CN1440197A - Positioned Chaos fragile digital implanting and extracting method - Google Patents

Abstract

This invention relates to a method for embedding and extracting the frail water mark of Chaos of positioning type. Based on the extremke sensitivity of Chaos to the original value, adopting the authentication of the positioning type and the technique of the frail digital water mark, the method projects the value of the gray level of the pixels of original image into a original value of Chaos, generates the image of the water mark through several iterations of Chaos and then embeds the water mark onto the LSB plane of the original image. The security of the algorithm is increased due to the application of the random sequence in Chaos and the integrity of the image of the water mark is examined through distorting the judging matrix when the water mark is extracted.

Description

Location type chaos fragile digital watermark embeds and extracting method

Technical field:

The present invention relates to a kind of location type chaos fragile digital watermark and embed and extracting method, is that a kind of fragile digital watermark that is specially adapted to authentication, content tampering location and the integrity certification of digital product embeds and extracting method.

Background technology:

In recent years, along with computer and Development of Multimedia Technology, the copyright protection and the content authentication of digital product more and more come into one's own, and digital watermarking is exactly the extensive concern that produces and be subjected to academia under such background.The research of robust digital watermark that is used for copyright protection is more deep; but consider the content integrity authentication of digital product; and the confidence level of the multi-medium data of transmission, this digital watermark can't adapt to this requirement fully, and the fragile watermark technology can realize this point.Fragile watermark should satisfy following characteristics usually: (1) invisibility, and the sensitiveness that distort the adding watermarking images (2), (3) tampering location ability, (4) extract watermark does not need original image.

In the middle of the location type fragile digital watermark of research at present, following two kinds of typical methods are arranged.

(1) index modulation method

In recent years, there are some scholars to propose watermark information to be embedded in the middle of the carrier image abroad with the method for index modulation.In these methods, it is that watermark information is embedded in the coefficient in transform domain greater than certain given threshold value by modulation that a kind of method is relatively effectively arranged, such as list of references " MultipurposeWatermarking for Image Authentication and Protection " (C S Lu, H Y Mark Liao.IEEE Trans on Image Processing.2001,10 (10): 1579-1592) image low frequency sub-band coefficient is just being passed through, negatively-modulated embeds the wavelet domain coefficients greater than certain given threshold value, the watermark information that just can obtain embedding by demodulation during extraction.

But there is following problems in this method:

Problem one is not because propose effective countermeasure at tampering location, so though above-mentioned fragile watermark has the tampering location ability, positioning accuracy is not high.

Problem two, above-mentioned algorithm be owing to exist positive and negative two kinds of modulation, and wherein also have selection of threshold and how to make it the problem that reaches optimum, and institute is not suitable for the occasion of handling in real time so that algorithm implements very complicatedly.

(2) based on the Hash function method of safety

The Hash function is widely used on conventional cipher is learned.Utilize the Hash function that the extreme sensitivity of initial value is realized that fragile watermark is to distorting the focus that responsive requirement also is the research of watermark in recent years.List of references " Hierarchical Watermarking for Secure Image Authentication with Localization " (M U Celik, G Sharma, E Saber, et al.IEEE Trans on Image Processing.2002,11 (6): 585-595) at list of references " Public Key Watermark for Image Verification andAuthentication " (P W Wong.IEEE Int Con on Image Processing.1998, on basis 1:455-459), a kind of classification fragile watermark with tampering location ability is proposed, the original image classification, and by Hash function generation summary, pass through afterwards encrypted private key with the watermark XOR, LSB (the Least Significant Bit that is embedded into original image at last, the minimum bit position) plane, this algorithm progressively improves positioning accuracy by continuous classification, and its major advantage is to have overcome the ineffectivity of vector being attacked (VQ attack).

But there are a lot of problems equally in these Fragile Watermarking Algorithm based on the Hash function of safety.

Problem one, above-mentioned algorithm causes embedding content and comprises mutually because different fractions is enjoyed same LSB plane, so the collision problem when needing partitioning algorithm to solve to embed.

Problem two, similar these traditional tampering location methods generally are based on the sensitiveness of Hash function to distorting, the drawback of this method is: (1) necessary piecemeal, cause the location accurate inadequately, (2) Hash function calculation complexity height, be not suitable for real-time processing, (3) approved safe Hash function kind is few.

