User-oriented hierarchical management image steganography method, device, equipment and medium

Abstract

The application provides an image steganography method and device for user hierarchical management, equipment and medium. The DWT coefficients of a plaintext image are compressed and blocked by using a compression sensing technology, and then a threshold secret image sharing method is used to dynamically set differentiated threshold values for each image block. The image blocks are steganographed into a carrier image by combining a double-image embedding method, so that a ciphertext image with visual significance is finally generated. Different threshold values are set for different image blocks, so that hierarchical steganography and extraction of the image are realized. When the number of users participating in decryption is less than a set threshold value, no effective information related to the plaintext image can be obtained. When the number of users participating in decryption is greater than or equal to the threshold value, information related to the plaintext image can be obtained, and with the increase of the number of participating users, the reconstruction quality of the plaintext image is improved.

Description

Technical Field

[0001] This invention belongs to the field of image privacy and security technology, and in particular to an image steganography method, apparatus, device and medium for user-level hierarchical management. Background Technology

[0002] Images, with their advantages of large information capacity, intuitive expression, and diverse forms, have become a key medium in the fields of visual information communication and human-computer interaction. To ensure the security of sensitive information during image transmission and storage, image steganography has emerged as an important privacy protection method. This technology aims to covertly embed secret information (such as text, keys, or other data) into publicly available carrier images, achieving covert communication and secure storage of secret information while maintaining the basic visual quality (i.e., visual security) of the carrier image.

[0003] In existing technologies, a mainstream image steganography strategy is "encryption before embedding." This strategy first uses cryptographic algorithms (such as symmetric or asymmetric encryption) to convert the original secret information into meaningless ciphertext data. Then, specific steganography algorithms (such as least significant bit (LSB) substitution, transform domain embedding, etc.) are used to hide the generated ciphertext data in the redundant space (such as pixel values, transform coefficients) of the carrier image, ultimately generating a visually secure ciphertext image. The core objective of this method is to ensure that unauthorized third parties cannot perceive or extract valid secret information from the carrier image.

[0004] However, existing image steganography techniques based on "encryption before embedding" generally have a significant limitation: their output exhibits an "all or nothing" access characteristic. Specifically:

[0005] Single access permission: The generated ciphertext image can only be fully decrypted and the complete original secret information can be extracted by a single user (or user group) holding the complete decryption key. It lacks the ability to allocate a subset of secret information or information with different access levels as needed.

[0006] Poor scenario adaptability: When facing the actual needs of multi-user collaboration (such as different users needing to obtain different parts of secret information) or differentiated application scenarios (such as dynamic changes in user permissions and information needing to be decrypted hierarchically), existing technologies appear extremely rigid. To meet the differentiated needs of different users or scenarios, it is often necessary to generate and distribute different encrypted image copies for each need, resulting in low efficiency and a significant increase in storage and transmission costs.

[0007] Lack of flexibility: Existing technologies struggle to dynamically adjust the embedded secret information content or its access permissions. Once the ciphertext image is generated, the secret information it contains and its access control policy are fixed. Any change in the number of users, permission levels, or required information segments inevitably requires re-executing the entire "encryption-embedding" process and distributing a new ciphertext image.

[0008] Therefore, while existing "encryption before embedding" image steganography techniques can provide basic confidentiality, their inherent "all or nothing" nature severely limits their practicality and scalability in complex scenarios involving multiple users, multiple permissions, and dynamic changes. Thus, there is an urgent need for a flexible, efficient image steganography method that can be dynamically adjusted according to the number of users. Summary of the Invention

[0009] In order to address the aforementioned problems in the existing technology and to meet the application needs of multiple users, multiple scenarios, and differentiating characteristics in complex environments, this invention provides an image steganography method, apparatus, device, and medium for user-level hierarchical management.

[0010] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0011] This invention provides an image steganography method for user-level hierarchical management, comprising:

[0012] Perform discrete wavelet transform on the plaintext image to generate a transform domain coefficient matrix, which is then used as the first matrix.

[0013] The first matrix is ​​zigzag scanned to generate a one-dimensional vector, which is then converted into a second matrix of the same size as the plaintext image. An initial key value is set, and a measurement matrix is ​​generated. The second matrix is ​​then subjected to compressed sensing measurement using the measurement matrix to generate a third matrix. The third matrix is ​​quantized to obtain a fourth matrix. The fourth matrix and the measurement matrix are divided into blocks according to the total number of participating users, with each participating user corresponding to a measurement matrix sub-block. A diffusion operation is performed on each image block obtained from the fourth matrix to obtain the corresponding secret image block.

[0014] The threshold value of each secret image block is dynamically set according to the total number of participating users. A polynomial is constructed for the corresponding secret image block according to the threshold value, and multiple shadow images corresponding to the secret image block are generated, which are the same as the total number of participating users. Each participating user corresponds to one shadow image of the secret image block.

[0015] For each participating user, the shadow images of all the image blocks corresponding to the participating user will be combined to form the shadow share held by the participating user;

[0016] Each participating user's shadow share is embedded into the carrier image to generate a ciphertext image, thus enabling the steganography and transmission of the shadow share.

