CN103886544A - Volume data robustness multi-watermark algorithm based on three-dimensional DCT perceptual Hash and chaos - Google Patents

Abstract

The invention discloses a volume data robustness multi-watermark algorithm based on three-dimensional DCT perceptual Hash and chaos and belongs to the field of multimedia signal processing. The volume data robustness multi-watermark algorithm comprises the steps of firstly utilizing a LogisticMap to generate a chaos sequence and utilizing the chaos sequence to scramble watermarks; then performing watermark embedding, perform global 3D-DCT transformation on volume data, selecting previous 4*4*2 coefficients to perform 3D-IDCT transformation, then extracting perceptual Hash, enabling multiple watermark sequences and the perceptual Hash to be associated to obtain a string of two-value secret key sequences, and storing the two-value secret key sequences in third party; calculating perceptual Hash values of volume data to be tested by using three-dimensional DCT and enabling the perceptual Hash values and the two-value secret key sequences stored in the third party to be associated to extract multiple watermarks, and finally utilizing the LogisticMap to restore the extracted watermarks.

Description

The many watermarking algorithms of volume data robust based on three-dimensional DCT perception Hash and chaos

Technical field

The present invention relates to a kind of volume data multiple digital digital watermark based on three-dimensional DCT perception Hash and chaos, is a kind of multi-media data protection method, belongs to field of multimedia signal processing.

Technical background

Utilize internet can realize distance medical diagnosis, when medical volume data carries out remote transmission on network, be recorded in the personal information of the patient on medical picture, be easy to be revealed.How to protect patient individual privacy, the data such as personal information, patient's electronic health record in patient's the medical imaging such as CT, MRI are not revealed, this problem is along with the universal of internet becomes day by day serious.Encryption method now and access control have been difficult to meet the requirement of medical volume data information security, if be embedded in medical picture using personal information as digital watermarking, be that medical image digital watermark technology (MedicalImageWatermarking is called for short MIW) can address this problem effectively.

Digital watermark technology is the copyright protection for the Digital Media on internet at first; utilize now the feature such as invisibility, robustness of digital watermarking; patient's personal information can be hidden in its medical image, to ensure its safe transmission on the internet.The appearance of medical image digital watermarking, when distance medical diagnosis, the required relevant patient data of remote operation are transmitted on the internet, can effectively protect patient's privacy, avoids patient's data to be tampered.

The research of the current volume data multiple digital watermarking algorithm for resist geometric attacks is less.And volume data exists in a large number in medical image, as: the volume data that CT, MRI image are all made up of section, therefore how research embeds multiple digital watermark in volume data greater significance, and for medical volume data, is generally not allow to revise its content.This has improved difficulty for embed many watermarks in volume data again.

In a word, the research that embeds the many watermarks of robust in three-dimensional data is less, still belongs at present blank for the research of the volume data multi-watermark technology based on perception Hash, has no open report.

Summary of the invention

The object of the invention is to propose a kind of based on three-dimensional DCT perception Hash and Chaotic Scrambling; many watermarking inset and distills of volume data method that realization can be resisted geometric attack and can be resisted again conventional attack; it has higher robustness; and the embedding of many watermarks does not affect the voxel data value of initial body data; be a kind of zero watermarking project, thereby protected preferably three-dimensional data.

Principle of the present invention is: first medical volume data is carried out to overall 3D-DCT conversion, choose front 4 × 4 × 2 coefficients, carry out again 3D-IDCT conversion, then in the coefficient after inverse transformation, extract the perception cryptographic hash of a robust, and multi-watermarking is associated with this perception cryptographic hash, utilize anti-geometry and the conventional attack of the robustness multiple digital watermark of perception Hash.

To achieve these goals, the present invention is performed such: application LogisticMap produces chaos sequence many watermarks are carried out to Chaotic Scrambling and reduction, improves the security of watermark; Adopt three-dimensional dct transform and inverse transformation, obtain the cryptographic hash of volume data.That is: based on the three-dimensional dct transform of the overall situation, 4 × 4 × 2 coefficients before choosing in three-dimensional dct transform coefficient, again the coefficient of choosing is carried out to three-dimensional anti-dct transform, ask for the mean value of coefficient after inverse transformation, then the coefficient after each inverse transformation and mean value are compared, carry out two-value quantification treatment, be more than or equal to mean value, be designated as 1; Be less than mean value, be designated as 0, obtain the perception cryptographic hash of volume data.This cryptographic hash has the ability of resist geometric attacks, and perception Hash, cryptographic Hash are combined, and has realized the robust multiple digital digital watermark of volume data.The method applied in the present invention comprises embedding, extraction and reduction four major parts of the Chaotic Scrambling of many watermarks, many watermarks, and the Chaotic Scrambling that Part I is watermark, comprising: (1) produces chaos sequence X (j) by LogisticMap; (2) according to X (j), many watermarks are carried out to scramble, obtain mixing the watermark BW of scramble ^g(i, j); Part II is the embedding of watermark, comprising: (3), by three-dimensional DCT perception hash algorithm, obtain the robust perception Hash H (j) of initial body data, and (4) utilize cryptography HASH Functional Quality, embed the watermark BW after multiple Chaotic Scrambling ^g(i, j), generates the two-value key sequence Key containing many watermark informations ^g(i, j), then by two-valued function sequence Key ^gthere is third party in (i, j); Part III is the extraction of watermark, comprising: (j), (6) utilize and are present in third-party two-valued function key sequence Key the perception cryptographic hash H ' that (5) obtain volume data to be measured ^gthe perception cryptographic hash H ' of (i, j) and volume data to be measured (j), extracts many watermarks BW ^g' (i, j); Part IV is the reduction of watermark, comprising: (7) application LogisticMap, obtain identical chaos sequence X (j), and reduce to watermark by X (j) (8).

