CN105389517A - Method for hiding secret information in images - Google Patents

Abstract

The invention discloses a method for hiding secret information in images, and mainly aims to solve the problem that consideration cannot be given to both robustness and security in the prior art. The method comprises the following steps: 1) a sender selects carrier images from a carrier image library via a shared key, conducts block segmentation on the selected carrier images, and selects a pseudorandom number generator for each block, so as to acquire the probability distribution of candidate hiding carriers; 2) the sender creates a characteristic matrix of the carrier images; 3) the sender acquires a final characteristic matrix set and the key space size; 4) the sender hides the secret information in carriers via the final characteristic matrix set and the key space size; 5) the sender sends an obtained carrier set in which the secret information is embedded to a receiver; and 6) the receiver acquires the secret information. The method can improve the security of a hiding system while guaranteeing the robustness of the hiding system, and can be used for changing the hiding position of the secret information in an image in real time when the contents of the image are changed.

Description

Image is utilized to carry out the hiding method of secret information

Technical field

The invention belongs to field of information security technology, particularly a kind ofly hide secret method in the picture, be used in picture material when changing, change secret information stowed position in the picture in real time.

Background technology

The international and domestic scientific and technical literature hidden about image secret information has a lot, in these documents, and the main consideration way in spatial domain and the way of transform domain.Concealing technology wherein based on spatial domain mainly adopts least significant bit (LSB) LSB algorithm.Transmit leg, before transmission information, is replaced the least significant bit (LSB) of carrier image pixel one by one, thus secret information is hidden in carrier image with the bit of scale-of-two secret information.The hidden capacity of spatial information (si) concealing technology is large, but it is difficult to attack such as opposing filtering, compression etc., and institute is inadequate robust in this way.

Information Hiding Techniques based on transform domain mainly adopts different conversion, transfers in the amendment to coefficient in transform domain to the amendment of pixel data.Conventional conversion comprises: discrete cosine transform, wavelet transform DWT, discrete Fourier transform (DFT) DFT.Principle of operation is, transmit leg, before transmission information, carries out piecemeal to carrier image, and chooses suitable piecemeal and convert, and such as, DCT, then embeds secret information in the middle of coefficient after the conversion.Then, the piecemeal after changing is carried out inverse transformation, obtains the carrier comprising secret information.The hidden capacity of transform domain information concealing technology is not high, but possesses strong robustness, can effectively resist the attack such as filtering, compression.

As seen from the above description, existing hiding way mainly considers hiding not sentience and robustness, does not relate to security.This will make assailant can be obtained the key of system by the security of breaking through system, and then obtains the message hidden, and now the strong robustness of system is lost its meaning existed.And show according to current research data, can simultaneously compatible robustness and security really blanking method not yet propose.

Therefore, while the robustness improving the system of hiding, can take into account security again, be prior art urgent problem.

Summary of the invention

The object of the invention is to the deficiency existed for above-mentioned prior art, propose a kind of image that utilizes and carry out the hiding method of secret information.With while taking into account hiding system robustness, improve its security.

Technical scheme of the present invention is achieved in that

One. know-why:

In order to solve the safety issue in Information Hiding Techniques, gordian technique of the present invention sets up the eigenmatrix set A of carrier image.

The foundation of carrier image eigenmatrix set refers to the distribution describing carrier data with matrix.The distribution of carrier data refers to the distribution of image pixel or the distribution of image block.Compared by the average distributed with carrier data, constantly divide raw data, the data of the cloth that makes to secure satisfactory grades occupy more element in a matrix, on the contrary, then occupy less element.So only by eigenmatrix just can very directly demonstrate different secret information under different choice with the corresponding relation of carrier.Meanwhile, in eigenmatrix, in every a line, non-vanishing unit have and only has one, and position with element non-vanishing in a line in different matrix is mutually orthogonal.Therefore, eigenmatrix set A not only make the selection of carrier follow himself distribution and also can corresponding relation between scramble itself and secret information.The security of the system of hiding can be expressed as: Security=I (A, k), wherein I represents the feature in carrier image, and k represents shared key.

