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

Abstract

The invention discloses a method for hiding secret information by using an image, which mainly solves the problem that the prior art cannot simultaneously take into account robustness and safety. The implementation steps are: 1) The sender uses the shared key to select the carrier image from the carrier image library for block segmentation, and selects a pseudo-random number generator for each block to obtain the probability distribution of candidate hidden carriers; 2) The sender creates carrier images Image feature matrix; 3) The sender obtains the final set of feature matrices and the size of the key space; 4) The sender uses the final set of feature matrices and the size of the key space to hide the secret information in the carrier; 5) The sender will obtain The carrier set embedded with secret information is sent to the receiver; 6) The receiver obtains the secret information. The invention improves the security of the concealment system while taking into account the robustness of the concealment system, and can be used to change the hiding position of secret information in the image in real time when the image content changes.

Description

A Method of Hiding Secret Information Using Images

technical field

The invention belongs to the technical field of information security, in particular to a method for hiding secrets in images, which can be used to change the hidden position of secret information in the images in real time when the content of the images changes.

Background technique

There are many international and domestic scientific and technological documents on image secret information hiding. In these documents, the method of air domain and the method of transform domain are mainly considered. Among them, the concealment technology based on the airspace mainly adopts the least significant bit LSB algorithm. Before sending the information, the sender replaces the least significant bits of the pixels of the carrier image with the bits of the binary secret information one by one, thereby hiding the secret information in the carrier image. The hiding capacity of spatial information hiding technology is large, but it is difficult to resist attacks such as filtering and compression, so this method is not robust enough.

The information hiding technology based on the transform domain mainly adopts different transformations, and transfers the modification of the pixel data to the modification of the coefficients of the transform domain. Commonly used transforms include: discrete cosine transform DCT, discrete wavelet transform DWT, discrete Fourier transform DFT. The operating principle is that before sending information, the sender divides the carrier image into blocks, selects appropriate blocks for transformation, for example, DCT, and then embeds the secret information into the transformed coefficients. Then, inverse transform the changed block to obtain the carrier containing the secret information. The hiding capacity of transform domain information hiding technology is not high, but it has strong robustness and can effectively resist attacks such as filtering and compression.

It can be seen from the above description that the existing hiding methods mainly consider the imperceptibility and robustness of hiding, and have never involved security. This will allow the attacker to obtain the key of the system by breaking through the security of the system, and then obtain the hidden message. At this time, the strong robustness of the system also loses its meaning of existence. And according to the latest research data, the exact method that can be compatible with both robustness and security has not yet been proposed.

Therefore, improving the robustness of the hidden system while taking into account the security is an urgent problem to be solved in the prior art.

Contents of the invention

The object of the present invention is to propose a method for hiding secret information by using images to address the shortcomings of the above-mentioned prior art. In order to improve its security while taking into account the robustness of the hidden system.

Technical scheme of the present invention is realized like this:

1. Technical principle:

In order to solve the security problem in the information hiding technology, the key technology of the present invention is to establish the feature matrix set A of the carrier image.

The establishment of carrier image feature matrix set refers to using matrix to describe the distribution of carrier data. The distribution of carrier data refers to the distribution of image pixels or the distribution of image blocks. By comparing with the mean value of the carrier data distribution, the original data is continuously divided, so that the highly distributed data occupies more elements in the matrix, on the contrary, it occupies fewer elements. In this way, only through the characteristic matrix, the corresponding relationship between different secret information and carriers under different selections can be directly displayed. At the same time, there is only one non-zero element in each row of the feature matrix, and the positions of non-zero elements in the same row in different matrices are orthogonal to each other. Therefore, the characteristic matrix set A not only makes the selection of the carrier follow its own distribution but also can scramble the corresponding relationship between it and the secret information. The security of the hidden system can be expressed as: Security=I(A,k), where I represents the feature in the cover image, and k represents the shared key.

In order to achieve the robustness of the hidden system, when embedding secret information, a variable embedding depth α is obtained according to the content of the carrier.