[0017] Furthermore, it also includes the decryption process of the encrypted image, including:

[0018] Determine the number of participating users in the current decryption, obtain the set of shadow shares held by each participating user in the current decryption, and reconstruct the shadow image corresponding to each participating user using Lagrange interpolation. Then, perform dediffusion operation to recover the corresponding image blocks and stitch the image blocks together to obtain the current reconstruction result.

[0019] Based on the current reconstruction results, an inverse quantization operation is performed. The SL0 algorithm is used to reconstruct the result after the inverse quantization operation. An inverse zigzag scan is performed on the reconstructed output matrix, and the image spatial domain data is recovered through the inverse wavelet transform. Finally, the decrypted image corresponding to the number of participating users is obtained.

[0020] On the other hand, an image steganography device for user-level hierarchical management is provided, including an encryption module for encrypting plaintext images, wherein the encryption module includes:

[0021] The wavelet transform module is used to perform discrete wavelet transform on plaintext images to generate a transform domain coefficient matrix, which serves as the first matrix.

[0022] The zigzag scanning module performs a zigzag scan on the first matrix to generate a one-dimensional vector, and then converts it into a second matrix of the same size as the plaintext image. The compressed sensing measurement module sets an initial key value, generates a measurement matrix, and performs compressed sensing measurement on the second matrix using the measurement matrix to generate a third matrix. The quantization processing module quantizes the third matrix to obtain a fourth matrix. The diffusion operation module divides the fourth matrix and the measurement matrix into blocks according to the total number of participating users, with each participating user corresponding to a measurement matrix sub-block. It performs a diffusion operation on each image block obtained from the fourth matrix to obtain the corresponding secret image block.

[0023] The shadow image generation module is used to dynamically set the threshold value of each secret image block according to the total number of participating users, construct a polynomial for the corresponding secret image block according to the threshold value, and generate multiple shadow images corresponding to the secret image block with the same number as the total number of participating users. Each participating user corresponds to one shadow image of the secret image block.

[0024] The user shadow share determination module is used to combine the shadow images of all image blocks corresponding to each user to form the shadow share held by the user.

[0025] The ciphertext image generation module is used to embed the shadow share held by each participating user into the carrier image to generate a ciphertext image, thereby realizing the steganography and transmission of the shadow share.

[0026] Furthermore, it also includes a decryption module for decrypting the ciphertext image generated by the encryption module, wherein the decryption module includes:

[0027] The recovery module determines the number of participating users in the current decryption, obtains the set of shadow shares held by each participating user in the current decryption, performs Lagrange interpolation reconstruction on the shadow image corresponding to each participating user, performs dediffusion operation to recover the corresponding image blocks, and stitches and combines the image blocks to obtain the current reconstruction result.

[0028] The image decryption module performs inverse quantization based on the current reconstruction result, reconstructs the result after inverse quantization using the SL0 algorithm, performs inverse zigzag scanning on the reconstructed output matrix, and recovers the image spatial domain data through inverse wavelet transform, finally obtaining the decrypted image corresponding to the number of participating users.

[0029] On the other hand, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described user-oriented hierarchical image steganography method.

[0030] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described user-oriented hierarchical image steganography method.

[0031] On the other hand, the present invention provides a computer program product stored on a computer-readable storage medium and including computer instructions that, when executed by a processor, cause a computer device to implement the steps of the above-described user-level hierarchical image steganography method.

[0032] Compared with the prior art, the technical effects of the present invention are as follows:

[0033] This invention provides an image steganography method for user-level hierarchical management. This method utilizes compressed sensing technology to compress the DWT coefficients of plaintext images. The images are then segmented and then utilized (…). k , n The threshold-based secret image sharing method dynamically sets differentiated threshold values ​​for each image block. The steganography is embedded into the carrier image using a dual-image embedding method, ultimately generating a visually meaningful ciphertext image. Hierarchical steganography and extraction are achieved by setting different threshold values ​​for different image blocks. The number of users participating in decryption... Less than At that time, no valid information related to the plaintext image could be obtained; when the number of users participating in the decryption met the requirement... At that time, relevant information about the plaintext image can be obtained, and the reconstruction quality of the plaintext image improves as the number of participating users increases. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0035] Figure 1 This is a flowchart of an image steganography method for user-level hierarchical management in one embodiment;

[0036] Figure 2 This is a flowchart of image encryption and steganography for user-level hierarchical management in one embodiment;

[0037] Figure 3 This is a flowchart of image extraction and decryption process for user-level hierarchical management in one embodiment;

[0038] Figure 4 This is an image steganography effect diagram for user-level hierarchical management in one embodiment, wherein... Figure 4 (a) in the image is a plaintext image. Figure 4 (b) in the image is the carrier image. Figure 4 (c) in the image is the encrypted image. Figure 4 Image (d) shows the decrypted image obtained when one user participates. Figure 4 Image (e) shows the decrypted image obtained when two users participated. Figure 4 (f) shows the decrypted image obtained when three participating users were present. Figure 4 (g) shows the decrypted image obtained when four participating users are present.

[0039] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0041] Reference Figure 1 One embodiment provides an image steganography method for user-level hierarchical management, comprising:

[0042] Step S1: Perform discrete wavelet transform on the plaintext image to generate a transform domain coefficient matrix, which is used as the first matrix.