Now be elaborated as follows to method of the present invention:

First select one group of significant bianry image as many watermarks that will embed medical volume data, be designated as W ^g={ w ^g(i, j) | w ^g(i, j)=0,1; 1≤i≤M1,1≤j≤N1}; Meanwhile, choose the MRI volume data carrying in Matlab as primitive medicine volume data, be expressed as: F={f (i, j, k) | f (i, j, k) ∈ R; 1≤i≤M, 1≤j≤N, 1≤k≤P}.Wherein, f (i, j, k) represents voxel (Voxel) data value of primitive medicine volume data, and the grey scale pixel value in this similar two dimensional image, for the purpose of facilitating, is established M1=M2, M=N.

Part I: to the Chaotic Scrambling of watermark

1) generate chaos sequence by LogisticMap;

By initial value x ₀generate chaos sequence X (j) by LogisticMap chaos system.

2) obtain many watermarks of Chaotic Scrambling;

First, original many watermarks are converted into many watermarks of two-value W ^g(i, j), then, value in chaos sequence X (j) is sorted according to order from small to large, finally, according to the change in location before and after each value sequence in X (j), the locational space of many watermarks pixel is carried out to scramble, obtain many watermarks BW of Chaotic Scrambling ^g(i, j).

Part II: the embedding of watermark

3) by volume data being carried out to three-dimensional dct transform and inverse transformation IDCT, obtain a robust perception cryptographic hash H (j) of initial body data;

First initial body data F (i, j, k) is carried out to overall three-dimensional dct transform, obtain three-dimensional DCT matrix of coefficients FD (i, j, k), in matrix of coefficients FD (i, j, k), choose front 4 × 4 × 2 coefficient FD ₄(i, j, k), then to the matrix of coefficients FD selecting ₄(i, j, k) carry out three-dimensional anti-dct transform, obtain the matrix of coefficients FID (i after inverse transformation, j, k), ask for the mean value of coefficient after inverse transformation, by sending out subtraction operation and the two-value quantification treatment of conversion coefficient and its mean value, perception cryptographic hash H (j) main process that obtains volume data is described below:

FD ₄(i,j,k)=DCT3(F(i,j,k))

FID(i,j,k)=IDCT3(FD ₄(i,j,k))

H(j)=BINARY(FID(i,j,k))

4) utilize HASH function, embed multi-watermarking; Generate the two-value key order Key containing many watermark informations ^g(i, j);

Key ^g(i,j)=H(j)⊕BW ^g(i,j)

Key ^g(i, j) is perception cryptographic hash H (j) and the many watermarking images BW by volume data ^g(i, j), the Hash function conventional by cryptography generates.Preserve Key ^g(i, j), will use while extracting many watermarks below.By by Key ^g(i, j) applies for to third party as key, to obtain entitlement and the right to use of medical volume data, reaches the object of copyright protection.And the embedding of many watermarks does not affect the quality of primitive medicine volume data, it is a kind of zero watermarking project.

Part III: the extraction of watermark

5) the perception cryptographic hash H ' that obtains testing data (j);

If volume data to be measured is F ' (i, j, k), after the three-dimensional dct transform of the overall situation of volume data, obtaining three-dimensional DCT matrix of coefficients is FD ' (i, j, k), then choose suitable coefficient and carry out inverse transformation, again by above-mentioned steps 1) similar method, try to achieve the perception cryptographic hash H ' of volume data to be measured (j);

FD’ ₄(i,j,k)=DCT3(F’(i,j,k))

FID’(i,j,k)=IDCT3(FD’ ₄(i,j,k))

H’(j)=BINARY(FID’(i,j,k))

6) in volume data to be measured, extract many watermarks BW ^g' (i, j);

BW ^g’(i,j)=Key ^g(i,j)⊕H’(j)

According to the logic key sequence Key generating in the time embedding many watermarks ^gthe perception cryptographic hash H ' of (i, j) and volume data to be measured (j), the many watermarks BW that utilizes Hash Functional Quality to extract to contain in volume data to be measured ^g' (i, j).