The embedding depth α is mainly achieved by computing the variance of the local data at the corresponding location. When the variance of the local data is larger, it means that the data changes more, and the local pixels of the image are scattered, which can embed more information, on the contrary, only a small amount of information can be embedded. The robustness of the hidden system can be expressed as Robust=I(α,λ), where I represents the features in the carrier image, and λ represents the stretch coefficient, which is used to control the overall embedding effect.

2. Symbols and abbreviations

k is the shared key;

T is the threshold when selecting the carrier;

C _S is the set of candidate hidden vectors;

n is the space size of the candidate hidden carrier set;

P _C ={p ₁ ,p ₂ ...,p _i ,...,p _n } is the probability distribution of the carrier;

A＝{A ₁ ,A ₂ ...,A _j ,...,A _m } is a feature matrix set;

|K| is the key space size;

m is the space size of secret information;

δ is the modification variable of the elements in the probability distribution P _C ;

A'={A ₁ ', A ₂ ',...,A _j ',...,A _m '} is the final feature matrix set;

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

is the carrier to hide the secret information x _j ;

α _j is the carrier the embedding depth;

λ is the stretch coefficient;

s _j is the carrier after hiding the secret information x _j ;

S={s ₁ ,s ₂ ,...,s _j ,...,s _m } is the set of carriers embedded with secret information;

X' is the secret information estimated from the carrier set S after embedding the secret information.

3. Implementation steps:

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

(1) The sender uses the shared key k to select the carrier image from the carrier image database for block segmentation, and selects a pseudo-random number generator for each block, and uses the key k as the initial seed to obtain the probability distribution of the candidate hidden carrier _CS P _C ={p ₁ ,p ₂ ...,p _i ,...,p _n }, p _i is the i-th element in the probability distribution P _C , i=1,...,n, n is The space size of the candidate hidden carrier set _CS ;

(2) The sender creates the feature matrix of the carrier image:

(2a) The sender initializes the feature matrix set A={A ₁ ,A ₂ ...,A _j ,...,A _m } and the size of the key space |K|, and assigns each matrix in the feature matrix set A Set them all to be empty, and at the same time set the size of the key space |K| to zero, and execute (2b), where A _j represents the jth matrix in the feature matrix set A, j=1,...,m;

(2b) The sender judges whether the elements of the probability distribution P _C are all zeros: if they are all zeros, the feature matrix set A and the size of the key space |K| are not changed, and the execution ends, and the final feature matrix set A′ is obtained. The spatial numerical size of A' is equal to the spatial numerical size m of the secret information, the numerical size of the row in A' is equal to the key space numerical size |K|', and the numerical value of the column in _A ' is equal to the spatial numerical size of the candidate hidden carrier set CS n; otherwise, execute (2c);

(2c) The sender _arranges the elements in the probability distribution PC in descending order, namely Execute (2d), where Indicates the i _- th element in PC after sorting, and σ _i represents The subscript of i=1,...,n;

(2d) The sender modifies the first m elements of the sorted probability distribution P _C such that The rest of the elements do not change, execute (2e), where is the element in the modified probability distribution P _C , the value of m is the same as the size of the secret information space, and 1≤m<n, δ is the modified variable of the element in the probability distribution P _C ;

(2e) Return to (2b) to re-judge the modified probability distribution P _C , and at the same time modify each feature matrix A _j in the feature matrix set A, so that A _j =A _j +δe _t , and increase the key space to | K|=|K|+m, where e represents the unit row vector, t represents the coordinates of the element with a value of 1 in e, t=((z+j)modm)+1, z,j=1,... ,m;

(3) The sender obtains the final feature matrix set A′={A ₁ ′,A ₂ ′,...,A _j ′,...,A _m ′} and the size of the key space |K|′, where A _j 'indicates the jth feature matrix of A' in the final set of feature matrices, |K|'indicates the size of the final key space;

(4) The sender uses the final feature matrix set A' and the size of the key space |K|' to hide the secret information in the carrier:

(4a) The sender uses the secret information X=(x ₁ ,x ₂ ,...x _j ,...,x _m ) to select the matrix A _j ′ corresponding to the secret information x _j from the feature matrix set A′, Where x _j represents the jth element in the secret information X, j=1,...,m;