[0043] Step S2: Perform a zigzag scan on the first matrix to generate a one-dimensional vector, and then convert it into a second matrix of the same size as the plaintext image;

[0044] Step S3: Set the initial key value, generate a measurement matrix, perform compressed sensing measurement on the second matrix using the measurement matrix, and generate the third matrix;

[0045] Step S4: Quantize the third matrix to obtain the fourth matrix;

[0046] Step S5: Divide the fourth matrix into blocks, and perform a diffusion operation on each image block obtained by the block division to obtain the corresponding secret image block;

[0047] Step S6: Dynamically set the threshold value for each secret image block according to the total number of participating users, construct a polynomial for the corresponding secret image block according to the threshold value, and generate multiple shadow images corresponding to the secret image block with the same number as the total number of participating users. Each participating user corresponds to one shadow image of the secret image block.

[0048] Step S7: For each participating user, combine the shadow images of all image blocks corresponding to the participating user to form the shadow share held by the participating user.

[0049] Step S8: Embed the shadow share held by each participating user into the carrier image to generate a ciphertext image, thereby realizing the steganography and transmission of the shadow share.

[0050] In steps S1 and S2 of the above embodiments, a Discrete Wavelet Transform (DWT) is performed on the plaintext image P to generate a corresponding transform domain coefficient matrix. Subsequently, the transform domain coefficient matrix is ​​arranged into a one-dimensional vector using a zigzag scan and converted into a matrix form of the same size as the plaintext image to preserve the spatial structure of the image for subsequent processing. Specifically, a preferred embodiment provides specific preferred implementation steps for steps S1 and S2, including:

[0051] Step S1, using the discrete wavelet transform matrix Ψ The size is N×N plaintext images PPerforming a discrete wavelet transform yields the corresponding transform domain coefficient matrix, i.e., the first matrix. P 1:

[0052]

[0053] Step S2, use zigzag scanning to scan the first matrix P 1 converted into a one-dimensional vector Then the vector Convert to N ×N The matrix and setting a threshold TS To make the absolute value of the matrix less than TS Setting the first zero to zero yields the second matrix.

[0054] In another embodiment, a preferred implementation of step S3, which involves setting an initial key value, generating a measurement matrix, and performing compressed sensing measurements on the second matrix using the measurement matrix to generate a third matrix, includes:

[0055] Step S3.1, settings 、 initial value and initial value and This is the initial key value, which, through an iterative improved Henon mapping, generates a floating-point form of chaotic sequence. The equation for the improved Henon mapping is:

[0056] ;

[0057] in , , , , , They represent the first The next iteration, the... In the next iteration The value of , , They represent the first The next iteration, the... In the next iteration y The value of . To obtain a sequence with better chaotic properties, it is recommended to skip the initial transition state. .

[0058] Step S3.2, convert the chaotic sequence into floating-point form Convert to integer chaotic sequence The conversion formula is as follows:

[0059] ;

[0060] Where mod represents the modulo operation. This indicates rounding down. The length of the desired integer chaotic sequence;

[0061] Step S3.3, convert the integer chaotic sequence X ( s Remove duplicate elements from the given information to obtain the result. ;

[0062] Step S3.4, set It did not appear in the middle The elements in the array are arranged in ascending order. X ( s The end of ') until a length of is obtained. L And a non-repeating password stream XS ;

[0063] Step S3.5, using a cipher stream XS The corresponding row vector in the Hadamard matrix is ​​selected as the index, and then a partial Hadamard matrix is ​​obtained by recombination, which is then used as the measurement matrix. .

[0064] Step S3.6, measure the second matrix using the measurement matrix. P 2. Perform compressed sensing measurements and generate a size of M×N The third matrix P 3, of which ， The compression ratio of the entire image. .

[0065] In any of the above embodiments of the image steganography method for user-level hierarchical management, step S4 involves processing the third matrix. P 3. Quantization processing is performed to obtain the fourth matrix. P 4. The formula that can be used for quantification is:

[0066] ;

[0067] Where max and min Representing the third matrix respectively P The maximum and minimum values ​​among all elements in 3. round This indicates taking the nearest integer value. Representing the third matrix P The z-th element of 3 Represents the fourth matrix P The z-th element of 4.

[0068] In another embodiment, step S5, which involves dividing the fourth matrix into blocks and performing a diffusion operation on each resulting image block to obtain the corresponding secret image block, is provided. A preferred implementation of this step includes: based on the total number of users participating during encryption. n , the fourth matrix P 4 and measurement matrix Divided by row n Non-overlapping image blocks , Each image block The size is Measurement matrix Divided by row n Measurement matrix sub-blocks ,in n The total number of participating users during encryption is represented by the measurement matrix sub-block, which corresponds to one participating user. During the recovery phase, the system selects the corresponding measurement matrix sub-block based on the participating user's index number. All image blocks are traversed. And perform the following operations respectively:

[0069] Based on the set initial key value, a modified Henon mapping is used to generate a length of (). M / n ) ×N×2 Non-repeating integer cryptographic stream sequence KS , non-repeating integer cryptographic stream KS Divided into two parts, each with a length of ( M / n ) ×N Non-repeating integer cryptographic stream sequence KS 1 and a sequence of unique integer cryptographic streams KS 2. Transform the sequence of unique integer cryptographic streams. KS 1 and a sequence of unique integer cryptographic streams KS 2 are applied to the forward and reverse diffusion processes of diffusion operations, respectively, for the first... i Image blocks B i Performing a diffusion operation yields the corresponding secret image block. C i The formula for the diffusion operation is:

[0070] ;

[0071] in, It is an intermediate variable in the diffusion process, representing the pixel value at the corresponding position after forward diffusion. express The previous element, This represents the index sequence corresponding to the forward diffusion process; Indicates the first iThe z-th element of the image patch is the input to the diffusion process; mod256 represents modulo operation, which limits the calculation result to the range of 0 to 255. This represents the pixel value at the corresponding position during backward diffusion, and is the output of the diffusion process; express The next element; This represents the index sequence corresponding to the backward diffusion process. The equation for the improved Henon map has been introduced in the previous embodiments and will not be repeated here.

[0072] In step S6 of the aforementioned embodiment, based on the total number of participating users n Dynamically set each secret image block Threshold value Threshold values ​​for each image block Customizable settings to ensure That's all.

[0073] According to secret image blocks Threshold value For secret image blocks Construct polynomials to generate secret image patches corresponding A shadow image The number of shadow images is the same as the total number of participating users. , No. Each participating user corresponds to a secret image block. The A shadow image , The participating user ID has a value range of [value range missing]. The polynomial is constructed as follows:

[0074] ;

[0075] in, Assign secret image blocks to the first The shadow image value of each participating user, i.e. ; User IDs for participating users; For polynomial coefficients, For secret image blocks pixel values, The coefficients are randomly generated polynomials. For secret image blocks The threshold value indicates the minimum requirement. Only a shadow image can recover the secret image block. ; It is a large prime number with a value of 251.

[0076] Step S7, for the first The participating user, the first of all secret image blocks The combination of shadow images forms the first... Shadow shares held by each participating user ,Right now:

[0077]

[0078] Step S8, transfer the shadow shares held by each participating user. Embedded into a carrier image to generate a ciphertext image Q This enables the steganography and transmission of shadow shares. Specifically, it includes:

[0079] Step S8.1, carrier image preprocessing: scale the pixel values ​​of the carrier image to prevent overflow;

[0080] Step S8.2, Index Sequence Generation: Generate an index sequence based on the initial key value using an improved Henon mapping. T ;

[0081] Step S8.3, Construction M matrix W m The preprocessed image on the carrier is segmented into... M matrix U m , M =2 n ;

[0082] Step S8.4, Shadow Share Embedding: Divide participating users into two groups, and the shadow shares held by one group of participating users correspond to the embedding matrix. W m In the embedding process, an index sequence is used. T Select matrix W m The element positions will embed the shadow shares held by participating users into the matrix. W m The corresponding bits at the corresponding element positions in the matrix are used to combine the embedded matrices to obtain the matrix. W The other group of participating users' shadow shares correspond to the embedded matrix. U m In the embedding process, an index sequence is used. T Select matrix U m The element positions will embed the shadow shares held by participating users into the matrix. U m The corresponding bits at the corresponding element positions in the matrix are used to combine the embedded matrices to obtain the matrix. U ;

[0083] Step S8.5, convert the matrix W As alpha channel and matrix U Merge to generate encrypted image Q .

[0084] In one embodiment, with n Taking a user count of 4 as an example, the embedding process includes the following steps:

[0085] Step S8.1: Scale the pixel values ​​of the carrier image to prevent data overflow that may occur during the embedding process, and obtain the scaled pixel value matrix of the carrier image.

[0086] Step S8.2, use the improved Henon mapping to generate a length of... N 2 / 8 Non-repeating integer cryptographic stream KS 3, denoted as index sequence T ;

[0087] Step S8.3, Construction M Each element has a value of 255, and its size is (1 / 8 × ... N )× N rectangular matrix The pixel value matrix of the scaled carrier image is then segmented to obtain... M The size is (1 / 8 × N )× N matrix , m =1,2,......,8;

[0088] Step S8.4: The first participating user and the second participating user are grouped together, and the shadow share held by the first participating user is determined using the least significant bit embedding algorithm. S Shadow shares held by the first and second participating users S 2. Randomly embedded into the matrix W m :

[0089] , , , ,

[0090] , , , .

