CN115131253A

CN115131253A - Secret image sharing method and system for resisting JPEG recompression

Info

Publication number: CN115131253A
Application number: CN202210581724.4A
Authority: CN
Inventors: 姜越; 杨国正; 刘京菊; 于龙; 刘林涛; 李龙龙; 陈佳
Original assignee: National University of Defense Technology
Current assignee: National University of Defense Technology
Priority date: 2022-05-26
Filing date: 2022-05-26
Publication date: 2022-09-30
Anticipated expiration: 2042-05-26
Also published as: CN115131253B

Abstract

The invention provides a secret image sharing method and system for resisting JPEG recompression, and belongs to the technical field of image processing. The method comprises the steps that a secret image to be shared is a JPEG image, secret information contained in the JPEG image is a quantized DCT coefficient, JPEG recompression refers to compression processing executed after the JPEG image is shared, and the compression processing is resisted while the JPEG image is shared.

Description

Secret image sharing method and system for resisting JPEG recompression

Technical Field

The invention belongs to the technical field of image processing, and particularly relates to a secret image sharing method and system for resisting JPEG recompression.

Background

Secret sharing technology encrypts secret information into a plurality of shadow images (shadow, shadow image or share) and distributes the shadow images to a plurality of participants, only a subset of authorized participants can be decrypted together, but an unauthorized subset cannot. A secret sharing algorithm generally includes two stages, namely, secret sharing (share or generate) and recovery (receiver), sometimes referred to as encryption (encrypt) and decryption (decrypt) or encoding (encode) and decoding (decode). In a (k, n) threshold secret sharing scheme (scheme), where k is less than or equal to n, secret information is encrypted into n shadow images. Only when k shadow images or more are obtained, the original secret can be decrypted; and less than k shadow images cannot obtain any secret.

Digital images are one of the most important media types, and researchers have extensively studied to apply a secret sharing technique to digital image objects, and the Secret Image Sharing (SIS) technique has been vigorously developed. With respect to data, the particularity of digital images in the field of secret image sharing lies in: (1) a special file storage structure for digital images. Taking a gray-scale BMP format digital image as an example, the pixel value space is [0, 255], so the value ranges of the secret value, the sharing value and the related parameters should be fully considered in the secret image sharing scheme, and the situation that the secret image cannot be recovered due to information loss in the sharing or recovery process is avoided. (2) The digital image is composed of a large number of pixel points, and secret sharing is performed only aiming at one or a plurality of pixel values at a time, so that the high efficiency of a sharing and recovery algorithm is emphasized in the scheme design process. (3) There is a correlation between adjacent pixel values. Continuity and relevance exist between adjacent pixel points of the image, which may cause leakage of secret information of the image. Therefore, the secret image sharing scheme considers both single-share security and visual security. (4) The image transmission is finally identified by a human eye vision system, and due to the low-pass filtering characteristic of human eyes, the lossless recovery image (5) is not required to be special data, and the secret image sharing scheme can be applied to the secret sharing occasion of general data through simple change. The performance evaluation indexes of the secret image sharing scheme comprise: the recovery quality of the secret image, whether pixel expansion exists or not, (k, n) threshold, the recovery complexity of the secret image, comprehensibility, progressiveness, type of the secret image, and the like.

The mainstream principles of secret sharing include: a polynomial-based (k, n) threshold secret sharing scheme, a Chinese remainder theorem-based secret sharing scheme, a visual encryption scheme. The technical scheme is a secret sharing scheme based on a polynomial. A polynomial based (k, n) threshold secret sharing scheme is described below.

In a polynomial secret sharing scheme in the prior art, a secret is embedded into a random k-1 degree polynomial, and the polynomial can be reconstructed by a Lagrange interpolation method during decryption, so that secret information embedded into the polynomial is acquired. Given secret information s, it is shared as n shadow shares sc ₁ ,sc ₂ ,…,sc _n The specific scheme is as follows:

(1) in an initialization phase, the value of a threshold (k, n) is determined, where k ≦ n. Choosing a large prime number p, satisfying p > n and p > s, letting gf (p) be a finite field, all elements being those of gf (p), and all operations being performed in the finite field gf (p).

(2) In the sharing phase, in order to encrypt s as a shadow value sc _i Randomly generating a k-1 degree polynomial in a finite field GF (p):

f(x)＝a ₀ +a ₁ x+…+a _k-1 x ^k-1

in which a secret s is embedded in the first coefficient of a polynomial, i.e. a ₀ S, the remaining coefficients a ₁ ，…，a _k-1 Randomly chosen in the finite field gf (p). Then calculate

sc ₁ ＝f(1),…,sc _k ＝f(k),…,sc _n ＝f(n)

Taking (i, sc) _i ) As a shadow pair, where i is used as an information tag or sequence number tag, sc _i As a shadow pixel value. And distributing the n shadow shares to the n participants respectively to complete secret sharing.

(3) In the recovery phase, any k secret pairs held in the acquiring n participants

Wherein the content of the first and second substances,

the following system of linear equations can be constructed:

because i is _l (1 ≦ l ≦ k) are all different, so the following polynomial can be constructed from the Lagrangian interpolation formula:

thus, the secret s ═ f (0) can be obtained. If k-1 participants want to obtain a secret, k-1 equations can be constructed and organized into a linear system of equations in which the k coefficients sharing the polynomial are unknowns. Due to the label i _l In contrast, each shadow share corresponds to a unique polynomial satisfying a formula linear equation system, so that the known k-1 shadows cannot solve the linear equation system containing k unknowns, so that any information about secrets cannot be obtained, and therefore, the scheme is complete.