In order to realize the robustness of hiding system, when embedding secret information, obtain transformable insert depth α according to carrier content.

Insert depth α is mainly realized by the variance calculating relevant position local data.When the variance of local data is larger, illustrate that the change of data is larger, image local pixel is comparatively disperseed, and can embed more information, then can only embed the information of small amount on the contrary.The robustness of the system of hiding can be expressed as Robust=I (α, λ), and wherein, I represents the feature in carrier image, and λ represents drawing coefficient, is used for controlling overall embedding effect.

Two. symbol and abbreviation

K is shared key;

T is threshold value when choosing carrier;

C _sfor candidate hides collection of vectors;

N is the space size that candidate hides collection of vectors;

P _c={ p ₁, p ₂..., p _i..., p _nit is the probability distribution of carrier;

A={A ₁, A ₂..., A _j..., A _mit is eigenmatrix set;

| K| is key space size;

M is the space size of secret information;

δ is probability distribution P _cthe amendment variable of middle element;

A '={ A ₁', A ₂' ..., A _j' ..., A _m' be final eigenmatrix set;

X=(x ₁, x ₂... x _j..., x _m) be secret information;

for hiding secret information x _jcarrier;

α _jfor carrier insert depth;

λ is drawing coefficient;

S _jfor hiding secret information x _jafter carrier;

S={s ₁, s ₂..., s _j..., s _mfor embedding the collection of vectors after secret information;

X ' is the secret information estimated from the collection of vectors S after embedding secret information.

Three. performing step:

According to above-mentioned principle, performing step of the present invention comprises as follows:

(1) transmit leg utilizes shared key k from carrier image storehouse, choose carrier image to carry out block comminute, and selects pseudorandom number generator to each block, using key k as initial seed, obtains candidate and hides support C _sprobability distribution P _c={ p ₁, p ₂..., p _i..., p _n, p _iprobability distribution P _cin i-th element, i=1 ..., n, n are that candidate hides collection of vectors C _sspace size;

(2) eigenmatrix of transmit leg creating plasmid image:

(2a) transmit leg initialization feature set of matrices A={A ₁, A ₂..., A _j..., A _mand key space size | K|, is set to sky, simultaneously by key space size by each matrix in eigenmatrix set A | and K| is set to zero, performs (2b), wherein A _ja jth matrix in representation feature set of matrices A, j=1 ..., m;

(2b) transmit leg judges probability distribution P _celement whether be zero entirely: if be zero entirely, do not change eigenmatrix set A and key space size | K|, perform end, obtain final eigenmatrix set A ', the space values size of this A ' equals the space values size m of secret information, the numerical values recited of the middle row of A ' equals key space numerical values recited | and the numerical values recited of K| ', A ' middle row equals candidate and hides collection of vectors C _sspace values size n; Otherwise, perform (2c);

(2c) transmit leg is to probability distribution P _cin element arrange according to order from big to small, namely perform (2d), wherein represent the rear P of sequence _cin i-th element, σ _irepresent subscript, i=1 ..., n;

(2d) transmit leg amendment sequence posterior probability distribution P _cfront m element, make all the other elements do not change, and perform (2e), wherein amended probability distribution P _cin element, the numerical values recited of m is identical with the space size of secret information, and 1≤m < n, and δ is probability distribution P _cthe amendment variable of middle element;

(2e) return (2b) and rejudge amended probability distribution P _c, revise each eigenmatrix A in eigenmatrix set A simultaneously _j, make A _j=A _j+ δ e _t, and key space is increased to | K|=|K|+m, wherein e representation unit row vector, t represents that e intermediate value is the coordinate of the element of 1, t=((z+j) modm)+1, z, j=1 ..., m; ;

(3) transmit leg obtains final eigenmatrix set A '={ A ₁', A ₂' ..., A _j' ..., A _m' and key space size | K| ', wherein A _j' represent the jth eigenmatrix of A ' in final eigenmatrix set, | K| ' represents final key space size;