In sum, for a well behaved fragile watermark, must meet these points basic demand: (1) is invisible.Be that digital multimedia data after watermarked must have very high Y-PSNR value; (2) has the tamper-proofing ability.Do not need original image when (3) detecting.(4) has the tampering location ability.And prior art still fails well to address these problems.

Summary of the invention:

The objective of the invention is at the deficiencies in the prior art, provide a kind of location type chaos fragile digital watermark to embed and extracting method based on chaotic maps, to the distorting sensitivity and have accurate tampering location ability of data, speed advantage is obvious, and visual effect is good simultaneously.

For realizing such purpose, the present invention is based on the extreme sensitivity of chaos to initial value, adopt the authentication and the Fragile Watermark Technique of location type, the pixel gray value of original image is mapped as the chaos initial value and generates watermarking images through the several times chaos iteration, the LSB plane that is embedded into original image then, utilize the extreme sensitivity of chaos to initial value, can accurately locate adding distorting of watermarking images, stable performance, principle is clear, realization speed is fast, and the extraction of watermark does not need original image.

The present invention includes following concrete steps:

1) chaos iteration: with system key K is the initial value of chaotic maps, obtains chaos random sequence { n _l| 1≤n _l≤ 2 ^m, l=1,2 ..., N ², this sequence is as the number of times of following chaos iteration.

2) watermark generates: (i, j), (1≤i, the LSB plane zero setting of j≤N) obtains original image X

By mapping

As the initial value of chaotic maps, and with sequence n _lAs iterations, obtain a real-valued sequence of chaos thus

By quantize function Q obtain watermark information S (i, j).

3) watermark embeds: right

Each point implement aforesaid operations, obtain watermark S.S is embedded into watermark

The LSB plane, obtained watermarked image WX thus.

4) watermark extracting: to the LSB plane zero setting of the watermarked image received, be used as the initial value of chaos, according to the chaos iteration number of times that generates by key, obtain watermark S then by chaos iteration with its pixel gray value _ExCompare S _ExWith the corresponding position on the LSB plane of WX, if S _ExAll the same with corresponding position, the LSB plane of WX, show that original image is not distorted, otherwise point inequality is detected distorting a little.

Fragile Watermark Algorithm based on chaotic maps of the present invention is for the authentication and the fragile digital watermark of location type have been opened up a new path.Watermarked by the LSB plane of revising original image, obtained higher Y-PSNR.The application of chaos random sequence has increased the fail safe of algorithm.Algorithm is simple, realizes easily, does not need original image when extracting watermark, and can accurate localization for adding distorting of watermarking images, and these good performances have increased its range of application greatly.

Description of drawings:

Fig. 1 is a watermarking algorithm block diagram of the present invention.

Fig. 2 is watermark extracting of the present invention and verification process block diagram.

Fig. 3 revises for picture material and detects embodiment.

Wherein, Fig. 3 (a) is the image that is embedded with watermark before distorting, the image of Fig. 3 (b) for distorting, and Fig. 3 (c) is for distorting testing result, and Fig. 3 (d) is the magnification region before distorting, and Fig. 3 (e) is the magnification region after distorting.

Fig. 4 detects embodiment for picture material increases.

Wherein, Fig. 4 (a) is the image that is embedded with watermark before distorting, the image of Fig. 4 (b) for distorting, and Fig. 4 (c) is for distorting testing result.

Embodiment:

Below in conjunction with drawings and Examples technical scheme of the present invention is further described.1) chaos iteration

With system key K is the initial value of chaos iteration, obtains chaos random sequence { n _l| 1≤n _l≤ 2 ^m, l=1,2 ..., N ², this sequence is as the number of times of following chaos iteration.