Part IV: the reduction of many watermarks

7) generate chaos sequence by LogisticMap;

By with initial value x that above step1 is identical ₀generate identical chaos sequence X (j) by LogisticMap chaos system;

8) watermark that reduction is extracted;

First by the value in chaos sequence X (j) according to sorting from small to large, then according to the change in location before and after each value sequence in X (j), the locational space of multi-watermarking pixel reduce and obtains many watermarks W of reducing ^g' (i, j).

Again according to W ^g(i, j) and W ^g' whether the degree of correlation of (i, j) is differentiated have watermark to embed, thereby confirm the entitlement of volume data to be measured and the safety issue of sufferer information.

The present invention and existing medical science multi-watermark technology relatively have following advantage:

First, because the present invention is the multiple digital digital watermark based on three-dimensional DCT perception hash algorithm, confirm by experimental data below, these many watermarks not only have stronger anti-conventional attack ability, and have stronger resist geometric attacks ability; Secondly, many watermarks of embedding, through LogisticMap Chaotic Scrambling, make many watermark informations become disorderly and unsystematic, have improved the security of watermark information; Finally, the embedding of many watermarks does not affect the voxel data value of initial body data, is a kind of zero digital watermark, better must protect medical volume data.This characteristic, especially has very high practical value at aspects such as medical image processing, and usable range is wide, and can realize embedding and the extraction of large watermark.

We illustrate from theoretical foundation and test figure below:

1) Three-dimensional DCT (3D-DCT)

Three-dimensional dct transform formula is as follows:

Corresponding size is M × N × P volume data, and 3 d-dem cosine direct transform (DCT) formula is as follows:

$F(u,v,w)=c\left(u\right)c\left(v\right)c\left(w\right)[\underset{x=0}{\overset{M-1}{\mathrm{\Σ}}}\underset{y=0}{\overset{N-1}{\mathrm{\Σ}}}\underset{p=0}{\overset{P-1}{\mathrm{\Σ}}}f(x,y,z)\·\cos \frac{(2x+1)\mathrm{u\π}}{2M}\cos \frac{(2y+1)\mathrm{v\π}}{2N}\cos \frac{(2z+1)\mathrm{w\π}}{2P}]$

u＝0,1,...,M-1;v＝0,1,...,N-1;w＝0,1,...,P-1;

In formula

$c\left(u\right)=\left\{\begin{array}{cc}\sqrt{1/M}& u=0\\ \sqrt{2/M}& u=\mathrm{1,2},...,M-1\end{array}\right.$

$c\left(v\right)=\left\{\begin{array}{cc}\sqrt{1/N}& v=0\\ \sqrt{2/N}& v=\mathrm{1,2},...,N-1\end{array}\right.$

$c\left(w\right)=\left\{\begin{array}{cc}\sqrt{1/P}& w=0\\ \sqrt{2/P}& w=\mathrm{1,2},...,P-1\end{array}\right.$

Here, f (x, y, z) is voxel (voxel) data value that volume data V locates at (x, y, z), and F (u, v, w) is the 3D-DCT conversion coefficient that this voxel data is corresponding.

3 d-dem cosine inverse transformation (IDCT) formula is as follows:

$f(x,y,z)=\underset{u=0}{\overset{M-1}{\mathrm{\Σ}}}\underset{v=0}{\overset{N-1}{\mathrm{\Σ}}}\underset{w=0}{\overset{P-1}{\mathrm{\Σ}}}[c\left(u\right)c\left(v\right)c\left(w\right)F(u,v,w)\cos \frac{(2x+1)\mathrm{u\π}}{2M}\cos \frac{(2y+1)\mathrm{v\π}}{2N}\cos \frac{(2z+1)\mathrm{w\π}}{2P}]$

x＝0,1,...,M-1;y＝0,1,...,N-1;z＝0,1,...,P-1

Wherein, (x, y, z) is spatial domain sampled value; (u, v, w) is frequency field sampled value.The acquisition of medical volume data can be by CT and MRI(MagneticResnaneIamge, magnetic resonance imaging), volume data (Volumedata) is made up of the section (slice) of many layers, and each section is a two dimensional image, size is M × N, and the number of plies of section is P.

2）LogisticMap

Chaos is one random motion seemingly, refers to the similar random process occurring in deterministic system.Therefore, had its initial value and parameter, we just can generate this chaos system.LogisticMap is foremost a kind of chaos system, and it is the Nonlinear Mapping being given by the following formula:

$x_{k+1}={\mathrm{\μx}}_k(1+x_k)$

Wherein, 0≤μ≤4 are growth parameter, x _k∈ (0,1) is system variable, and k is iterations.The research work of Chaos dynamic system points out, in the time of growth parameter 3.569945≤μ≤4, LogisticMap works in chaos state.Can see that initial value has a slight difference will cause the significant difference of chaos sequence.Therefore, above sequence is a desirable key sequence.Set μ=4 herein, chaos sequence is by different initial value x ₀produce.