(4) transmit leg utilizes final eigenmatrix set A ' and key space size | and secret information is hidden in carrier by K| ':

(4a) transmit leg utilizes secret information X=(x ₁, x ₂... x _j..., x _m) from eigenmatrix set A ' choose and secret information x _jcorresponding matrix A _j', wherein x _jrepresent the jth element in secret information X, j=1 ..., m;

(4b) transmit leg utilizes A _j' choose A with key k _jthe carrier that the element column of ' (k, i) ≠ 0 is corresponding as hiding secret information x _jcarrier, wherein A _j' (k, i) representing matrix A _jthe element that ' middle row k i-th arranges, i=1 ..., n, j=1 ..., m, carrier that candidate hides collection of vectors C _sin an element,

(4c) transmit leg calculates selected carrier variance definition secret information insert depth α _j, j=1 ..., m;

(4d) transmit leg utilizes insert depth α _j, secret information is embedded in corresponding carrier data, and the main additivity data modification mode that adopts embeds secret information, wherein s _jrepresent and hide secret information x _jafter carrier;

(5) transmit leg is by the collection of vectors S={s after the embedding secret information of acquisition ₁, s ₂..., s _j..., s _msend to take over party;

(6) take over party obtains secret information:

(6a) the secret information X ' in the collection of vectors S after take over party adopts filtering mode to estimate embedding secret information;

(6b) take over party according to the secret information X ' estimated utilize final eigenmatrix set A ', obtain the corresponding relation estimated between secret information X ' and hiding carrier;

(6c) take over party is according to the corresponding relation between the estimation secret information X ' obtained and hiding carrier, carries out reciprocal permutation obtain real secret information X to estimation secret information X '.

The present invention compared with prior art has the following advantages:

The first, security is high.

In the present invention, the random selecting of secret information embedded and the random selecting of carrier are independently, carry out respectively, for assailant, according to single impersonation attack, two-way random selecting obviously than unidirectional random selecting secret information or the unidirectional difficulty choosing carrier larger, thus the security of system is higher.

The second, promote and extend range of application

The present invention realizes carrying out in the picture by corresponding relation between eigenmatrix scramble carrier and secret information the way that secret information hides, and when when carrier and secret information, one of them changes, corresponding relation therebetween also changes.Therefore, in actual application, the present invention is not only applicable to hiding of single image, and is applicable to hiding of multiple image, has more extensibility.

3rd, compatible good

The present invention introduces transformable insert depth α and controls secret information embedding in the picture, insert depth α can change along with the change of view data, thus the embedding of secret information is also changed with the change of view data, finally reach the good compatibility of robustness and not sentience.

Accompanying drawing explanation

Fig. 1 is overall flow figure of the present invention;

Fig. 2 is the sub-process figure of transmit leg creating plasmid image characteristic matrix in the present invention;

Fig. 3 is the sub-process figure that in the present invention, transmit leg embeds secret information;

Fig. 4 is the sub-process figure that in the present invention, take over party extracts secret information.

Embodiment

Embodiment of the present invention are further illustrated below by the drawings and specific embodiments.

With reference to Fig. 1, the present invention is that performing step is as follows:

Step 1, the probability distribution of transmit leg determination carrier.

(1a) transmit leg utilizes shared key k to choose carrier image from carrier image storehouse;

(1b) transmit leg is divided into the carrier image chosen the fritter of 8 × 8, and calculates the fuzzy value of each fritter;

(1b1) ambiguity in definition set degree of membership is as follows:

$u(p,q)=\frac{1}{1+|d(p,q)-d^{\&prime;}|/N},$

Wherein, d (p, q) is image (p, q) pixel value at place, d ' is the normalized value of image (p, q) place pixel, p, q=1 ..., 8, N is constant, and its value will ensure that the codomain scope of u (p, q) is between 0.5 to 1, i.e. 0.5≤u (d (p, q))≤1;

(1b2) fuzzy value defining image (p, q) position pixel on the basis of fuzzy set degree of membership is as follows:

H(p,q)＝-(u(p,q))log ₂(u(p,q))-(1-u(p,q))log ₂(1-u(p,q)),

(1b3) fuzzy value of fritter is defined according to the fuzzy value H (p, q) of image (p, q) position pixel as follows:

$R\left(p_0,q_0\right)=\frac{1}{a\&times;a}\underset{k=-\left(a-1\right)/2}{\overset{\left(a-1\right)/2}{\mathrm{\&Sigma;}}}\underset{l=-\left(a-1\right)/2}{\overset{\left(a-1\right)/2}{\mathrm{\&Sigma;}}}H\left(p_0+k,q_0+l\right),$

Wherein, p ₀represent the horizontal ordinate of fritter central point, q ₀represent the ordinate of fritter central point, a represents the length of fritter, and a=8, k are the variablees controlling the movement of fritter central point horizontal ordinate, and l is the variable controlling the movement of fritter central point ordinate, $k=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},l=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},k\&NotEqual;l;$

(1c) fuzzy value of each fritter and given threshold value T compare by transmit leg: if fuzzy value is more than or equal to threshold value T, then fritter corresponding for this fuzzy value is joined candidate and hide collection of vectors C _sin; Otherwise, abandon, until calculated all image blocks;

(1d) transmit leg hides collection of vectors C to candidate _sin each fritter select pseudorandom number generator, using key k as initial seed, obtain candidate hide support C _sprobability distribution P _c={ p ₁, p ₂..., p _i..., p _n, p _iprobability distribution P _cin i-th element, i=1 ..., n, n are that candidate hides collection of vectors C _sspace size.

Step 2, the eigenmatrix of transmit leg creating plasmid.

With reference to Fig. 2, being implemented as follows of this step:

(2a) transmit leg initialization feature set of matrices A={A ₁, A ₂..., A _j..., A _mand key space size | K|, is set to sky, simultaneously by key space size by each matrix in eigenmatrix set A | and K| is set to zero, performs (2b), wherein A _ja jth matrix in representation feature set of matrices A, j=1 ..., m;

(2b) transmit leg judges probability distribution P _celement whether be zero entirely: if be zero entirely, then do not change eigenmatrix set A and key space size | K|, perform end, obtain final eigenmatrix set A ', the space values size of this A ' equals the space values size m of secret information, the numerical values recited of the middle row of A ' equals key space numerical values recited | and the numerical values recited of K| ', A ' middle row equals candidate and hides collection of vectors C _sspace values size n; Otherwise, perform (2c);

(2c) transmit leg is to probability distribution P _cin element arrange according to order from big to small, namely perform (2d), wherein represent the rear P of sequence _cin i-th element, σ _irepresent subscript, i=1 ..., n;

(2d) transmit leg amendment sequence posterior probability distribution P _cfront m element, make all the other elements do not change, and perform (2e), wherein amended probability distribution P _cin element, the numerical values recited of m is identical with the space size of secret information, and 1≤m < n, and δ is probability distribution P _cthe amendment variable of middle element;

(2e) return (2b) and rejudge amended probability distribution P _c, revise each eigenmatrix A in eigenmatrix set A simultaneously _j, make A _j=A _j+ δ e _t, and key space is increased to | K|=|K|+m, wherein e representation unit row vector, t represents that e intermediate value is the coordinate of the element of 1, t=((z+j) modm)+1, z, j=1 ..., m, z ≠ j;

Step 3, transmit leg obtains final eigenmatrix set A '={ A ₁', A ₂' ..., A _j' ..., A _m' and key space size | K| ', wherein A _j' represent the jth eigenmatrix of A ' in final eigenmatrix set, | K| ' represents final key space size.

Step 4, transmit leg utilizes final eigenmatrix set A ' and key space size | and secret information is hidden in carrier by K| '.