With the increasing influence of social networks, Facebook, Twitter, Instagram, Wechat, and the microblog of new waves deeply permeate people's daily lives, and photo sharing has become a popular activity for users to communicate with friends. 350 hundred million photos have been uploaded to Facebook by 2 months 2022. The information is transmitted or stored by utilizing the image on the social network, the hidden transmission and storage of the secret information can be realized, the requirements of convenient and safe communication of the country and the society are met, and the method has important value for guaranteeing the information safety. At present, Secret Image Sharing (SIS) can solve the problem of covert communication and covert storage by taking an Image as a medium. The secret image sharing technology utilizes the idea of secret sharing, divides and stores secrets to prevent the secret image sharing technology from utilizing the idea of secret sharing, divides and stores secrets to prevent secrets from being too concentrated to achieve the purposes of dispersing risks and tolerating intrusion (loss), encrypts secret information into a plurality of shadow images (shadow images or shares) and distributes the shadow images or the shares to a plurality of participants, only subsets of authorized participants can be decrypted together, but unauthorized subsets cannot be decrypted. The general secret image sharing-based covert communication is multi-channel, and the problems that multi-channel covert communication, authority control, loss tolerance and the like cannot be realized in single image steganography can be solved.

However, in a large-scale social network environment, limited by social network performance and a background server, an image passing through a lossy channel of the social network may be subjected to lossy operations such as recompression, which causes a decrease in quality of a shadow image and information loss. Traditional secret image sharing techniques are designed for lossless channels, so that the traditional techniques are not applicable to social network environments. When the existing secret image sharing technology is applied to a social network on the public internet, a distributor shares a secret image into a plurality of shadow images to be delivered to a plurality of participants; a participant uploads a held shadow image to social network accounts such as Facebook, Twitter, WeChat and other social platforms, and the shadow image is transmitted through a channel of the public Internet; the uploaded shadow image can be subjected to lossy operations such as recompression and the like, so that the quality of the shadow image is reduced and information is lost; if a restorer wants to restore successfully after receiving the damaged shadow image, a robust secret image sharing scheme needs to be designed to generate the shadow image which is robust to JPEG recompression.

A robust secret sharing scheme against JPEG recompression is a mandatory way and foundation to apply secret image sharing to social networks. In addition, better secret image sharing properties such as comprehensibility of shadow images, high image quality of restored secret images, and the like should be pursued.

The traditional secret image sharing technology is designed for a lossless channel, and the image is subjected to recompression lossy operation when the image is uploaded to a social network, so that the traditional secret image sharing technology is not suitable for the social network environment. There is currently no robust secret image sharing scheme that is effective for JPEG recompression. A robust secret sharing scheme against JPEG recompression is the mandatory way and foundation to apply secret image sharing to social networks. In order to resist loss of the shadow image caused by JPEG recompression, the patent finds out a stable quantity before and after the JPEG recompression, and constructs a robust shadow image capable of resisting the JPEG recompression by utilizing a screening mechanism of a secret image sharing scheme based on a polynomial and a stable block condition.

Disclosure of Invention

In view of the above technical problem, the present invention provides a secret image sharing scheme for resisting JPEG recompression.

The invention discloses a secret image sharing method for resisting JPEG recompression. The method comprises the steps that a secret image to be shared is a JPEG image, secret information contained in the JPEG image is a quantized DCT (Discrete Cosine Transform) coefficient, JPEG recompression refers to compression processing executed after the JPEG image is shared, and the compression processing is resisted while the JPEG image is shared; the method comprises the following steps:

step S1, extracting n +1 acquired images for preprocessing to extract a complete DCT coefficient list of each image in the n +1 images, wherein the n +1 images comprise 1 secret image to be shared and n carrier images;

step S2, determining a DCT coefficient list to be shared of the secret image to be shared and n DCT coefficient lists to be used corresponding to the n carrier images based on n +1 complete DCT coefficient lists, and determining a prime number p according to the DCT coefficient list to be shared and the maximum DCT coefficient value in the n DCT coefficient lists to be used;

step S3, obtaining n sharing value lists corresponding to the n DCT coefficient lists to be shared and including the secret information of the secret image to be shared by calculating, using the DCT coefficient list to be shared, the n DCT coefficient lists to be used, the prime number p, and a threshold value k;

step S4, for each of the n shared value lists, executing: b multiplied by B shadow DCT blocks are formed according to each sharing value, and decompression processing is carried out on the B multiplied by B shadow DCT blocks to obtain B multiplied by B shadow image null field blocks;

step S5, determining whether element values in each image null field block of n × B shadow image null field blocks obtained based on n sharing value lists are all within a specified range, and if so, regarding the B × B shadow image null field blocks corresponding to each sharing value list, taking the B × B shadow image null field blocks as B × B shadow DCT blocks resisting the JPEG recompression;

step S6, for B × B shadow DCT blocks corresponding to each sharing value list and resisting the JPEG recompression, determining 1 shadow image resisting the JPEG recompression, obtaining n shadow images resisting the JPEG recompression in total, and transmitting the n shadow images resisting the JPEG recompression to a receiving party by a transmitting party to realize sharing the secret image and resisting the JPEG recompression;

wherein n, p, k and B are positive integers, k is less than or equal to n, and a threshold value k represents the number of the minimum shadow images required for recovering the secret image.