[0091] The third and fourth participating users are grouped together, and the shadow share held by the third participating user is determined using the least significant bit embedding algorithm.S Shadow shares held by the 3rd and 4th participating users S 4. Randomly embedded into the matrix :

[0092] , , , ,

[0093] , , , ;

[0094] In the formula, This represents the shadow shares held by the first participating user. S The first in 1 n The first and second bits of each element, This represents the shadow shares held by the first participating user. S The first in 1 n The 3rd and 4th bits of each element, This represents the shadow shares held by the first participating user. S The first in 1 n The 5th and 6th bits of each element, This represents the shadow shares held by the first participating user. S The first in 1 n The 7th and 8th bits of each element, This indicates the shadow shares held by the second participating user. S The second one n The first and second bits of each element, This indicates the shadow shares held by the second participating user. S The second one n The 3rd and 4th bits of each element, This indicates the shadow shares held by the second participating user. S The second one n The 5th and 6th bits of each element, This indicates the shadow shares held by the second participating user. S The second one n The 7th and 8th bits of each element, This indicates the shadow shares held by the third participating user. S The third n The first and second bits of each element, This indicates the shadow shares held by the third participating user. S The third n The 3rd and 4th bits of each element, This indicates the shadow shares held by the third participating user. S The third n The 5th and 6th bits of each element, This indicates the shadow shares held by the third participating user. S The third n The 7th and 8th bits of each element, This indicates the shadow shares held by the fourth participating user. S The fourth one n The first and second bits of each element, This indicates the shadow shares held by the fourth participating user. S The fourth one n The 3rd and 4th bits of each element, This indicates the shadow shares held by the fourth participating user. S The fourth one n The 5th and 6th bits of each element, This indicates the shadow shares held by the fourth participating user. S The fourth one n The 7th and 8th bits of each element, Representation matrix W The first in 1 T ( n The first and second bits of the ) elements, T ( n ) represents an index sequence T The first in n One element, Representation matrix W The second one T ( n The 3rd and 4th bits of the element, Representation matrix W The third T ( n The 5th and 6th bits of the element, Representation matrix W The fourth one T ( n The 7th and 8th bits of the element, Representation matrix W The fifth one T ( n The first and second bits of the ) elements, Representation matrix W The 6th T ( n The 3rd and 4th bits of the element, Representation matrix W The 7th T (n The 5th and 6th bits of the element, Representation matrix W The 8th T ( n The 7th and 8th bits of the ) elements; Representation matrix U The first in 1 T ( n The first and second bits of the ) elements, Representation matrix U The second one T ( n The 3rd and 4th bits of the element, Representation matrix U The third T ( n The 5th and 6th bits of the element, Representation matrix U The fourth one T ( n The 7th and 8th bits of the element, Representation matrix U The fifth one T ( n The first and second bits of the ) elements, Representation matrix U The 6th T ( n The 3rd and 4th bits of the element, Representation matrix U The 7th T ( n The 5th and 6th bits of the element, Representation matrix U The 8th T ( n The 7th and 8th bits of the elements.

[0095] Step S8.5, will , , , , , , , The matrix obtained after combination W ;Will , , , , , , , The matrix obtained after combinationU， matrix W As alpha channel and matrix U Together, they generate ciphertext images with visual significance. .

[0096] In another embodiment, a method for decrypting the ciphertext image generated in the above embodiments is provided. Specifically, the decryption process of the ciphertext image includes:

[0097] Determine the number of participating users in the current decryption, obtain the set of shadow shares held by each participating user in the current decryption, and reconstruct the shadow image corresponding to each participating user using Lagrange interpolation. Then, perform dediffusion operation to recover the corresponding image blocks and stitch the image blocks together to obtain the current reconstruction result.

[0098] Based on the current reconstruction results, after inverse quantization, the SL0 algorithm is used for reconstruction. The inverse zigzag scan is performed on the reconstructed output matrix, and the image spatial domain data is recovered through the inverse wavelet transform. Finally, the decrypted image corresponding to the number of participating users is obtained.

[0099] Specifically, in one embodiment, the decryption process of the ciphertext image generated in the above embodiment is completed using the following steps:

[0100] (1) Based on the initial key value, iteratively improve the Henon mapping to generate the measurement matrix. A diffuse sequence is a stream of non-repeating integer cryptographic sequences. KS (used for dediffusion operations) and index sequences T Using index sequences T And least significant bit extraction algorithm, from ciphertext image Q The shadow share is extracted from the pixel value channel and alpha channel. The initial key value serves as the initial input parameter for the improved Henon mapping. Iteratively, it generates a subsequent cipherstream sequence for the diffusion operation and for generating the measurement matrix (steps S3.1 to S3.5). The specific method has been described in detail above and will not be repeated here. The measurement matrix Φ is divided into rows... n Measurement matrix sub-block Φ i ,in n The total number of participating users during encryption is represented by the measurement matrix sub-block, which corresponds to one participating user. During the decryption phase, the corresponding measurement matrix sub-block is selected based on the participating user's index number.

[0101] (2) Determine the number of participating users in the current decryption and collect the shadow image corresponding to each participating user. and the shadow image Lagrange interpolation reconstruction is performed to recover pixel values ​​using the Lagrange interpolation algorithm. The recovered pixel values ​​are then arranged to generate secret image patches. The formula for recovering pixel values ​​using the Lagrange interpolation algorithm is as follows:

[0102] ;

[0103] In the formula This represents the number of users currently participating in the decryption process. For the first j The pixel values ​​of the shadow image corresponding to each participating user.

[0104] The polynomial is used for reconstruction, and the polynomial coefficients are the recovered pixel values. The calculation process is performed in the finite field GF(251). That is, the shadow image is provided by the currently participating user in the decryption process. Then, the polynomial coefficients can be restored using Lagrange interpolation, thereby recovering the original polynomial. , For secret image blocks The pixel value.