(4b) The sender uses A _j ' and key k to select the carrier corresponding to the column where the element of A _j '(k,i)≠0 is located As a carrier for hiding secret information x _j , where A _j ′(k,i) represents the element of row k and column i in matrix A _j ′, i=1,...,n, j=1,... ,m, carrier is an element in the set of _candidate hidden vectors CS,

(4c) The sender calculates the selected bearer The variance of defines the embedding depth α _j of secret information, j=1,...,m;

(4d) The sender uses the embedding depth α _j to embed the secret information into the corresponding carrier data, mainly adopting the additive data modification method to embed the secret information, Where s _j represents the carrier after hiding the secret information x _j ;

(5) The sender sends the obtained carrier set S={s ₁ ,s ₂ ,...,s _j ,...,s _m } after embedding the secret information to the receiver;

(6) The receiver obtains the secret information:

(6a) The receiver uses filtering to estimate the secret information X' in the carrier set S after embedding the secret information;

(6b) The receiver uses the final feature matrix set A' according to the estimated secret information X' to obtain the correspondence between the estimated secret information X' and the hidden carrier;

(6c) According to the corresponding relationship between the acquired estimated secret information X' and the hidden carrier, the receiver reverses the estimated secret information X' to obtain the real secret information X.

Compared with the prior art, the present invention has the following advantages:

First, high security.

In the present invention, the random selection of the embedded secret information and the random selection of the carrier are carried out independently and separately. For the attacker, if a single imitation attack is adopted, the two-way random selection is obviously better than the one-way random selection of the secret information or It is more difficult to select a carrier one-way, so the security of the system is higher.

Second, it improves and expands the scope of application

The present invention implements a method of hiding secret information in an image by scrambling the corresponding relationship between the carrier and the secret information through the characteristic matrix. When one of the carrier and the secret information changes, the corresponding relationship between the two also changes. Therefore, in the actual application process, the present invention is not only applicable to the concealment of a single image, but also applicable to the concealment of multiple images, and has more scalability.

Third, good compatibility

The present invention introduces a variable embedding depth α to control the embedding of secret information in the image. The embedding depth α will change with the change of image data, so that the embedding of secret information will also change with the change of image data, and finally achieves Good compatibility of stickiness and imperceptibility.

Description of drawings

Fig. 1 is the overall flowchart of the present invention;

Fig. 2 is a sub-flow chart of creating a carrier image feature matrix by the sender in the present invention;

Fig. 3 is a sub-flow chart of the sender embedding secret information in the present invention;

Fig. 4 is a sub-flow chart of the receiving party extracting secret information in the present invention.

detailed description

Embodiments of the present invention will be further described below with reference to the drawings and specific embodiments.

With reference to Fig. 1, the present invention is that realization steps are as follows:

Step 1, the sender determines the probability distribution of the bearers.

(1a) The sender uses the shared key k to select the carrier image from the carrier image library;

(1b) The sender divides the selected carrier image into small blocks of 8×8, and calculates the blur value of each small block;

(1b1) Define the membership degree of fuzzy set as follows:

$\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; ,</span>$

Among them, d(p,q) is the pixel value at the image (p,q), d′ is the normalized value of the pixel at the image (p,q), p,q=1,...,8,N It is a constant, and its value should ensure that the value range of u(p,q) is between 0.5 and 1, that is, 0.5≤u(d(p,q))≤1;

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

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

(1b3) Define the blur value of the small block according to the blur value H(p,q) of the pixel at the position (p,q) of the image as follows:

$\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; q</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; a</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; q</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$

Wherein, p ₀ represents the abscissa of the small block center point, q ₀ represents the ordinate of the small block center point, a represents the length of the small block, a=8, k is the variable that controls the movement of the small block center point abscissa, and l is The variable that controls the vertical coordinate movement of the center point of the small block, $k=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},l=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},k\&NotEqual;l;$

(1c) The sender compares the fuzzy value of each small block with a given threshold T: if the fuzzy value is greater than or equal to the threshold T, then add the small block corresponding to the fuzzy value to the candidate hidden carrier set _CS ; Otherwise, discard until all image blocks are calculated;

(1d) The sender selects a pseudo-random number generator for each small block in the candidate hidden carrier set CS, uses the key k as the initial seed, and obtains the probability distribution _{P C} ={p ₁ ,p ₂ of the candidate hidden carrier _CS ...,p _i ,...,p _n }, p _i is the i-th element in the probability distribution P _C , i=1,...,n, n is the space size of the candidate hidden carrier set C _S .