According to the method of the first aspect of the present invention, in step S1, the preprocessing specifically includes, for each of the n +1 images:

extracting a DCT coefficient matrix after quantization of a current image through entropy decoding, wherein the DCT coefficient matrix comprises M multiplied by M DCT coefficients, and carrying out blocking processing on the DCT coefficient matrix to divide the DCT coefficient matrix into B multiplied by B DCT blocks, each DCT block comprises A multiplied by A DCT coefficients, and M is multiplied by B multiplied by A;

for each DCT block containing A multiplied by A DCT coefficients, extracting the first C DCT coefficients in a zigzag order to obtain a DCT coefficient list of each DCT block so as to construct a complete DCT coefficient list of the current image, wherein the DCT coefficient list of each DCT block has the length of C, and the complete DCT coefficient list of the current image has the length of C multiplied by B;

wherein M, A, C are all positive integers.

According to the method of the first aspect of the present invention, the step S2 specifically includes:

judging whether the minimum DCT coefficient in the n +1 complete DCT coefficient lists is larger than 0 or not;

if so, taking the complete DCT coefficient list of 1 secret image to be shared in the n +1 complete DCT coefficient lists as the DCT coefficient list to be shared, and taking the complete DCT coefficient list of n carrier images in the n +1 complete DCT coefficient lists as the n DCT coefficient lists to be used;

if not, performing value translation on all DCT coefficients in the n +1 complete DCT coefficient lists, wherein the translation amount of the value translation is the absolute value of the minimum DCT coefficient, taking the complete DCT coefficient list of the 1 to-be-shared secret image after the value translation as the to-be-shared DCT coefficient list, and taking the complete DCT coefficient list of the n carrier images after the value translation as the n to-be-used DCT coefficient lists;

and acquiring the maximum DCT coefficient values in the DCT coefficient list to be shared and the n DCT coefficient lists to be used, and taking the minimum prime number larger than the maximum DCT coefficient as the prime number p.

According to the method of the first aspect of the present invention, the length of the DCT coefficient list to be shared, the length of each of the n DCT coefficient lists to be used, and the length of each of the n shared value lists are C × B; the step S3 specifically includes:

for each position in each of the n lists of shared values, using the formula f (x) s + a ₁ x+a ₂ x ² +…+a _k-1 x ^k-1 (mod p) computing its DCT shadow values;

wherein f (x) is a DCT shadow value at the current position in the current list in the n sharing value lists, s is a DCT coefficient at a position corresponding to the current position in the current list in the DCT coefficient list to be shared, a ₁ 、a ₂ 、...、a _k-1 To be selected arbitrarilyX is a selected value, and modp represents modulo p operation;

judging whether the high delta bit of f (x) is equal to the high delta bit of the DCT coefficient at the position corresponding to the current position in the current list in the n DCT coefficient lists to be used or not, wherein the high delta bit of f (x) is equal to the high delta bit of the DCT coefficient at the position corresponding to the current position in the current list in the n DCT coefficient lists to be used or not

If yes, taking the DCT shadow value f (x) as a sharing value at the current position in a current list in the n sharing value lists of the current position;

if not, adjusting a ₁ 、a ₂ 、...、a _k-1 And recalculating f (x) until the high delta bit of the DCT coefficient is equal to the high delta bit of the DCT coefficient at the position corresponding to the current position in the current list in the n DCT coefficient lists to be used, and acquiring the shared value at the current position.

According to the method of the first aspect of the present invention, in step S3, x remains unchanged when calculating the DCT shadow value for each position in the current list, and the selected values x of n sharing value lists are different, f (x), x, and a ₁ 、a ₂ 、...、a _k-1 Has a value range of [0, p-1 ]]The above integer.

According to the method of the first aspect of the present invention, the step S4 specifically includes:

for each shared value list in the n shared value lists: every time C sharing values are extracted, the C sharing values are spliced with the C +1 th to A x A th DCT coefficients in the DCT blocks corresponding to the carrier images in the n carrier images to form 1 complete shadow DCT list; repeating the operations to obtain n complete shadow DCT lists;

for each DCT list in the n shadow DCT lists: b multiplied by B shadow DCT blocks with the size of A multiplied by A are obtained through inverse zigzag arrangement, decompression processing is respectively carried out on the B multiplied by B shadow DCT blocks, and the decompression processing comprises inverse DCT transformation and rounding processing, so that B multiplied by B shadow image space domain blocks are obtained; repeating the operations to obtain n multiplied by B shadow image spatial domain blocks in total;

before performing the inverse zigzag arrangement, it is determined whether all DCT coefficients in the n +1 complete DCT coefficient lists have been subjected to value translation in step S2, and if yes, all shared values and all DCT values in the n shadow DCT lists are subjected to inverse value translation, where a translation amount of the inverse value translation is an absolute value of the minimum DCT coefficient.

According to the method of the first aspect of the present invention, in said step S5:

the specified range is [ -128,127);

if the element values in each image space domain block in the n multiplied by B shadow image space domain blocks obtained based on the n sharing value lists are not all in the designated range, adjusting a ₁ 、a ₂ 、...、a _k-1 And re-executing steps S3-S5 until the values of the elements in each of the image space blocks are within the specified range.