[0105] (3) Recover all the secret image blocks C i The images are pieced together according to their original positions to form a complete encrypted image. E i .

[0106] (4) Based on the diffusion sequence, i.e., the non-repeating integer cryptographic stream KS For encrypted images Perform dediffusion to recover the image patch. The formula used for the dediffusion operation is:

[0107] ;

[0108] in Indicates the first The z-th element of the image patch is the input to the de-diffusion process. This represents the output of the de-diffusion process; This is an intermediate variable representing the pixel value at the corresponding position after backward de-divergence; for The previous element; for The next element; This indicates the index corresponding to the forward diffusion; mod256 indicates modulo operation, which limits the calculation result to the range of 0~255. This indicates the index corresponding to the backward diffusion.

[0109] (5) Recover the image patch data splicing and assembling M×N matrix Y， Obtain the current reconstruction results;

[0110] (6) For the matrix Y Perform inverse quantization to obtain the matrix Y1 The formula used for the inverse quantization operation is:

[0111] ;

[0112] in, max and min It was obtained via a secret channel before being decrypted.

[0113] (7) Select the corresponding measurement matrix sub-block according to the index number of the currently decrypted participating user. The SL0 reconstruction algorithm is used to reconstruct the matrix. Y 1. Reconstruct the matrix and output the matrix. Y 2:

[0114]

[0115] (8) For the reconstructed output matrix Y 2. Perform an inverse zigzag scan and recover the image spatial domain data through the inverse wavelet transform to finally obtain the decrypted image corresponding to the number of participating users. Y 3.

[0116] This invention encrypts the image to be protected by following a specific encryption process, sets a threshold for each image block based on the number of users participating in decryption, and is based on ( k , n The threshold secret image sharing method divides each secret image block into... n The image is divided into shadow images and distributed. A security protection system for multi-level user collaborative decryption is constructed, effectively realizing hierarchical steganography and progressive recovery based on the number of users. To demonstrate the effectiveness of the user-level hierarchical image steganography method provided by this invention, an example uses a plaintext image of size 512×512, with an initial key value of... x 0 = 0.2, y 0 = 0.3 CR = 0.25; the number of blocks and the total number of users participating in decryption are both 0.25. n = 4. Among them, image patches B 1 、B 2 、B 3 、B The thresholds corresponding to 4 are respectively K 1=3、 K 2=4、 K 3=4、K 4=4. Reference Figure 2 , Figure 3 ,in Figure 2 A flowchart for user-level hierarchical image encryption and steganography management. Figure 3 A flowchart for image extraction and decryption in a user-centric hierarchical management system. Figure 4 This is an image protection effect diagram for user-level hierarchical management, in which... Figure 4 (a) in the image is a plaintext image. Figure 4 (b) in the image is the carrier image. Figure 4 (c) in the figure represents the ciphertext image. The plaintext image was successfully embedded into the carrier image. There was no significant visual difference between the encrypted ciphertext image and the carrier image, and the peak signal-to-noise ratio (PSNR) between the two was 35.3733 dB. Figure 4 Image (d) shows the decrypted image obtained when one user participates. Figure 4 Image (e) shows the decrypted image obtained when two users participated. Figure 4 (f) shows the decrypted image obtained when three participating users were present. Figure 4 Figure (g) shows the decrypted images obtained when there are 4 participating users, demonstrating the differentiated decryption effects obtained with different numbers of participants. When there are only 1 or 2 participants (less than the minimum threshold)... K With 1=3 participants, the PSNR values ​​of the decrypted images were 5.7635 dB and 5.8846 dB, respectively, and no useful information about the plaintext images could be obtained. As the number of participants increased, the details and clarity of the recovered images became higher, until the extraction and optimal reconstruction of the plaintext images were achieved with 4 participants, and the PSNR of the decrypted images increased to 31.9565 dB.

[0117] On the other hand, another embodiment provides an image steganography device for user-level hierarchical management, including an encryption module for encrypting plaintext images, wherein the encryption module includes:

[0118] The wavelet transform module is used to perform discrete wavelet transform on plaintext images to generate a transform domain coefficient matrix, which serves as the first matrix.

[0119] The zigzag scanning module is used to perform a zigzag scan on the first matrix to generate a one-dimensional vector, and then convert it into a second matrix of the same size as the plaintext image.

[0120] The compressed sensing measurement module is used to set the initial key value, generate a measurement matrix, perform compressed sensing measurement on the second matrix using the measurement matrix, and generate a third matrix.

[0121] The quantization module is used to quantize the third matrix to obtain the fourth matrix;

[0122] The diffusion operation module is used to divide the fourth matrix and the measurement matrix into blocks according to the total number of participating users. Each participating user corresponds to a measurement matrix sub-block. The diffusion operation is performed on each image block obtained by dividing the fourth matrix to obtain the corresponding secret image block.

[0123] The shadow image generation module is used to dynamically set the threshold value of each secret image block according to the total number of participating users, construct a polynomial for the corresponding secret image block according to the threshold value, and generate multiple shadow images corresponding to the secret image block with the same number as the total number of participating users. Each participating user corresponds to one shadow image of the secret image block.