Step 2, the sender creates the feature matrix of the carrier.

Referring to Figure 2, the specific implementation of this step is as follows:

(2a) The sender initializes the feature matrix set A={A ₁ ,A ₂ ...,A _j ,...,A _m } and the size of the key space |K|, and assigns each matrix in the feature matrix set A Set them all to be empty, and at the same time set the size of the key space |K| to zero, and execute (2b), where A _j represents the jth matrix in the feature matrix set A, j=1,...,m;

(2b) The sender judges whether the elements of the probability distribution P _C are all zeros: if they are all zeros, the characteristic matrix set A and the size of the key space |K| are not changed, and the execution ends, and the final characteristic matrix set A′ is obtained. The spatial value of A' is equal to the spatial value of secret information m, the value of rows in A' is equal to the value of the key space |K|', and the value of columns in _A ' is equal to the spatial value of the candidate hidden carrier set CS size n; otherwise, execute (2c);

(2c) The sender _arranges the elements in the probability distribution PC in descending order, namely Execute (2d), where Indicates the i _- th element in PC after sorting, and σ _i represents The subscript of i=1,...,n;

(2d) The sender modifies the first m elements of the sorted probability distribution P _C such that The rest of the elements do not change, execute (2e), where is the element in the modified probability distribution P _C , the value of m is the same as the size of the secret information space, and 1≤m<n, δ is the modified variable of the element in the probability distribution P _C ;

(2e) Return to (2b) to re-judge the modified probability distribution P _C , and at the same time modify each feature matrix A _j in the feature matrix set A, so that A _j =A _j +δe _t , and increase the key space to | K|=|K|+m, where e represents the unit row vector, t represents the coordinates of the element with a value of 1 in e, t=((z+j)modm)+1, z,j=1,... , m, z≠j;

Step 3, the sender obtains the final feature matrix set A′={A ₁ ′,A ₂ ′,...,A _j ′,...,A _m ′} and the size of the key space |K|′, where A _j ′ denotes the jth feature matrix of A′ in the final feature matrix set, and |K|′ denotes the size of the final key space.

Step 4, the sender uses the final feature matrix set A' and the key space size |K|' to hide the secret information into the carrier.

Referring to Figure 3, the specific implementation of this step is as follows:

(4a) The sender uses the secret information X=(x ₁ ,x ₂ ,...x _j ,...,x _m ) to select the matrix A _j ′ corresponding to the secret information x _j from the feature matrix set A′, Where x _j represents the jth element in the secret information X, j=1,...,m;

(4b) The sender uses A _j ' and key k to select the carrier corresponding to the column where the element of A _j '(k,i)≠0 is located As a carrier for hiding secret information x _j , where A _j ′(k,i) represents the element of row k and column i in matrix A _j ′, i=1,...,n, j=1,... ,m, carrier is an element in the set of _candidate hidden vectors CS,

(4c) The sender selects the carrier according to The variance of calculates the embedding depth α _j of the secret information:

(4c1) Calculate the mean value of the carrier data, the formula is as follows:

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\underset{{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\underset{{\mathrm{<span\; class="notranslate">\; l</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; l</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$

Among them, Δ represents the control variable of the moving step, k' is the variable that controls the movement of the abscissa of the carrier data, l' is the variable that controls the movement of the ordinate of the carrier data, k'=1,...,(2Δ+1), l'=1,...,(2Δ+1), k'≠l';

(4c2) Using the mean Define the variance of the carrier data as follows:

$\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\sqrt{\underset{\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{\overset{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\underset{\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{\overset{\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ,</span>$

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

(4c3) Define the embedding depth α _j of the carrier image block according to the variance σ(u,v) of the carrier data as follows:

α _j =λσ(u,v),

Wherein, λ represents stretch coefficient;

(4d) The sender uses the embedding depth α _j to embed the secret information into the corresponding carrier data, mainly adopting the additive data modification method to embed the secret information, where s _j represents the carrier after hiding the secret information x _j .