3) choosing method of volume data perception hash function

Perception cryptographic hash is after quantizing, to be obtained by the proper vector of an individual data items.The main cause of current most of watermarking algorithm resist geometric attacks ability is: people are embedded in digital watermarking in voxel or conversion coefficient, and the slight geometric transformation of volume data usually can cause the larger variation suddenly of voxel data value or transform coefficient values.The watermark being embedded in like this in volume data is just attacked easily.If can find the perception cryptographic hash of an antimer data geometrical feature, in the time there is little geometric transformation in volume data, can there is not obvious sudden change in this cryptographic hash, then we are associated the perception Hash of the digital watermarking that will embed and this volume data, and the digital watermarking embedding so just has good resist geometric attacks ability.

The experimental data that we choose some conventional attacks and geometric attack is shown in Table 1.The former figure that is used as test in table 1 is Fig. 1, it is a section (getting the tenth) of a MRI volume data carrying in matlab, in table 1, " the 1st row " demonstration is volume data type under attack, this sectioning image being subject to after conventional attack is shown in Fig. 2 to Fig. 4, and Fig. 5 to Fig. 8 is shown in the three-dimensional imaging that conventional attack is corresponding; The sectioning image being subject to after geometric attack is shown in Fig. 9 to Figure 12, and Figure 13 to Figure 16 is shown in its corresponding three-dimensional imaging.Y-PSNR (PSNR) after volume data that what " the 2nd row " of table 1 represented is is under attack; " the 3rd row " of table 1 arrive " the 10th row ", are in the three-dimensional data after anti-dct transform, to choose arbitrarily eight pixel values such as " F (1, Isosorbide-5-Nitrae), F (1,2,1) "." the 11st row " of table 1 are that DCT perception hash algorithm two-value quantification treatment is obtained the average pixel value coming.For conventional attack, may there are some conversion in these pixel values F (1, Isosorbide-5-Nitrae), F (1,2,1) etc., but the size of it and average pixel value is more constant, and we will be more than or equal to mean value, be designated as 1; Be less than mean value, be designated as 0, so for initial body data, pixel value F (1,1,4), the cryptographic hash sequence of the correspondence such as F (1,2,1) is: " 00010101 ", specifically be listed as in the 12nd of table 1, observing these row can find, no matter conventional attack or this symbol sebolic addressing of geometric attack are similar with the maintenance of initial body data, all larger with initial body data normalization related coefficient, this meets the robustness feature of perception hash function, for similar image, its perception functional similarity, their related coefficient is larger.In table 1 " the 13rd row ".

Table 1 volume data DCT and perception Hash are processed rear section coefficient and are subject to the changing value after different attack

* DCT and perception Hash are processed rear coefficient unit 1.0e+001

Robustness and not sentience are two key properties of perception hash function.Be a perception hash function of this volume data for what further prove to extract as stated above, we can not collision property detect it, and for different volume datas, its perception cryptographic hash is different; The value of the related coefficient between them is less; We,, different tested objects (seeing Figure 17 to Figure 23), test, and by three-dimensional DCT perception hash algorithm, they are processed.From angle of statistics, front 8 × 8 × 4 256 DCT coefficients are got here.And the perception cryptographic hash of obtaining every individual data items related coefficient each other, result of calculation is as shown in table 2.

As can be seen from Table 2, first, the related coefficient maximum between volume data self, is 1.00; Secondly, the related coefficient between Figure 22 and Figure 23 also more greatly 0.72, and the volume data that these two figure are two similar livers of shape;

Related coefficient (vector length 256bit) between the different volume data perception of table 2 cryptographic hash

?	Ha	Hb	Hc	Hd	He	Hf	Hg
								Ha	1.00	0.58	0.46	-0.09	-0.53	0.11	0.15
Hb	0.58	1.00	0.28	0.12	-0.49	0.19	0.19
								Hc	0.46	0.28	1.00	0.01	-0.30	-0.07	-0.05
Hd	-0.09	0.12	0.01	1.00	-0.07	0.14	0.14
								He	-0.53	-0.49	-0.30	-0.07	1.00	-0.23	-0.21
Hf	0.11	0.19	-0.07	0.14	-0.23	1.00	0.74
								Hg	0.15	0.19	-0.05	0.14	-0.21	0.74	1.00

Figure 17 and Figure 18, related coefficient is 0.61, also larger, be the third-largest related coefficient, and these two heads that figure is human body is also more similar in table.Facies relationship numerical value between other perception cryptographic hash is less, this with our eye-observation to be consistent, the perception cryptographic hash that this explanation is extracted by the method for this invention, has reflected the main resemblance of volume data.

In sum, we,, by the analysis to real part coefficient after the anti-dct transform of the three-dimensional of volume data, utilize the positive inverse transformation of three-dimensional DCT, obtain a perception cryptographic hash of volume data.