[0124] The user shadow share determination module is used to combine the shadow images of all image blocks corresponding to each user to form the shadow share held by the user.

[0125] The ciphertext image generation module is used to embed the shadow share held by each participating user into the carrier image to generate a ciphertext image, thereby realizing the steganography and transmission of the shadow share.

[0126] Furthermore, it also includes a decryption module for decrypting the encrypted image, wherein the decryption module includes:

[0127] The recovery module is used to determine the number of participating users in the current decryption, obtain the set of shadow shares held by each participating user in the current decryption, and reconstruct the shadow image corresponding to each participating user using Lagrange interpolation. Then, it performs dediffusion operation to recover the corresponding image blocks and stitches the image blocks together to obtain the current reconstruction result.

[0128] The image decryption module is used to perform inverse quantization based on the current reconstruction result, reconstruct the result after inverse quantization using the SL0 algorithm, perform inverse zigzag scanning on the reconstructed output matrix, and recover the image spatial domain data through inverse wavelet transform, finally obtaining the decrypted image corresponding to the number of participating users.

[0129] On the other hand, the present invention provides a computer device including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the user-oriented hierarchical image steganography method provided in any of the above embodiments. The computer device may be a server. The computer device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device stores sample data. The network interface of the computer device is used for communication with external terminals via a network connection.

[0130] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the image steganography method for user-oriented hierarchical management provided in any of the above embodiments.

[0131] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0132] Matters not covered in this invention are common knowledge.

[0133] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0134] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application.

[0135] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An image steganography method for user-level hierarchical management, characterized in that, include: Perform discrete wavelet transform on the plaintext image to generate a transform domain coefficient matrix, which is then used as the first matrix. The first matrix is ​​zigzag scanned to generate a one-dimensional vector, and then converted into a second matrix of the same size as the plaintext image. Set the initial key value, generate a measurement matrix, perform compressed sensing measurement on the second matrix using the measurement matrix, and generate the third matrix; The third matrix is ​​quantized to obtain the fourth matrix; The fourth matrix and the measurement matrix are divided into blocks according to the total number of participating users. Each participating user corresponds to a measurement matrix sub-block. A diffusion operation is performed on each image block obtained by dividing the fourth matrix to obtain the corresponding secret image block. Based on the total number of participating users n Dynamically set the differential threshold value for each secret image block k i Threshold values ​​for each secret image block k i Freely set, where 2≤ k i ≤ n Based on the threshold value, a polynomial is constructed for the corresponding secret image block to generate multiple shadow images corresponding to the secret image block, the same number as the total number of participating users. Each participating user corresponds to one shadow image of the secret image block. The polynomial is constructed as follows: in, Assign secret image blocks to the first j The shadow image value of each participating user; j For participating user IDs, ; For polynomial coefficients, For secret image blocks C i pixel values, The coefficients are randomly generated polynomials. For secret image blocks C i The threshold value indicates the minimum requirement. Only a shadow image can recover the secret image block. C i ; p It is a large prime number with a value of 251; For each participating user, the shadow images of all the image blocks corresponding to the participating user will be combined to form the shadow share held by the participating user; Each participating user's shadow share is embedded into the carrier image to generate a ciphertext image, thus enabling the steganography and transmission of the shadow share. The number of users participating in the decryption process when decrypting the encrypted image. t Less than min( k i When the number of users participating in the decryption is sufficient, no valid information related to the plaintext image can be obtained; when the number of users participating in the decryption is sufficient... At that time, relevant information about the plaintext image can be obtained, and the reconstruction quality of the plaintext image improves as the number of participating users increases.

2. The image steganography method for user-oriented hierarchical management according to claim 1, characterized in that, The measurement matrix is ​​obtained through the following steps: set up 、 initial value and initial value and This is the initial key value, which, through an iterative improved Henon mapping, generates a floating-point form of chaotic sequence. The equation for the improved Henon mapping is: in , , , , , They represent the first The next iteration, the... In the next iteration The value of , , They represent the first The next iteration, the... In the next iteration y The value of ; Chaotic sequences in floating-point form Convert to integer chaotic sequence The conversion formula is as follows: Where mod represents the modulo operation. floor Indicates rounding down. The length of the desired integer chaotic sequence; Integer chaotic sequence X ( s Remove duplicate elements from the given information to obtain the result. X ( s '); set It did not appear in the middle X ( s' The elements in the array are arranged in ascending order. X ( s The end of ') until a length of is obtained. L And a non-repeating password stream XS ; With cryptographic stream XS The corresponding row vector in the Hadamard matrix is ​​selected as the index, and then a partial Hadamard matrix is ​​obtained by recombination, which is then used as the measurement matrix. .

3. The image steganography method for user-oriented hierarchical management according to claim 2, characterized in that, The third matrix is ​​quantized to obtain the fourth matrix, where the quantization formula is: in, max and min Let represent the maximum and minimum values ​​of the elements in the third matrix, respectively. round This indicates taking the nearest integer value. Representing the third matrix P The z-th element of 3 Represents the fourth matrix P The z-th element of 4.