With reference to Fig. 3, the specific implementation of this step is as follows:

(4a) transmit leg utilizes secret information X=(x ₁, x ₂... x _j..., x _m) from eigenmatrix set A ' choose and secret information x _jcorresponding matrix A _j', wherein x _jrepresent the jth element in secret information X, j=1 ..., m;

(4b) transmit leg utilizes A _j' choose A with key k _jthe carrier that the element column of ' (k, i) ≠ 0 is corresponding as hiding secret information x _jcarrier, wherein A _j' (k, i) representing matrix A _jthe element that ' middle row k i-th arranges, i=1 ..., n, j=1 ..., m, carrier that candidate hides collection of vectors C _sin an element,

(4c) transmit leg is according to selected carrier variance calculate the insert depth α of secret information _j:

(4c1) calculate the average of carrier data, formula is as follows:

$\stackrel{\&OverBar;}{d}=\frac{1}{{(2\mathrm{\&Delta;}+1)}^2}\underset{k^{\&prime;}=1}{\overset{2\mathrm{\&Delta;}+1}{\&Sigma;}}\underset{l^{\&prime;}=1}{\overset{2\mathrm{\&Delta;}+1}{\&Sigma;}}d(k^{\&prime;},l^{\&prime;}),$

Wherein, Δ represents the control variable of moving step length, and k ' is the variable controlling the movement of carrier data horizontal ordinate, and l ' is the variable controlling the movement of carrier data ordinate, k '=1 ..., (2 Δ+1), l '=1, ..., (2 Δ+1), k ' ≠ l ';

(4c2) average is utilized the variance of definition carrier data is as follows:

$\mathrm{\&sigma;}(u,v)=\frac{1}{{(2\mathrm{\&Delta;}+1)}^2}\sqrt{\underset{e=u-\mathrm{\&Delta;}}{\overset{u+\mathrm{\&Delta;}}{\&Sigma;}}\underset{f=v-\mathrm{\&Delta;}}{\overset{v+\mathrm{\&Delta;}}{\&Sigma;}}{\left(d\right(u+e,v+f)-\stackrel{\&OverBar;}{d})}^2},$

Wherein, u represents the horizontal ordinate of carrier center point, v represents the ordinate of carrier center point, and e is the variable controlling the movement of carrier center point horizontal ordinate, and f is the variable controlling the movement of carrier data ordinate, e=(u-Δ), ...., (u+ Δ), f=(v-Δ), ..., (v+ Δ);

(4c3) the insert depth α of carrier image piecemeal is defined according to the variances sigma (u, v) of carrier data _jas follows:

α _j＝λσ(u,v)，

Wherein, λ represents drawing coefficient;

(4d) transmit leg utilizes insert depth α _j, secret information is embedded in corresponding carrier data, and the main additivity data modification mode that adopts embeds secret information, wherein s _jrepresent and hide secret information x _jafter carrier.

Step 5, transmit leg is by the collection of vectors S={s after the embedding secret information of acquisition ₁, s ₂..., s _j..., s _msend to take over party.

Step 6, take over party obtains secret information.

With reference to Fig. 4, the specific implementation of this step is as follows:

(6a) the secret information X ' in the collection of vectors S after take over party adopts filtering mode to estimate embedding secret information;

(6b) take over party according to the secret information X ' estimated utilize final eigenmatrix set A ', obtain the corresponding relation estimated between secret information X ' and hiding carrier;

(6c) take over party is according to the corresponding relation between the estimation secret information X ' obtained and hiding carrier, carries out reciprocal permutation obtain real secret information X to estimation secret information X '.

More than describing is only example of the present invention, does not form any limitation of the invention.Obviously for those skilled in the art; after having understood content of the present invention and principle; all may when not deviating from the principle of the invention, structure; carry out the various correction in form and details and change, but these corrections based on inventive concept and change are still within claims of the present invention.