4) length of perception cryptographic hash and the relation of robustness

According to human visual system (HVS), Low Medium Frequency signal is larger to people's visual impact, is image outline for two dimensional image, is exactly the appearance profile of volume data for 3-D view.Therefore, we choose the Low Medium Frequency coefficient of volume data in the time that volume data is chosen to proper transformation coefficient, the size of the initial body data of overall three-dimensional dct transform is selected and carried out to the number of Low Medium Frequency coefficient, and the quantity of information of disposable embedding is relevant with the robustness of requirement, the length L of the perception cryptographic hash of choosing is less, the quantity of information of disposable embedding is fewer, but robustness is higher.Consider in experiment below, the length that we choose L in the time of specific experiment chamber is 32.

Brief description of the drawings

Fig. 1 is a section (acquiescence is the 10th section of volume data) of initial body data.

Fig. 2 is the sectioning image after 10% Gauusian noise jammer.

Fig. 3 is the sectioning image after JPEG compression (compression quality is 2%).

Fig. 4 is the sectioning image (filtering parameter is [5x5]) after medium filtering.

Fig. 5 is three-dimensional imagings corresponding to initial body data.

Fig. 6 is that to be subject to intensity be the three-dimensional imaging that 10% Gauss disturbs rear correspondence to volume data.

Fig. 7 is corresponding three-dimensional imaging after JPEG compression (compression quality is 2%).

Fig. 8 is three-dimensional imaging (filtering parameter is [5x5]) corresponding after medium filtering.

Fig. 9 is the sectioning image through up time rotation 20 degree.

Figure 10 is the sectioning image through 0.5 times of convergent-divergent.

Figure 11 is that vertical direction moves down 10% sectioning image.

Figure 12 is that Z-direction is sheared first sectioning image after 10%.

Figure 13 is the three-dimensional imaging of up time rotation 20 degree.

Figure 14 is that zoom factor is 0.5 three-dimensional imaging.

Figure 15 is that vertical direction moves down 10% three-dimensional imaging.

Figure 16 is that Z-direction is sheared 10% three-dimensional imaging.

Figure 17 is the three-dimensional imaging of volume data MRI_1.

Figure 18 is the three-dimensional imaging of volume data MRI_2.

Figure 19 is the three-dimensional imaging of volume data MRI_3.

Figure 20 is the three-dimensional imaging of volume data Teddybear.

Figure 21 is the three-dimensional imaging of volume data Tooth.

Figure 22 is the three-dimensional imaging of volume data Liver_1.

Figure 23 is the three-dimensional imaging of volume data Liver_2.

Figure 24 is original watermark HN.

Figure 25 is original watermark CN.

Figure 26 is the watermark HN after LogisticMap Chaotic Scrambling.

Figure 27 is the watermark CN after LogisticMap Chaotic Scrambling.

Figure 28 is the many watermark sections that do not add while interference.

Figure 29 is the volume data three-dimensional reconstruction figure not adding while interference.

Figure 30 does not add the watermark HN extracting while interference.

Figure 31 does not add the watermark CN that interference is extraction.

Figure 32 is the sectioning image (Gaussian noise intensity 10%) after Gauusian noise jammer.

Figure 33 is the three-dimensional reconstruction figure (Gaussian noise intensity 10%) after Gauusian noise jammer.Figure 34 is the watermark HN(Gaussian noise intensity 10% of extracting after Gauusian noise jammer).Figure 35 is the watermark CN(Gaussian noise intensity 10% of extracting after Gauusian noise jammer).Figure 36 is the sectioning image (compression quality parameter is 5%) after JPEG compression.

Figure 37 is the volume data three-dimensional imaging (compression quality parameter is 5%) after JPEG compression.Figure 38 is that the watermark HN(compression quality parameter of extracting after JPEG compression is 5%).

Figure 39 is that the watermark CN(compression quality parameter of extracting after JPEG compression is 5%).

Figure 40 is the sectioning image (filtering parameter is [5x5], and filter times is 1 time) after medium filtering.

Figure 41 is the three-dimensional imaging (filtering parameter is [5x5], and filter times is 1 time) of the volume data after medium filtering.

Figure 42 is the watermark HN (filtering parameter is [5x5], and filter times is 1 time) extracting after medium filtering.

Figure 43 is the watermark CN (filtering parameter is [5x5], and filter times is 1 time) extracting after medium filtering.

Figure 44 is the sectioning image after up time rotation 20 degree.

Figure 45 is the three-dimensional imaging of volume data after up time rotation 20 degree.

Figure 46 is the watermark HN extracting after up time rotation 20 degree.

Figure 47 is the watermark CN extracting after up time rotation 20 degree.

Figure 48 is that zoom factor is 0.5 sectioning image.

Figure 49 is that zoom factor is 0.5 three-dimensional imaging.

Figure 50 is that zoom factor is the watermark HN extracting for 0.5 o'clock.

Figure 51 is that zoom factor is the watermark CN extracting for 0.5 o'clock.

Figure 52 vertically moves down 5% sectioning image.

Figure 53 vertically moves down 5% three-dimensional imaging corresponding to volume data.

Figure 54 is the watermark HN that vertically moves down 5% rear extraction.

Figure 55 is the watermark CN that vertically moves down 5% rear extraction.

Figure 56 is after Z-direction shears 10%, first sectioning image of volume data.