4. The image steganography method for user-oriented hierarchical management according to claim 3, characterized in that, Embedding the shadow share held by each participating user into the carrier image generates a ciphertext image, including: Carrier image preprocessing: scaling the pixel values ​​of the carrier image; Index sequence generation: Generates an index sequence using an improved Henon mapping based on the set initial key value. T ; structure M matrix W m The preprocessed image on the carrier is segmented into... M matrix U m , M =2 n ; Shadow share embedding: Participating users are divided into two groups, and the shadow shares held by one group of participating users correspond to an embedding matrix. W m In the embedding process, an index sequence is used. T Select Matrix W m The element positions will embed the shadow shares held by participating users into the matrix. W m The corresponding bits at the corresponding element positions in the matrix are used to combine the embedded matrices to obtain the matrix. W The other group of participating users' shadow shares correspond to the embedded matrix. U m In the embedding process, an index sequence is used. T Select Matrix U m The element positions will embed the shadow shares held by participating users into the matrix. U m The corresponding bits at the corresponding element positions in the matrix are used to combine the embedded matrices to obtain the matrix. U ; matrix W As alpha channel and matrix U Merge to generate encrypted image Q .

5. The image steganography method for user-level hierarchical management according to any one of claims 1 to 4, characterized in that, It also includes the decryption process for the encrypted image, including: Determine the number of participating users in the current decryption, obtain the set of shadow shares held by each participating user in the current decryption, and reconstruct the shadow image corresponding to each participating user using Lagrange interpolation. Then, perform dediffusion operation to recover the corresponding image blocks and stitch the image blocks together to obtain the current reconstruction result. Based on the current reconstruction results, an inverse quantization operation is performed. The SL0 algorithm is used to reconstruct the result after the inverse quantization operation. An inverse zigzag scan is performed on the reconstructed output matrix, and the image spatial domain data is recovered through the inverse wavelet transform. Finally, the decrypted image corresponding to the number of participating users is obtained.

6. An image steganography device for user-level hierarchical management, characterized in that, Includes an encryption module for encrypting plaintext images, wherein the encryption module includes: The wavelet transform module is used to perform discrete wavelet transform on plaintext images to generate a transform domain coefficient matrix, which serves as the first matrix. The zigzag scanning module performs a zigzag scan on the first matrix to generate a one-dimensional vector, and then converts it into a second matrix of the same size as the plaintext image. The compressed sensing measurement module sets an initial key value, generates a measurement matrix, and performs compressed sensing measurement on the second matrix using the measurement matrix to generate a third matrix. The quantization processing module quantizes the third matrix to obtain a fourth matrix. The diffusion operation module divides the fourth matrix and the measurement matrix into blocks according to the total number of participating users, with each participating user corresponding to a measurement matrix sub-block. It performs a diffusion operation on each image block obtained from the fourth matrix to obtain the corresponding secret image block. The shadow image generation module is used to generate images based on the total number of participating users. n Dynamically set the differential threshold value for each secret image block k i Threshold values ​​for each secret image block k i Freely set, where 2≤ k i ≤ n Based on the threshold value, a polynomial is constructed for the corresponding secret image block to generate multiple shadow images corresponding to the secret image block, the same number as the total number of participating users. Each participating user corresponds to one shadow image of the secret image block. The polynomial is constructed as follows: in, Assign secret image blocks to the first j The shadow image value of each participating user; j For participating user IDs, ; For polynomial coefficients, For secret image blocks C i pixel values, The coefficients are randomly generated polynomials. For secret image blocks C i The threshold value indicates the minimum requirement. Only a shadow image can recover the secret image block. C i ; p It is a large prime number with a value of 251; The user shadow share determination module is used to combine the shadow images of all image blocks corresponding to each user to form the shadow share held by the user. The ciphertext image generation module is used to embed the shadow share held by each participating user into the carrier image to generate a ciphertext image, thereby realizing the steganography and transmission of the shadow share. The number of users participating in the decryption process when decrypting the encrypted image. t Less than min( k i When the number of users participating in the decryption is sufficient, no valid information related to the plaintext image can be obtained; when the number of users participating in the decryption is sufficient... At that time, relevant information about the plaintext image can be obtained, and the reconstruction quality of the plaintext image improves as the number of participating users increases.

7. The image steganography device for user-level hierarchical management according to claim 6, characterized in that, It also includes a decryption module for decrypting encrypted images, wherein the decryption module includes: The recovery module is used to determine the number of participating users in the current decryption, obtain the set of shadow shares held by each participating user in the current decryption, and reconstruct the shadow image corresponding to each participating user using Lagrange interpolation. Then, it performs dediffusion operation to recover the corresponding image blocks and stitches the image blocks together to obtain the current reconstruction result. The image decryption module is used to perform inverse quantization based on the current reconstruction result, reconstruct the result after inverse quantization using the SL0 algorithm, perform inverse zigzag scanning on the reconstructed output matrix, and recover the image spatial domain data through inverse wavelet transform, finally obtaining the decrypted image corresponding to the number of participating users.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes a computer program to implement the steps of the user-oriented hierarchical image steganography method as described in claim 1.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When a computer program is executed by a processor, it implements the steps of the user-oriented hierarchical image steganography method as described in claim 1.

Images