Step 5, the sender sends the obtained carrier set S={s ₁ , s ₂ ,...,s _j ,...,s _m } after embedding the secret information to the receiver.

Step 6, the receiver obtains the secret information.

Referring to Figure 4, the specific implementation of this step is as follows:

(6a) The receiver uses filtering to estimate the secret information X' in the carrier set S after embedding the secret information;

(6b) The receiver uses the final feature matrix set A' according to the estimated secret information X' to obtain the correspondence between the estimated secret information X' and the hidden carrier;

(6c) According to the corresponding relationship between the acquired estimated secret information X' and the hidden carrier, the receiver reverses the estimated secret information X' to obtain the real secret information X.

The above description is only a specific example of the present invention, and does not constitute any limitation to the present invention. Obviously, for those skilled in the art, after understanding the content and principles of the present invention, it is possible to make various modifications and changes in form and details without departing from the principles and structures of the present invention, but these are based on the present invention. The modification and change of the inventive concept are still within the protection scope of the claims of the present invention.

Claims (4)

1. A method for hiding secret information using images, comprising: (1) The sender uses the shared key k to select the carrier image from the carrier image database for block segmentation, and selects a pseudo-random number generator for each block, and uses the key k as the initial seed to obtain the probability distribution of the candidate hidden carrier _CS P _C ={p ₁ ,p ₂ ...,p _i ,...,p _n }, p _i is the i-th element in the probability distribution P _C , i=1,...,n, n is The space size of the candidate hidden carrier set _CS ; (2) The sender creates the feature matrix of the carrier image: (2a) The sender initializes the feature matrix set A={A ₁ ,A ₂ ...,A _j ,...,A _m } and the size of the key space |K|, and assigns each matrix in the feature matrix set A Set them all to be empty, and at the same time set the size of the key space |K| to zero, and execute (2b), where A _j represents the jth matrix in the feature matrix set A, j=1,...,m; (2b) The sender judges whether the elements of the probability distribution P _C are all zeros: if they are all zeros, the feature matrix set A and the size of the key space |K| are not changed, and the execution ends, and the final feature matrix set A′ is obtained. The spatial numerical size of A' is equal to the spatial numerical size m of the secret information, the numerical size of the row in A' is equal to the key space numerical size |K|', and the numerical value of the column in _A ' is equal to the spatial numerical size of the candidate hidden carrier set CS n; otherwise, execute (2c); (2c) The sender _arranges the elements in the probability distribution PC in descending order, namely , execute (2d), where Indicates the i _- th element in PC after sorting, and σ _i represents The subscript of i=1,...,n; (2d) The sender modifies the first m elements of the sorted probability distribution P _C such that The rest of the elements do not change, execute (2e), where is the element in the modified probability distribution P _C , the value of m is the same as the size of the secret information space, and 1≤m<n, δ is the modified variable of the element in the probability distribution P _C ; (2e) Return to (2b) to re-judge the modified probability distribution P _C , and at the same time modify each feature matrix A _j in the feature matrix set A, so that A _j =A _j +δe _t , and increase the key space to | K|=|K|+m, where e represents the unit row vector, t represents the coordinates of the element with a value of 1 in e, t=((z+j)modm)+1, z,j=1,... ,m; (3) The sender obtains the final feature matrix set A'={A' ₁ ,A' ₂ ,...,A' _j ,...,A' _m } and the size of the key space |K|', Where A' _j represents the jth feature matrix of A' in the final set of feature matrices, and |K|' represents the final key space size; (4) The sender uses the final feature matrix set A' and the size of the key space |K|' to hide the secret information in the carrier: (4a) The sender uses the secret information X=(x ₁ ,x ₂ ,...x _j ,...,x _m ) to select the matrix A′ _j corresponding to the secret information x _j from the feature matrix set A′, Where x _j represents the jth element in the secret information X, j=1,...,m; (4b) The sender uses A' _j and key k to select the carrier corresponding to the column where the element of A' _j (k, i)≠0 is located As a carrier for hiding secret information x _j , where A′ _j (k, i) represents the element of row k and column i in matrix A′ _j , i=1,...,n, j=1,... ,m, carrier is an element in the set of _candidate hidden vectors CS, (4c) The sender calculates the selected bearer The variance