In the scheme that the present invention proposes, repeatedly use chaos system to produce pseudo random sequence.Chaos is nonlinear equation (group) separating under given conditions, and exporting at random by an equation (group) of determining of it determines.The initial parameter of chaos system can be used as the key of whole watermaking system, and the characteristics of performance at random that chaos system is good and regeneration have easily increased the security performance of system.Below be example with the one-dimensional discrete chaotic maps, f:U → U, U R

Z (n+1)=f (λ, z (n)), λ ∈ R, z (n) ∈ U is n=0 wherein, and 1, Λ represents iterations, λ is the parameter of control system chaotic behavior.The following formula iteration obtains real-valued chaos sequence for n time and is designated as

Z ⁿ=z (1), z (2) ..., z (n) } and be defined as follows quantization function

V=Q (Z ⁿ, d) wherein $V\stackrel{\mathrm{\Δ}}{=}\{v_1,v_2,...{,v}_d|v_i\∈\left\{\mathrm{0,1}\right\}\},$ , d represent to return 0, the number of 1} bit.Because chaos to the extreme sensitivity of initial value, therefore for different slightly initial values, will obtain diverse two sequences.

If E={E ₀, E ₁..., E _Q-1Be limited cutting apart on the U, $Y_{j=0}^{q-1}{E}_j=U,E_i\∩E_j=\mathrm{\Φ}(i\≠j)$ 。The limited E of cutting apart according to mapping f can obtain symbol sebolic addressing ω, and if only if measure entropy h (f)=log ₂During q, ω is the Bernoulli sequence.Decimal system chaos sequence can be got the m bit in turn by binary symbol sequence ω and obtain.2) generation of watermark

For size is that (i, j), (1≤i, j≤N), its LSB plane zero setting obtains for the original image X of N * N

With Initial value z as chaos iteration _{I, j}(0), establishes mapping C for this reason ₀Satisfy $C_0\left(\tilde{X}(i,j)\right)\∈U,$ , then watermark signal S:S (i, j) ∈ 0,1} is produced by following formula: $z_{i,j}\left(0\right)=C_0\left(\tilde{X}(i,j)\right)$ $S(i,j)=Q(Z_{i,j}^{n_l},1)$ Iterations n wherein _lGet the decimal system chaos sequence { n that is produced by system key K successively _l| 1≤n _l≤ 2 ^m, l=1,2 ..., N ²In element, K can be used as the key of whole fragile watermark system.3) embedding of watermark

Watermarked image is obtained by following formula $\mathrm{WX}=\mathrm{Add}(\tilde{X},S)$ (a, b) expression is embedded into two-value plane b the operation on the LSB plane of a to function Add.

Such scheme causes the increase of computation complexity because each picture element is all wanted the chaotic maps iteration repeatedly.Improve such scheme below, Be divided into size and be the fritter of c * c, for each fritter

(1≤i, j≤c), calculate $z_r\left(0\right)=C_0\left(\underset{i=1}{\overset{c}{\mathrm{\Σ}}}\underset{i=1}{\overset{c}{\mathrm{\Σ}}}\tilde{X}(i,j)\right)$ $S_r=Q(Z_r^{n_r},c^2)$ Wherein $\left\{n_r\right|1\≤n_r{\≤2}^m,r=\mathrm{1,2},...,{\left(\frac{N}{c}\right)}^2\}$ 。Improve algorithm with all pixel gray values of each fritter and be mapped as the chaos initial value, each chaotic maps iteration is returned the binary sequence of block sizes number.Note S ': $S^{\′}\stackrel{\mathrm{\Δ}}{=}(S_1,S_2,...,S_{{\left(\frac{N}{c}\right)}^2})$ Be the watermark of improvement project, the samely obtain watermarked image $W{X}^{\′}=\mathrm{Add}(\tilde{X},S^{\′})$ 4) watermark extracting and authentication

The watermark extracting algorithm block diagram to the LSB plane zero setting of the watermarked image received, is used as the initial value of chaos with its pixel gray value as shown in Figure 2, according to the chaos iteration number of times that is generated by key, obtains watermark S by chaos iteration then _Ex

Discriminant function is distorted in definition:

T=|S _Ex-LSB (WX) | wherein LSB (.) is defined as the LSB plane of getting image.If S _Ex=LSB (WX), then T=[0] _{N * N}, show that original image is not distorted.Otherwise be that 1 point is represented the point distorted in the original image in the matrix T.

Effect of the present invention can be verified by following performance evaluation: sensitivity analysis

Fragile watermark requires the watermark of adding not discernable.Watermark signal is embedded into the LSB plane of image, establishes P _E1For declaring into 1 or declare into 0 error probability to 1 to 0, know P by the independence of each bit of LSB plane _E1=0.5.Analyze theoretically, algorithm of the present invention will obtain very high Y-PSNR value, satisfy the requirement of fragile watermark to imperceptibility.The tampering location capability analysis

Utilize the characteristics of chaotic maps, to adding distorting this algorithm and locating exactly of watermarking images, thereby make it to have content tampering proof and integrity certification ability the initial value extreme sensitivity.