Figure 57 is after Z-direction shears 10%, the three-dimensional imaging of volume data.

Figure 58 is after Z-direction shears 10%, the watermark HN of extraction.

Figure 59 is after Z-direction shears 10%, the watermark CN of extraction.

Figure 60 is that distortion frequency factor is the sectioning image of 13 o'clock.

Figure 61 is that distortion frequency factor is the three-dimensional imaging of the volume data of 13 o'clock.

Figure 62 is that distortion frequency factor is the watermark HN extracting for 13 o'clock.

Figure 63 is that distortion frequency factor is the watermark CN extracting for 13 o'clock.

Embodiment

Below in conjunction with accompanying drawing, the invention will be further described, selects one group of significant bianry image as original many watermarks, is designated as W ^g={ w ^g(i, j) | w ^g(i, j)=0,1; 1≤i≤M1,1≤j≤N1}, is shown in Figure 24 and Figure 25, the size of many watermarks is here all 32 × 32.By the many watermarks after Logistic Map Chaotic Scrambling, see Figure 26 and Figure 27, can obviously see that watermark has a very large change, security improves.Emulation platform is Matlab2010a, and Fig. 1 is shown in a section of primitive medicine volume data, is to take from the magnetic resonance three-dimensional image volumetric data (MRI.mat) carrying in matlab, and the size of volume data is 128x128x27, sees Fig. 5.Initial body data are expressed as F (i, j, k), wherein 1≤i, j≤128; 1≤k≤27, corresponding 3D-DCT matrix of coefficients is FD (i, j, k), wherein 1≤i, j≤128; 1≤k≤27.While considering the capacity of robustness and disposable embed watermark, we get front 4 × 4 × 2 coefficients.Carry out 3D-IDCT conversion again, and ask for the mean value of coefficient after inverse transformation, then by mean value, the coefficient after to inverse transformation carries out two-value quantification treatment, obtains the cryptographic hash sequence H (j) of volume data.Detect after W ' (i, j) by watermarking algorithm, we have judged whether that by calculating normalized correlation coefficient NC (NormalizedCross Correlation) watermark embeds.

Do not add while interference, Figure 28 is the sectioning image (acquiescence is selected the tenth section here, and test is made up of 27 sections altogether by volume data) not adding while interference;

Figure 29 is the volume data three-dimensional imaging not adding while interference;

Figure 30 does not add the watermark HN extracting while interference, can see that NC1=1.00, Figure 31 do not add the watermark CN that interference is extraction, and NC2=1.00, can accurately must extract many watermarks.

We judge anti-conventional attack ability and the resist geometric attacks ability of this multiple digital water mark method by specific experiment below.

First test the ability of the anti-conventional attack of these many watermarking algorithms.

(1) add Gaussian noise

Use imnoise () function to add Gaussian noise in many watermarking images.

Table 3 is experimental datas of the anti-Gauusian noise jammer of many watermarks.Therefrom can see, when Gaussian noise intensity is up to 25% time, the PSNR of many watermarks volume data is down to 0.10dB, many watermarks of at this moment extracting, and related coefficient NC1=1.00, NC2=1.00, still can accurately must extract many watermarks.The anti-Gaussian noise ability that this explanation adopts this invention to have.

Sectioning image when Figure 32 is Gaussian noise intensity 10% is visually very fuzzy;

Figure 33 is corresponding volume data three-dimensional imaging, visually very fuzzy, and the PSNR=3.30dB of volume data is lower;

Figure 34 and Figure 35 are respectively watermark HN and the watermark CN extracting, and can accurately must extract many watermarks, NC1=1.00, NC2=1.00.

The anti-Gauusian noise jammer data of watermark more than table 3

Noise intensity (%)	1	3	5	10	15	20	25
								PSNR(dB)	12.52	8.02	6.03	3.32	1.80	0.82	0.10
NC1	1.00	1.00	1.00	1.00	1.00	1.00	1.00
								NC2	1.00	1.00	1.00	1.00	1.00	1.00	1.00

(2) JPEG compression is processed

Adopt image compression quality percentage, as parameter, many watermarks volume data is carried out to JPEG compression; Table 4 is the anti-JPEG compression experiment of many watermarks volume data data.When compression quality is only 2%, at this moment compression quality is lower, still can extract many watermarks, NC1=1.00, NC2=1.00.

Figure 36 is that compression quality is 5% sectioning image, and blocking artifact has appearred in this figure;

Figure 37 is corresponding volume data three-dimensional imaging, and three-dimensional blocking artifact has appearred in this figure;

Figure 38 and Figure 39 are respectively watermark HN and the watermark CN extracting, NC1=1.00,

NC2=1.00, can accurately extract many watermarks.

The anti-JPEG compression experiment of watermark more than table 4 data

Compression quality (%)	2	4	8	10	20	40	60	80
									PSNR(dB)	16.57	17.82	20.21	21.20	23.10	25.06	26.61	29.31
NC1	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
									NC2	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00

(3) medium filtering processing

Table 5 is the anti-medium filtering ability of many watermarks volume data, and it can be seen from the table, when medium filtering parameter is [7x7], filtering multiplicity is 20 o'clock, still can record the existence of many watermarks, NC1=1.00, NC2=1.00.