of defines the embedding depth α _j of secret information, j=1,...,m; (4d) The sender uses the embedding depth α _j to embed the secret information into the corresponding carrier data, mainly adopting the additive data modification method to embed the secret information, Where s _j represents the carrier after hiding the secret information x _j ; (5) The sender sends the obtained carrier set S={s ₁ ,s ₂ ,...,s _j ,...,s _m } after embedding the secret information to the receiver; (6) The receiver obtains the secret information: (6a) The receiver uses filtering to estimate the secret information X' in the carrier set S after embedding the secret information; (6b) The receiver uses the final feature matrix set A' according to the estimated secret information X' to obtain the correspondence between the estimated secret information X' and the hidden carrier; (6c) According to the corresponding relationship between the acquired estimated secret information X' and the hidden carrier, the receiver reverses the estimated secret information X' to obtain the real secret information X. 2. The method for hiding secret information using images according to claim 1, characterized in that, in the step (1), the carrier image is divided into blocks by first dividing the carrier image into small blocks of 8×8, and calculating The fuzzy value of each small block; then compare the fuzzy value with a given threshold T: if the fuzzy value is greater than or equal to the threshold _T , then add the small block corresponding to the fuzzy value to the candidate hidden carrier set CS; otherwise , discarded until all image blocks are calculated. 3. the method for carrying out block segmentation to carrier image according to claim 2, is characterized in that, calculates the fuzzy value of each small block, carries out as follows: First, define the fuzzy set membership degree as follows: $\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; ,</span>$ Among them, d(p,q) is the pixel value at the image (p,q), d′ is the normalized value of the pixel at the image (p,q), p,q=1,...,8,N It is a constant, and its value should ensure that the value range of u(p,q) is between 0.5 and 1, that is, 0.5≤u(d(p,q))≤1; Secondly, the fuzzy value of the pixel at the position (p, q) of the image is defined on the basis of the membership degree of the fuzzy set as follows: H(p,q)＝-(u(p,q))log ₂ (u(p,q))-(1-u(p,q))log ₂ (1-u(p,q)), Finally, define the blur value of the small block according to the blur value H(p,q) of the pixel at the position (p,q) of the image as follows: $\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; q</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; a</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; q</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$ Wherein, p ₀ represents the abscissa of the small block center point, q ₀ represents the ordinate of the small block center point, a represents the length of the small block, a=8, k is the variable that controls the movement of the small block center point abscissa, and l is The variable that controls the vertical coordinate movement of the center point of the small block, $k=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},l=\frac{-(a-1)}{2},...,\frac{(a-1)}{2},k\&NotEqual;l.$ 4. The method for hiding secret information using images according to claim 1, characterized in that in the step 4c), the embedding depth α _j of the carrier is calculated as follows: First calculate the mean value of the carrier data, the formula is as follows: $\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\underset{{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\underset{{\mathrm{<span\; class="notranslate">\; l</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; l</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$ Among them, Δ represents the control variable of the moving step, k' is the variable that controls the movement of the abscissa of the carrier data, l' is the variable that controls the movement of the ordinate of the carrier data, k'=1,...,(2Δ+1), l'=1,...,(2Δ+1),; Second, using the mean Define the variance of the carrier data as follows: $\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\sqrt{\underset{\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{\overset{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\underset{\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{\overset{\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ,</span>$ Wherein, u represents the abscissa of the carrier center point, v represents the ordinate of the carrier center point, e is the variable that controls the movement of the abscissa of the carrier center point, and f is the variable that controls the movement of the ordinate of the carrier data, e=(u-Δ) ,....,(u+Δ), f=(v-Δ),...,(v+Δ); Finally, according to the variance σ(u,v) of the carrier data, the embedding depth α _j of the carrier image block is defined as follows: α _j =λσ(u,v). Among them, λ represents the stretching coefficient.