In the step S6:

b multiplied by B shadow DCT blocks corresponding to each sharing value list and resisting the JPEG recompression are spliced to form 1 shadow DCT matrix, and the 1 shadow DCT matrix is subjected to entropy coding, so that 1 shadow image resisting the JPEG recompression is obtained; repeating the above operations to obtain n pieces of shadow images resisting the JPEG recompression;

obtaining n selected values x of the list of shared values ₁ 、x ₂ 、...、x _n The sender will select the value x ₁ 、x ₂ 、...、x _n Sending the n shadow images to the receiver together, wherein the receiver receives the n shadow images and the selected value x ₁ 、x ₂ 、...、x _n And restoring the secret image, wherein k is less than or equal to l and less than or equal to n.

The invention discloses a secret image sharing system for resisting JPEG recompression. The secret image to be shared is a JPEG image, the secret information contained in the JPEG image is quantized DCT (Discrete Cosine Transform) coefficients, the JPEG recompression refers to compression processing executed after sharing the JPEG image, and the compression processing is resisted by the system while the JPEG image is shared; the system comprises:

a first processing unit configured to: extracting n +1 acquired images for preprocessing so as to extract a complete DCT coefficient list of each image in the n +1 images, wherein the n +1 images comprise 1 secret image to be shared and n carrier images;

a second processing unit configured to: determining a DCT coefficient list to be shared of the secret image to be shared and n DCT coefficient lists to be used corresponding to the n carrier images based on the n +1 complete DCT coefficient lists, and determining a prime number p according to the DCT coefficient list to be shared and the maximum DCT coefficient value in the n DCT coefficient lists to be used;

a third processing unit configured to: acquiring n sharing value lists which correspond to the n DCT coefficient lists to be shared and contain the secret information of the secret image to be shared through calculation by using the DCT coefficient list to be shared, the n DCT coefficient lists to be used, the prime number p and a threshold value k;

a fourth processing unit configured to: for each of the n shared value lists, executing: b multiplied by B shadow DCT blocks are formed according to each sharing value, and decompression processing is carried out on the B multiplied by B shadow DCT blocks to obtain B multiplied by B shadow image space domain blocks;

a fifth processing unit configured to: judging whether the element value of each image null field block in the total n multiplied by B shadow image null field blocks obtained based on the n sharing value lists is within a specified range, if so, regarding the B multiplied by B shadow image null field blocks corresponding to each sharing value list, and taking the B multiplied by B shadow image null field blocks as B multiplied by B shadow DCT blocks resisting the JPEG recompression;

a sixth processing unit configured to: for B multiplied by B (sub) shadow DCT blocks corresponding to each sharing value list and resisting the JPEG recompression, determining 1 shadow image resisting the JPEG recompression, and obtaining n shadow images resisting the JPEG recompression in total;

the sender realizes sharing of the secret image and simultaneously resists the JPEG recompression by sending n shadow images resisting the JPEG recompression to a receiver;

According to the system of the second aspect of the present invention, the preprocessing specifically includes, for each of the n +1 images:

extracting a DCT coefficient matrix after quantization of a current image through entropy decoding, wherein the DCT coefficient matrix comprises M multiplied by M DCT coefficients, and partitioning the DCT coefficient matrix into B multiplied by B DCT blocks, wherein each DCT block comprises A multiplied by A DCT coefficients, and M is multiplied by B multiplied by A;

wherein M, A, C are all positive integers.

According to the system of the second aspect of the invention, the second processing unit is specifically configured to:

According to the system of the second aspect of the present invention, the length of the list of DCT coefficients to be shared, the length of each of the n lists of DCT coefficients to be used, and the length of each of the n lists of shared values are cxbxb; the second processing unit is specifically configured to:

wherein f (x) is a DCT shadow value at the current position in the current list in the n lists of sharing values, s is a DCT coefficient at the position corresponding to the current position in the current list in the list of DCT coefficients to be shared, a ₁ 、a ₂ 、...、a _k-1 Is a randomly selected random number, x is a selected value, and modp represents modulo p operation;

if not, adjusting a ₁ 、a ₂ 、...、a _k-1 And recalculating f (x) until the high delta bit is equal to the high delta bit of the DCT coefficient at the position corresponding to the current position in the current list in the n DCT coefficient lists to be used, and acquiring the score at the current positionAnd (4) sharing the value.

According to the system of the second aspect of the present invention, for each position in the current list, x remains unchanged when calculating its DCT shadow value, and the selected values x of n of the shared value lists are different, f (x), x, and a ₁ 、a ₂ 、...、 a _k-1 Has a value range of [0, p-1 ]]The integer of (1) above.

According to the system of the second aspect of the present invention, the fourth processing unit is specifically configured to:

for each shared value list in the n shared value lists: every time C sharing values are extracted, the C sharing values are spliced with the C +1 th to A x A th DCT coefficients in the DCT blocks corresponding to the carrier images in the n carrier images to form 1 complete shadow DCT list; repeating the above operations to obtain n complete shadow DCT lists;

According to the system of the second aspect of the present invention, the specified range is [ -128,127); the fifth processing unit is specifically configured to: if the element values in each image null field block in the n multiplied by B shadow image null field blocks obtained based on the n sharing value lists are not all in the designated range, adjusting a ₁ 、a ₂ 、...、a _k-1 And re-executing steps S3-S5 until the values of the elements in each of the image space blocks are within the specified range.