Pseudo-random characteristics by chaos knows, as fruit dot WX (i j) is distorted, and then passes through chaos iteration, detect this point and distorted, promptly T (i, probability j)=1 is:

P{T (i j)=so 1}=0.5 is for the image that adds watermark, on the probability statistics meaning, will have 50% distort a little and be detected, and these be detected distort the tampered region that is distributed in of naming a person for a particular job at random.Computation complexity is analyzed

For the watermarking algorithm of a practicality, require its computation complexity low, the real-time processing that so easy realization watermark embeds and extracts.Watermark proposed by the invention embeds and extracts is the process of two symmetries, and its computation complexity is the same order of magnitude.For embedding algorithm, its computation complexity is by the number of times decision of original image size and chaos iteration.If the mathematic expectaion of the number of times of the chaos iteration of each point is E (n ^r), the computation complexity that then embeds algorithm is: $R_c=O(E\left(n^r\right)\×{\left(\frac{N}{c}\right)}^2)$ As seen, $R_c\∝\frac{1}{c^2},$ The c value is big more, and computation complexity is low more, but positioning accuracy will decrease, and compromisely considers so must do on computation complexity and positioning accuracy.

Embodiment

With Matlab 6 simulation algorithm of the present invention, what be of practical significance below the consideration distorts embodiment.

Revise among the embodiment in a picture material of the present invention, for 512 * 512 * 8 ' Jet ' gray-scale map, the image that adds watermark is shown in Fig. 3 (a), and Y-PSNR is 51.1dB.Suppose that " US AIR FORCE " on the airframe among ' Jet ' figure made into " UN AIR FORCE " promptly makes alphabetical S into N, shown in Fig. 3 (b), for the content that observation post more clearly distorts, Fig. 3 (d) and Fig. 3 (e) are the enlarged image of this zone before and after distorting.Utilize the algorithm that the present invention proposes to detect distorting of being done, the tampered region testing result is shown in Fig. 3 (c), and visible the present invention has good sensitiveness and the tampering location ability of distorting.

Increase among the embodiment in a picture material of the present invention, for 512 * 512 * 8 ' Peppers ' gray-scale map, the image that adds watermark is shown in Fig. 4 (a), and Y-PSNR is 51.2dB.If add two capsicums again shown in Fig. 4 (b) on the image that adds watermark, the algorithm that utilizes the present invention to propose detects this and distorts, the tampered region testing result is shown in Fig. 4 (c).As seen the present invention can point-devicely locate distorting of picture material, and the tampered region profile is obvious.

The present invention is very responsive as additive Gaussian noise, JPEG compression, stripping, medium filtering, sharpening, histogram equalization, veining, shearing, Gaussian Blur etc. to common image processing operation in addition, can both locate seizure well to the tampered region.

Claims (1)

1, a kind of location type chaos fragile digital watermark embeds and extracting method, comprises following concrete steps:

1) is the initial value of chaotic maps with system key K, obtains the chaos random sequence

{ n _l| 1≤n _l≤ 2 ^m, l=1,2 ..., N ², this sequence is as the number of times of following chaos iteration;

2) (i, j), (1≤i, the minimum bit position LSB plane zero setting of j≤N) obtains original image X

By mapping As the initial value of chaotic maps, and with sequence n _lAs iterations, obtain a real-valued sequence of chaos thus ,

By quantize function Q obtain watermark information S (i, j);

3) right

Each point implement aforesaid operations, obtain watermark S, S is embedded into watermark The LSB plane, obtained watermarked image WX thus;

4) during watermark extracting,, be used as the initial value of chaos, according to the chaos iteration number of times that generates by key, obtain watermark S then by chaos iteration with its pixel gray value to the LSB plane zero setting of the watermarked image received _Ex, compare S _ExWith the corresponding position on the LSB plane of WX, if S _ExAll the same with corresponding position, the LSB plane of WX, show that original image is not distorted, otherwise point inequality is detected distorting a little.