Claims (4)

1. utilize image to carry out the hiding method of secret information, comprising:

(1) transmit leg utilizes shared key k from carrier image storehouse, choose carrier image to carry out block comminute, and selects pseudorandom number generator to each block, using key k as initial seed, obtains candidate and hides support C _sprobability distribution P _c={ p ₁, p ₂..., p _i..., p _n, p _iprobability distribution P _cin i-th element, i=1 ..., n, n are that candidate hides collection of vectors C _sspace size;

(2) eigenmatrix of transmit leg creating plasmid image:

(2a) transmit leg initialization feature set of matrices A={A ₁, A ₂..., A _j..., A _mand key space size | K|, is set to sky, simultaneously by key space size by each matrix in eigenmatrix set A | and K| is set to zero, performs (2b), wherein A _ja jth matrix in representation feature set of matrices A, j=1 ..., m;

(2b) transmit leg judges probability distribution P _celement whether be zero entirely: if be zero entirely, do not change eigenmatrix set A and key space size | K|, perform end, obtain final eigenmatrix set A ', the space values size of this A ' equals the space values size m of secret information, the numerical values recited of the middle row of A ' equals key space numerical values recited | and the numerical values recited of K| ', A ' middle row equals candidate and hides collection of vectors C _sspace values size n; Otherwise, perform (2c);

(2c) transmit leg is to probability distribution P _cin element arrange according to order from big to small, namely , perform (2d), wherein represent the rear P of sequence _cin i-th element, σ _irepresent subscript, i=1 ..., n;

(2d) transmit leg amendment sequence posterior probability distribution P _cfront m element, make all the other elements do not change, and perform (2e), wherein amended probability distribution P _cin element, the numerical values recited of m is identical with the space size of secret information, and 1≤m < n, and δ is probability distribution P _cthe amendment variable of middle element;

(2e) return (2b) and rejudge amended probability distribution P _c, revise each eigenmatrix A in eigenmatrix set A simultaneously _j, make A _j=A _j+ δ e _t, and key space is increased to | K|=|K|+m, wherein e representation unit row vector, t represents that e intermediate value is the coordinate of the element of 1, t=((z+j) modm)+1, z, j=1 ..., m;

(3) transmit leg obtain final eigenmatrix set A '=A ' ₁, A ' ₂..., A ' _j..., A ' _mand key space size | K| ',

Wherein A ' _jrepresent the jth eigenmatrix of A ' in final eigenmatrix set, | K| ' represents final key space size;

(4) transmit leg utilizes final eigenmatrix set A ' and key space size | and secret information is hidden in carrier by K| ':

(4a) transmit leg utilizes secret information X=(x ₁, x ₂... x _j..., x _m) from eigenmatrix set A ' choose and secret information x _jcorresponding matrix A ' _j, wherein x _jrepresent the jth element in secret information X, j=1 ..., m;

(4b) transmit leg utilizes A ' _ja ' is chosen with key k _jthe carrier that the element column of (k, i) ≠ 0 is corresponding as hiding secret information x _jcarrier, wherein A ' _j(k, i) representing matrix A ' _jthe element that middle row k i-th arranges, i=1 ..., n, j=1 ..., m, carrier that candidate hides collection of vectors C _sin an element,

(4c) transmit leg calculates selected carrier variance definition secret information insert depth α _j, j=1 ..., m;

(4d) transmit leg utilizes insert depth α _j, secret information is embedded in corresponding carrier data, and the main additivity data modification mode that adopts embeds secret information, wherein s _jrepresent and hide secret information x _jafter carrier;

(5) transmit leg is by the collection of vectors S={s after the embedding secret information of acquisition ₁, s ₂..., s _j..., s _msend to take over party;

(6) take over party obtains secret information:

(6a) the secret information X ' in the collection of vectors S after take over party adopts filtering mode to estimate embedding secret information;

(6b) take over party according to the secret information X ' estimated utilize final eigenmatrix set A ', obtain the corresponding relation estimated between secret information X ' and hiding carrier;

(6c) take over party is according to the corresponding relation between the estimation secret information X ' obtained and hiding carrier, carries out reciprocal permutation obtain real secret information X to estimation secret information X '.