According to the system of the second aspect of the present invention, the sixth processing unit is specifically configured to:

for B × B shadow DCT blocks corresponding to each sharing value list and resisting the JPEG recompression, splicing to form the 1 shadow DCT matrix, and performing entropy coding on the 1 shadow DCT matrix to obtain 1 shadow image resisting the JPEG recompression; repeating the operation to obtain n pieces of shadow images resisting the JPEG recompression;

obtaining n selected values x of the list of shared values ₁ 、x ₂ 、...、x _n The sender will select the value x ₁ 、x ₂ 、...、 x _n Sending the n shadow images to the receiver together;

wherein the receiving party is based on the received one shadow image and the selected value x ₁ 、x ₂ 、...、x _n And recovering the secret image, wherein k is less than or equal to l and less than or equal to n.

A third aspect of the invention discloses an electronic device. The electronic device includes a memory storing a computer program and a processor, which when executing the computer program, implements the steps in a secret image sharing method for resisting JPEG recompression according to any one of the first aspects of the present disclosure.

A fourth aspect of the invention discloses a computer-readable storage medium. The computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps in a secret image sharing method for countering JPEG recompression of any one of the first aspects of the present disclosure.

In conclusion, the technical scheme provided by the invention applies the secret image sharing technology to the social network, so that the secret information can be transmitted and stored in a concealed manner, the requirements of convenient and safe communication of the country and the society are met, and the important value is provided for guaranteeing the information safety. The scheme provided by the invention realizes a white box robust scheme aiming at JPEG recompression and realizes comprehensibility of (k, n) threshold and shadow image. The scheme can be applied to the field of social network-oriented covert communication.

Drawings

In order to more clearly illustrate the embodiments or prior art solutions of the present invention, the drawings used in the embodiments or prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without inventive efforts.

FIG. 1 is a flowchart illustrating a secret image sharing method for resisting JPEG recompression according to an embodiment of the present invention;

fig. 2(a-k) is a diagram illustrating the results of a (3,3) threshold, δ 4, num (i.e., C) 9, id [11,13,19], and QF 75 according to an embodiment of the present invention;

fig. 3(a-h) is a diagram illustrating the results of a (2,2) threshold, δ — 3, num (i.e., C) 9, id ═ 11,13, QF ═ 75, according to an embodiment of the present invention;

fig. 4 is a block diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.

Secret sharing: secret Sharing (SS) technology encrypts Secret information into multiple shadow images (or shares) and distributes them to multiple participants, and only a subset of authorized participants can be decrypted together, while an unauthorized subset cannot.

Secret image sharing: secret Image Sharing (SIS) encrypts a Secret Image into a plurality of shadow images (shadows) and distributes the shadow images to a plurality of participants, and only a subset of authorized participants can be decrypted together, while an unauthorized subset cannot.

The shadow image can be understood as follows: the shadow image is understandable, rather than meaningless, which can reduce the doubt of encryption and increase the management efficiency of the shadow image. The quantitative indicators of intelligibility of the shadow images are evaluated by visual quality (PSNR is used herein).

Peak signal-to-noise ratio (PSNR): the image quality of the shadow image and the recovered secret image is evaluated by adopting the indexes of the PSNR.

(k, n) threshold: of the n shadow images, k or more shadow images are required to recover the secret. When the threshold is k, certain fault tolerance is provided, and at most n-k shadows can be allowed to be lost.

And (3) mean filtering: the method is the most commonly used means in image processing, and the mean value filtering is a low-pass filter from the viewpoint of frequency domain, so that high-frequency signals can be removed, and the method can help to eliminate sharp noise of images and realize the functions of smoothing and blurring the images. Ideally, the mean filtering is performed by replacing each pixel in the image with the mean value calculated for each pixel and its surrounding pixels. The sampled Kernel data is typically a 3 x 3 matrix, but may be of any shape.

JPEG image: jpeg (joint Photographic Experts group), a standard for compression of continuous tone still images, has a file suffix of jpg or jpeg, which is the most commonly used format for image files.

The compression Quality Factor (QF) is obtained by calculation (the calculation method is shown in the following formula). Table 1 shows a quantization table with a compression quality factor QF of 75, where the elements in the quantization table control the compression ratio, with larger values resulting in greater compression.

Where Q0(u, v) represents the quantization step size at the (u, v) position in the standard quantization table.

8	6	5	8	12	20	26	31
								6	6	7	10	13	29	30	28
7	7	8	12	20	29	35	28
								7	9	11	15	26	44	40	31
9	11	19	28	34	55	52	39
								12	18	28	32	41	52	57	46
25	32	39	44	52	61	60	51
								36	46	48	49	56	50	52	50

TABLE 1 QF 75 quantization table

White-box robust: in this application, white-box robustness refers to the known compression quality factor QF of the recompressed channel or the use of the recompressed channel at its own discretion, where robustness to JPEG recompression is a kind of white-box robustness.

The invention discloses a secret image sharing method for resisting JPEG recompression. The secret image to be shared is a JPEG image, the secret information contained in the JPEG image is quantized DCT (Discrete Cosine Transform) coefficients, the JPEG recompression refers to compression processing executed after sharing the JPEG image, and the method resists the compression processing while sharing the JPEG image.