Figure 40 is that medium filtering parameter is [5x5], the sectioning image that filtering multiplicity is 1, and image has occurred fuzzy;

Figure 41 is corresponding volume data three-dimensional imaging, and at this moment the profile such as ear is not too clearly demarcated;

Figure 42 and 43 is respectively watermark HN and the watermark CN extracting, and NC1=l.00, NC2=1.00, can accurately extract many watermarks.

The anti-medium filtering experimental data of watermark more than table 5

Many watermarks resist geometric attacks ability

(1) rotational transform

Table 6 is the anti-rotation attack experimental datas of many watermarks.From table, can see that NC1=0.71, NC2=0.73, still can extract many watermarks in the time that many watermarks volume data up time is rotated 35 °.

Figure 44 is many watermarks sectioning image of up time rotation 20 degree;

Figure 45 is corresponding volume data three-dimensional imaging, and at this moment, the signal to noise ratio (S/N ratio) of many watermarks volume data is lower, PSNR=12.44dB;

Figure 46 and Figure 47 are respectively watermark HN and the watermark CN extracting, and NC1=0.95, NC2=0.95, can extract many watermarks exactly.

The anti-rotation attack experimental data of watermark more than table 6

Up time rotation	5 degree	10 degree	15 degree	20 degree	25 degree	30 degree	35 degree
								PSNR(dB)	16.54	13.97	12.98	12.44	12.04	11.68	11.33
NC1	1.00	1.00	1.00	0.95	0.71	0.71	0.71
								NC2	1.00	1.00	1.00	0.95	0.73	0.73	0.73

(2) scale transformation

Table 7 is the nonshrink attack experimental data of putting of many watermarks volume data, and when many watermarks volume data zoom factor is little to 0.2 time, related coefficient NC1=1.00, NC2=1.00, can accurately extract many watermarks as can be seen from Table 7.

Figure 48 is the many watermarks sectioning image (zoom factor is 0.5) after convergent-divergent;

Figure 49 is after convergent-divergent is attacked, the three-dimensional imaging (zoom factor is 0.5) that volume data is corresponding;

Figure 50 and 51 is respectively after convergent-divergent is attacked, the watermark HN of extraction and watermark CN, and NC1=1.00, NC2=1.00, can accurately must extract many watermarks.

The nonshrink attack experimental data of putting of watermark more than table 7

Zoom factor	0.2	0.5	0.8	1.2	2.0	4.0
							NC1	1.00	1.00	1.00	1.00	1.00	1.00
NC2	1.00	1.00	1.00	1.00	1.00	1.00

(3) translation transformation

Table 8 is the anti-translation transformation experimental datas of many watermarks.When learning when level from table or vertically moving 10%, the value of NC1, NC2, all higher than 0.5, can accurately be extracted many watermarks, therefore these many water mark methods have stronger anti-translation transformation ability.

Figure 52 cuts into slices vertically to move down 5% image;

Figure 53 is after each section of volume data vertically moves down 5%, corresponding three-dimensional imaging, and at this moment PSNR=11.97dB, signal to noise ratio (S/N ratio) is lower;

Figure 54 and Figure 55 are watermark HN and the watermark CN extracting, and can accurately extract many watermarks, NC1=1.00, NC2=1.00.

The anti-translation transformation experimental data of watermark more than table 8

(4) shearing attack

Table 9 is the anti-shearing attack experimental datas of many watermarks, from table, can see, when shearing from Z-direction, when shearing displacement is 40%, still can extract many watermarks, and NC1=0.88, NC2=0.87, illustrate that these many watermarking algorithms have stronger anti-shearing attacking ability.

Figure 56 is after shearing 10% by Z-direction, first sectioning image;

Figure 57 is the three-dimensional imaging of shearing 10% rear correspondence by Z-direction, can find the successful of shearing attack; The three-dimensional imaging of the relatively former figure in top, has cut one.

Figure 58 and Figure 59 are respectively watermark HN and the watermark CN extracting, and can accurately must extract many watermarks, NC1=1.00, NC2=1.00.

The anti-shearing attack experimental data of watermark more than table 9

Z axis is sheared (%)	2	4	6	8	10	20	40
								NC1	1.00	1.00	1.00	1.00	1.00	1.00	1.00
NC2	1.00	1.00	1.00	1.00	1.00	1.00	1.00

(5) distortion is attacked

Table 10 is the anti-twist attack experimental datas of many watermarks, and distortion parameter is the distortion factor, and the distortion factor is larger, the frequency that represents distortion is higher, in the time that the distortion factor is 24, and the at this moment lower PSNR=9.68dB of the signal to noise ratio (S/N ratio) of volume data, but at this moment NC1=0.79, NC2=0.79, still can extract many watermarks; In the time that the distortion factor is 3, the at this moment lower PSNR=10.13dB of the signal to noise ratio (S/N ratio) of volume data, but at this moment NC1=0.73, NC2=0.73, now NC value is relatively low, but still can extract many watermarks.And from table 10, find, in the time that the distortion factor is lower, larger on the low frequency characteristic impact of volume data, so NC value is less; And in the time that the distortion factor is larger, larger on the high frequency characteristics impact of volume data, less on the exterior contour impact of volume data, so NC value is larger; Data in table are consistent in the analysis of the medium and low frequency coefficient to volume data above with us.

