further comprising:

[4] The operation means is

When the ratio is equal to or less than a predetermined threshold value, the breaking coefficient included in the coefficient group is Erase, and if the ratio is larger than a predetermined threshold, change the embedding coefficient included in the coefficient group to a destruction coefficient

The information processing apparatus according to claim 3.

[5] A plurality of identical pixels constituting the digital watermark information is a second sub-band associated with an embedding coefficient randomly determined in advance from the first sub-band coefficient in wavelet-transformed target information. Modulated based on the relevant coefficient and embedded in the embedded coefficient

The information processing apparatus according to any one of claims 1 to 4.

[6] The first subband is a subband other than the lowest subband,

The second subband is the lowest subband

The information processing apparatus according to claim 5.

[7] While performing multi-stage wavelet transform as needed,

6. The information processing method according to claim 5, wherein each time wavelet transform is performed, a related coefficient of a second subband associated with an embedding coefficient determined in random from a coefficient of the first subband is detected. apparatus.

[8] The first sub-band is a sub-band other than the highest sub-band,

The second sub-band is the highest band sub-band

The information processing apparatus according to claim 7.

[9] The apparatus further comprises demodulation means for performing demodulation based on the related coefficient,

9. The apparatus according to any one of claims 5 to 8, wherein said counting means counts and counts the embedding coefficients demodulated by said demodulation means according to the sub-band to which each of the embedding coefficients belongs. The information processing apparatus according to claim 1.

[10] An information processing method for detecting falsification of target information in which wavelet transformation is performed and electronic penetration information is embedded.

Transformation step of wavelet transforming the object information;

An embedding coefficient detection step of detecting a plurality of embedding coefficients in which a plurality of identical pixels constituting the electron penetration information are respectively embedded;

The plurality of embedding coefficients detected in the processing of the embedding coefficient detection step A counting step of weighting and counting according to a sub-band to which each of the embedding coefficients belongs

Information processing method including:

In a program that causes a computer to execute processing for detecting falsification of target information in which wavelet transformation is performed and electronic penetration information is embedded.

Transformation step of wavelet transforming the object information;

A counting step of weighting and counting the plurality of embedding coefficients detected in the process of the embedding coefficient detection step according to a sub-band to which each of the embedding coefficients belongs;

A program that causes a computer to execute a process that includes