Specifically, the present application finds a stable amount that is stable and unchanging before and after recompression, i.e., when QM1 is QM2(QM1 represents the quantization table matrix for secret jpeg images, and QM2 represents the quantization table matrix used in recompression), and the recompression channel QF ≦ 92, as long as: -128 ≦ udct (d) < 127, and the DCT coefficients before and after recompression do not change, which condition is referred to herein as the stable block condition, i.e., the block is said to be a stable block if the spatial pixel values obtained after IDCT transform of each 8 × 8 of the DCT coefficient matrix (hereinafter referred to as the original DCT coefficient matrix) after entropy decoding of the original image are found to be present in [ -128,127). The application provides a method for constructing a robust shadow image based on a stable block condition, so that the DCT number of the constructed shadow image is not changed when the constructed shadow image passes through a recompression channel, namely the constructed shadow image is robust to the recompression channel.

The method comprises the following steps:

step S4, for each of the n shared value lists, executing: b multiplied by B shadow DCT blocks are formed according to each sharing value, and the B multiplied by B shadow DCT blocks are decompressed to obtain B multiplied by B shadow image space domain blocks;

In some embodiments, in the step S1, the preprocessing specifically includes, for each image of the n +1 images:

wherein M, A, C are all positive integers.

In some embodiments, the step S2 specifically includes:

if yes, taking the complete DCT coefficient list of 1 secret image to be shared in the n +1 complete DCT coefficient lists as the DCT coefficient list to be shared, and taking the complete DCT coefficient list of n carrier images in the n +1 complete DCT coefficient lists as the n DCT coefficient lists to be used;

In some embodiments, the length of the list of DCT coefficients to be shared, the length of each of the n lists of DCT coefficients to be used, and the length of each of the n lists of shared values are C × B; the step S3 specifically includes:

wherein f (x) is a DCT shadow value at the current position in the current list in the n sharing value lists, s is a DCT coefficient at a position corresponding to the current position in the current list in the DCT coefficient list to be shared, a ₁ 、a ₂ 、...、a _k-1 Is a randomly selected random number, x is a selected value, and modp represents modulo p operation;

if not, adjusting a ₁ 、a ₂ 、...、a _k-1 And recalculating f (x) until the high delta bit of the DCT coefficient is higher than the high delta bit of the DCT coefficient at the position corresponding to the current position in the current list in the n lists of DCT coefficients to be usedThe bits are equal and the shared value at the current position is obtained.

In some embodiments, in step S3, for each position in the current list, x remains unchanged when calculating its DCT shadow value, and the selected values x of n pieces of the shared value list are different from each other, f (x), x, and a ₁ 、a ₂ 、...、a _k-1 Has a value range of [0, p-1 ]]The above integer.

In some embodiments, the step S4 specifically includes:

before performing the inverse zigzag arrangement, it is determined whether all DCT coefficients in the n +1 complete DCT coefficient lists have been value-translated in step S2, if yes, performing inverse value translation on all shared values and all DCT values in the n shadow DCT lists, where a translation amount of the inverse value translation is an absolute value of the minimum DCT coefficient.

In some embodiments, in said step S5:

the specified range is [ -128,127);

if the element values in each image null field block in the n multiplied by B shadow image null field blocks obtained based on the n sharing value lists are not all in the designated range, adjusting a ₁ 、a ₂ 、...、a _k-1 And re-executing steps S3-S5 until the values of the elements in each of the image space blocks are within the specified range.

In some embodiments, in said step S6:

b multiplied by B shadow DCT blocks corresponding to each sharing value list and resisting the JPEG recompression are spliced to form 1 shadow DCT matrix, and the 1 shadow DCT matrix is subjected to entropy coding, so that 1 shadow image resisting the JPEG recompression is obtained; repeating the operation to obtain n shadow images resisting the JPEG recompression;

obtaining n selected values x of the list of shared values ₁ 、x ₂ 、...、x _n Said sender will select said value x ₁ 、x ₂ 、...、 x _n Sending the n shadow images to the receiver together, wherein the receiver receives the n shadow images and the selected value x ₁ 、x ₂ 、...、x _n And restoring the secret image, wherein k is less than or equal to l and less than or equal to n.

The algorithm of the specific embodiment is as follows (in conjunction with fig. 1):

for the ith original DCT coefficient matrix block S _ DCTblock of the secret jpeg image S and the n carrier jpeg images _i ,cover ₁ _DCTblock _i ,…,cover_DCTblock _i Firstly, extracting the first num bits of the zigzag arrangement of the blocks and inputting the bits into a secret sharing algorithm which is understandable by a shadow based on a polynomial to obtain n shadow DCT lists, splicing the n shadow DCT lists with the last 64-num of the corresponding shadow image, and obtaining n shadow DCT blocks SC after inverse zigzag arrangement ₁ _DCTblock' _i ,SC ₂ _DCTblock' _i ,…,SC _n _DCTblock' _i 。

Then decompressing it, which includes IDC conversionRounding off, decompressing into shadow image space block SC ₁ _Spatialblock _i ,SC ₂ _Spatialblock _i ,…,SC _n _Spatialblock _i 。

The next step is the crucial step, namely, whether all elements of the n spatial pixel blocks are in the range of [ -128,127) is judged, if all elements are in the range, the shadow DCT blocks generated in the round are indicated as stable blocks, and the stable blocks are directly stored; if any element exceeds the range of [ -128,127), the generated shadow DCT block is declared to be an unstable block, the secret sharing algorithm which is understandable by the polynomial-based shadow is returned, and the random number is continuously screened until the stable shadow DCT block is generated. It is noted that there may not be a set of random numbers that will stabilize all shadow DCT blocks (a specific analysis will be given below), so the maximum filter number MAX is set here.