2. the method utilizing image to carry out secret information to hide according to claim 1, it is characterized in that, step carries out block comminute to carrier image in (1), be first carrier image is divided into 8 × 8 fritter, and calculate the fuzzy value of each fritter; Again fuzzy value and given threshold value T are compared: if fuzzy value is more than or equal to threshold value T, then fritter corresponding for this fuzzy value is joined candidate and hide collection of vectors C _sin; Otherwise, abandon, until calculated all image blocks.

3. method of carrier image being carried out to block comminute according to claim 2, is characterized in that, calculates the fuzzy value of each fritter, carries out as follows:

First, ambiguity in definition set degree of membership is as follows:

$u(p,q)=\frac{1}{1+|d(p,q)-d^{\&prime;}|/N},$

Wherein, d (p, q) is image (p, q) pixel value at place, d ' is the normalized value of image (p, q) place pixel, p, q=1 ..., 8, N is constant, and its value will ensure that the codomain scope of u (p, q) is between 0.5 to 1, i.e. 0.5≤u (d (p, q))≤1;

Secondly, the fuzzy value that the basis of fuzzy set degree of membership defines image (p, q) position pixel is as follows:

H(p,q)＝-(u(p,q))log ₂(u(p,q))-(1-u(p,q))log ₂(1-u(p,q)),

Finally, the fuzzy value of fritter is defined according to the fuzzy value H (p, q) of image (p, q) position pixel as follows:

$R(p_0,q_0)=\frac{1}{a\&times;a}\underset{k=-(a-1)/2}{\overset{(a-1)/2}{\&Sigma;}}\underset{l=-(a-1)/2}{\overset{(a-1)/2}{\&Sigma;}}H(p_0+k,q_0+l),$

Wherein, p ₀represent the horizontal ordinate of fritter central point, q ₀represent the ordinate of fritter central point, a represents the length of fritter, and a=8, k are the variablees controlling the movement of fritter central point horizontal ordinate, and l is the variable controlling the movement of fritter central point ordinate, $k=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},l=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},k\&NotEqual;l.$

4. the image that utilizes according to claim 1 carries out the hiding method of secret information, it is characterized in that step 4c) the middle insert depth α calculating carrier _j, carry out as follows:

First calculate the average of carrier data, formula is as follows:

$\stackrel{\&OverBar;}{d}=\frac{1}{{(2\mathrm{\&Delta;}+1)}^2}\underset{k^{\&prime;}=1}{\overset{2\mathrm{\&Delta;}+1}{\&Sigma;}}\underset{l^{\&prime;}=1}{\overset{2\mathrm{\&Delta;}+1}{\&Sigma;}}d(k^{\&prime;},l^{\&prime;}),$

Wherein, Δ represents the control variable of moving step length, and k ' is the variable controlling the movement of carrier data horizontal ordinate, and l ' is the variable controlling the movement of carrier data ordinate, k '=1 ..., (2 Δ+1), l '=1 ..., (2 Δ+1);

Secondly, average is utilized the variance of definition carrier data is as follows:

$\mathrm{\&sigma;}(u,v)=\frac{1}{{(2\mathrm{\&Delta;}+1)}^2}\sqrt{\underset{e=u-\mathrm{\&Delta;}}{\overset{u+\mathrm{\&Delta;}}{\&Sigma;}}\underset{f=v-\mathrm{\&Delta;}}{\overset{v+\mathrm{\&Delta;}}{\&Sigma;}}{\left(d\right(u+e,v+f)-\stackrel{\&OverBar;}{d})}^2},$

Wherein, u represents the horizontal ordinate of carrier center point, v represents the ordinate of carrier center point, and e is the variable controlling the movement of carrier center point horizontal ordinate, and f is the variable controlling the movement of carrier data ordinate, e=(u-Δ), ...., (u+ Δ), f=(v-Δ), ..., (v+ Δ);

Finally, the insert depth α of carrier image piecemeal is defined according to the variances sigma (u, v) of carrier data _jas follows:

α _j＝λσ(u,v).

Wherein, λ represents drawing coefficient.