Figure 60 is the sectioning image (the distortion factor is 13) after distortion is attacked;

Figure 61 is corresponding volume data three-dimensional imaging after distortion is attacked, PSNR=9.83dB, and signal to noise ratio (S/N ratio) is lower;

Figure 62 and Figure 63 are respectively watermark HN and the watermark CN extracting, and NC1=0.91, NC2=0.90, can extract many watermarks comparatively exactly.

The anti-twist attack experimental data of watermark more than table 10

Distortion frequency factor	3	5	7	9	13	20	24
								PSNR(dB)	10.13	10.16	9.89	9.58	9.83	9.68	9.68
NC1	0.73	0.88	0.88	0.73	0.88	0.79	0.79
								NC2	0.73	0.87	0.87	0.73	0.87	0.79	0.79

Claims (1)

1. the many water mark methods of volume data robust based on three-dimensional DCT perception Hash and chaos, it is characterized in that: convert based on 3D-DCT, choose front 4 × 4 × 4 coefficients, carry out again 3D-IDCT conversion, then in inverse transformation coefficient, extract the perception hash function of robust, and many watermark sequences are associated with the perception cryptographic hash of this robust, realized resist geometric attacks and the conventional attack of medical volume data multiple digital watermark, this volume data multiple digital watermark implementing method is divided into four major parts, amounts to eight steps:

Part I is the Chaotic Scrambling of many watermarks: utilize LogisticMap character to produce chaos sequence many watermarks are carried out to scramble, obtain many watermarks BW of Chaotic Scrambling ^g(i, j);

1) by logic initial value x ₀generate chaos sequence X (j) by LogisticMap;

2) value in chaos sequence X (j) is arranged according to order from small to large, then according to the change in location before and after each value sequence in X (j), scramble is carried out in the locus of watermark pixel, obtain many watermarks BW of Chaotic Scrambling ^g(i, j);

Part II is the embedding of many watermarks: by the embedding operation to many watermarks, obtain corresponding two-valued function sequence Key ^g(i, j);

3) by volume data being carried out to three-dimensional dct transform and inverse transformation IDCT, obtain a robust perception cryptographic hash H (j) of initial body data;

First initial body data F (i, j, k) is carried out to overall three-dimensional dct transform, obtain three-dimensional DCT matrix of coefficients FD (i, j, k), in matrix of coefficients FD (i, j, k), choose front 4 × 4 × 2 coefficient FD ₄(i, j, k), then to the matrix of coefficients FD selecting ₄(i, j, k) carry out three-dimensional anti-dct transform, obtain the matrix of coefficients FID (i after inverse transformation, j, k), ask for the mean value of coefficient after inverse transformation, by sending out subtraction operation and the two-value quantification treatment of conversion coefficient and its mean value, perception cryptographic hash H (j) main process that obtains volume data is described below:

FD ₄(i,j,k)=DCT3(F(i,j,k))

FID(i,j,k)=IDCT3(FD ₄(i,j,k))

H(j)=BINARY(FID(i,j,k))

4) utilize many watermark sequences BW of cryptography Hash Functional Quality and Chaotic Scrambling ^g(i, j), generates the two-value key sequence Key containing many watermark informations ^g(i, j);

Key ^g(i,j)=H(j)⊕BW ^g(i,j)；

Preserve Key ^g(i, j), will use while extracting many watermarks below, by Key ^g(i, j) applies for to third party as key, to obtain the entitlement to primitive medicine volume data;

Part III is the extraction of many watermarks: by two-valued function sequence Key ^gthe robust perception cryptographic hash H ' of (i, j) and volume data to be measured (j), extracts many watermarks BW ^g' (i, j);

5), for volume data to be measured, according to the method for step 3), obtain the perception cryptographic hash H ' of volume data to be measured (j);

6) utilize cryptography Hash Functional Quality and be present in third-party Key ^g(i, j), extracts many watermarks BW ^g' (i, j)=Key ^g(i, j) ⊕ H ' (j);

The reduction of the many watermarks of Part IV;

7) by logic initial value x ₀generate chaos sequence X (j) by LogisticMap;

8) to ascending sequence of value in chaos sequence X (j), according to the change in location before and after each value sequence in X (j), the locus of the watermark pixel of extracting is reduced, obtain many watermarks W of reduction ^g' (i, j);

By W ^g(i, j) and W ^g' (i, j) carry out the calculating of degree of correlation NC, whether utilize NC value to differentiate has watermark to embed.

Images