In the recovery stage, entropy decoding the obtained shadow images which are more than or equal to k to obtain quantized DCT coefficients of the shadow images; dividing the quantized DCT coefficients into 8 x 8 blocks; then, each block is subjected to zigzag arrangement, and the front num bits of the data after zigzag arrangement are extracted as an object to be recovered; and translating all DCT coefficients to a positive number range according to the determined minimum translation value in the sharing process, recovering by using a Lagrange interpolation method, then reversely translating, then supplementing 64 bits by using 64-num zeros, then entropy coding, and finally storing as a secret image.

And (3) experimental verification process:

in order to verify the effectiveness of the robust secret image sharing scheme of the robust shadow image construction based on the stable block condition, the sharing algorithm and the recovery method provided by the application implement a local simulation experiment. The experimental pictures of the application are derived from BOSSbase1.0, 4 images with the size of 256 multiplied by 256 gray scales are randomly selected and converted into JPEG images with the quality factor of 75. Entropy decoding is simulated by a read () function of the JPEGIO packet, and entropy encoding is simulated by a write () function of the JPEGIO packet.

A total of 2 experiments are presented here. Fig. 2 shows the (3,3) threshold, δ 4, num 9, id [11,13,19]]Robust carrier based on stable block condition, QF 75Experimental results of robust secret image sharing for volumetric image construction. Fig. 2(a) shows an input grayscale secret JPEG image S with a size of 256 × 256, QF 75. Fig. 2(b-d) shows input 3 JPEG images cover1, cover2 and cover3, with 256 × 256 size, QF being 75. The obtained robust carriers JPEG image stable _ SC1, stable _ SC2 and stable _ SC3 after applying the algorithm proposed by the application are shown in FIG. 2(e-g), and the size is also 256 × 256. Fig. 2(h-j) shows the 3 recompressed grayscale carrier JPEG images of size 256 × 256 from the compression channel with compression quality factor of 75, recom _ SC1, recom _ SC2, and recom _ SC 3. FIG. 2(k) is a recovered secret JPEG image S ^* . It can be seen from the recovered secret image that there are several blocks whose recovery is not normal, since no stable carrier image is screened when the screening number reaches the maximum condition. Experiments show that the scheme of the threshold (3,3) is feasible, and the quality of the generated shadow image and the quality of the recovered secret image are both higher.

Fig. 3 shows the (2,2) threshold, δ being 3, num being 9, id being [11,13]]QF is 75. the robust carrier image based on the stable block condition constructs the experimental result of the robust secret image sharing. Fig. 3(a) shows an input grayscale secret JPEG image S with a size of 256 × 256, QF 75. Fig. 3(b-c) shows input 2 grayscale carrier JPEG images cover1 and cover2 with size 256 × 256, QF 75. The JPEG images of robust vector, stable _ SC1, stable _ SC2 and stable _ SC3 obtained after applying the algorithm proposed in the present application are shown in fig. 3(d-e), and the size thereof is also 256 × 256. Fig. 3(f-g) shows the 2 recompressed grayscale carrier JPEG images recom _ SC1 and recom _ SC2 after passing through the compression channel with a compression quality factor of 75. FIG. 3(h) is a recovered secret JPEG image S ^* . From the recovered secret image, it can be seen that the recovery of several blocks is not normal similarly to the previous experiment, since a stable carrier image is not screened when the screening number reaches the maximum condition. Experiments show that the scheme of the threshold (2,2) is feasible, and the quality of the generated shadow image and the quality of the recovered secret image are both higher.

The invention discloses a secret image sharing system for resisting JPEG recompression. The secret image to be shared is a JPEG image, the secret information contained in the JPEG image is a quantized DCT (Discrete Cosine Transform) coefficient, the JPEG recompression refers to compression processing executed after the JPEG image is shared, and the compression processing is resisted by the system while the JPEG image is shared; the system comprises:

wherein M, A, C are all positive integers.

According to the system of the second aspect of the present invention, the length of the DCT coefficient list to be shared, the length of each of the n DCT coefficient lists to be used, and the length of each of the n sharing value lists are C × B; the second processing unit is specifically configured to:

According to the system of the second aspect of the present invention, for each position in the current list, x remains unchanged when calculating its DCT shadow value, and the selected values x of n of the shared value lists are different from each other, f (x), x, and a ₁ 、a ₂ 、...、 a _k-1 Has a value range of [0, p-1 ]]The above integer.

for each shared value list in the n shared value lists: every time C sharing values are extracted, the C sharing values are spliced with C +1 th to A x A th DCT coefficients in corresponding DCT blocks of corresponding carrier images in the n carrier images to form 1 complete shadow DCT list; repeating the operations to obtain n complete shadow DCT lists;

According to the system of the second aspect of the present invention, the specified range is [ -128,127); the fifth processing unit is specifically configured to: if the element value of each image empty domain block in the n multiplied by B shadow image empty domain blocks obtained based on the n sharing value lists is not all in the designated rangeInner, then adjust a ₁ 、a ₂ 、...、a _k-1 And re-executing steps S3-S5 until the values of the elements in each of the image space blocks are within the specified range.

obtaining n selected values x of the list of shared values ₁ 、x ₂ 、...、x _n Said sender will select said value x ₁ 、x ₂ 、...、x _n Sending the n shadow images to the receiver together;

wherein the receiving side is based on the received one shadow image and the selected value x ₁ 、x ₂ 、...、x _n And recovering the secret image, wherein k is less than or equal to l and less than or equal to n.

A third aspect of the invention discloses an electronic device. The electronic device comprises a memory storing a computer program and a processor implementing the steps of a secret image sharing method for resisting JPEG recompression according to any one of the first aspect of the present disclosure when the computer program is executed.

Fig. 4 is a block diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 4, the electronic device includes a processor, a memory, a communication interface, a display screen, and an input device, which are connected by a system bus. Wherein the processor of the electronic device is configured to provide computing and control capabilities. The memory of the electronic device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The communication interface of the electronic device is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be realized through WIFI, an operator network, Near Field Communication (NFC) or other technologies. The display screen of the electronic equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the electronic equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the electronic equipment, an external keyboard, a touch pad or a mouse and the like.

It will be understood by those skilled in the art that the structure shown in fig. 4 is only a partial block diagram related to the technical solution of the present disclosure, and does not constitute a limitation to the electronic device to which the solution of the present application is applied, and a specific electronic device may include more or less components than those shown in the drawings, or combine some components, or have different component arrangements.

It should be noted that the technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, however, as long as there is no contradiction between the combinations of the technical features, the scope of the present description should be considered. The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, and these are all within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims

1. A secret image sharing method for resisting JPEG recompression is characterized in that a secret image to be shared is a JPEG image, secret information contained in the JPEG image is quantized DCT (Discrete Cosine Transform) coefficients, the JPEG recompression refers to compression processing executed after the JPEG image is subjected to sharing processing, and the method resists the compression processing while the JPEG image is shared; the method comprises the following steps:

step S5, determining whether element values in each image null field block of the total of n × B shadow image null field blocks obtained based on the n sharing value lists are all within a specified range, if yes, regarding the B × B shadow image null field blocks corresponding to each sharing value list, taking the B × B shadow image null field blocks as B × B shadow DCT blocks resisting the JPEG recompression;

step S6, for B × B shadow DCT blocks corresponding to each sharing value list and resisting the JPEG recompression, determining 1 shadow image resisting the JPEG recompression, and obtaining n shadow images resisting the JPEG recompression in total, wherein a sender realizes sharing the secret image and resisting the JPEG recompression by sending the n shadow images resisting the JPEG recompression to a receiver;

2. A secret image sharing method against JPEG recompression as claimed in claim 1, characterized in that in said step S1, said preprocessing specifically includes, for each of said n +1 images:

wherein M, A, C are all positive integers.

3. The secret image sharing method for resisting JPEG recompression according to claim 2, characterized in that said step S2 specifically includes:

4. A secret image sharing method against JPEG recompression as claimed in claim 3, characterized in that the length of said list of DCT coefficients to be shared, the length of each of said n lists of DCT coefficients to be used, and the length of each of said n lists of sharing values are all cxbxb; the step S3 specifically includes:

If yes, taking the DCT shadow value f (x) as a sharing value at the current position in a current list in the n sharing value lists at the current position;

5. The secret image sharing method against JPEG recompression of claim 4, wherein in said step S3, for each position in the current list, x is kept unchanged while calculating its DCT shadow value, and the selected values x of n sharing value lists are different, f (x), x, and a ₁ 、a ₂ 、...、a _k-1 Has a value range of [0, p-1 ]]The above integer.

6. The secret image sharing method for resisting JPEG recompression according to claim 5, characterized in that said step S4 specifically includes:

for each shared value list in the n shared value lists: every time C sharing values are extracted, the C sharing values are spliced with C +1 th to AxA th DCT coefficients in corresponding DCT blocks of corresponding carrier images in the n carrier images to form 1 complete shadow DCT list; repeating the above operations to obtain n complete shadow DCT lists;

for each DCT list in the n shadow DCT lists: b multiplied by B shadow DCT blocks with the size of A multiplied by A are obtained through inverse zigzag arrangement, decompression processing is respectively carried out on the B multiplied by B shadow DCT blocks, the decompression processing comprises inverse DCT transformation and rounding processing, and therefore B multiplied by B shadow image space domain blocks are obtained; repeating the above operations to obtain n multiplied by B shadow image empty domain blocks in total;

7. A secret image sharing method against JPEG recompression as claimed in claim 6, characterized in that:

in the step S5:

the specified range is [ -128,127);

In the step S6:

obtaining n selected values x of the list of shared values ₁ 、x ₂ 、...、x _n Said sender will select said value x ₁ 、x ₂ 、...、x _n Sending the n shadow images to the receiver together, wherein the receiver receives the n shadow images and the selected value x ₁ 、x ₂ 、...、x _n And restoring the secret image, wherein k is less than or equal to l and less than or equal to n.

8. A secret image sharing system for resisting JPEG recompression, wherein a secret image to be shared is a JPEG image, secret information contained in the JPEG image is quantized DCT (Discrete Cosine Transform) coefficients, the JPEG recompression refers to compression processing executed after the JPEG image is subjected to sharing processing, and the system resists the compression processing while the JPEG image is shared; the system comprises:

a third processing unit configured to: acquiring n sharing value lists which correspond to the n DCT coefficient lists to be shared and contain the secret information of the secret image to be shared through calculation by utilizing the DCT coefficient list to be shared, the n DCT coefficient lists to be used, the prime number p and a threshold value k;

a fourth processing unit configured to: for each shared value list in the n shared value lists, executing: b multiplied by B shadow DCT blocks are formed according to each sharing value, and the B multiplied by B shadow DCT blocks are decompressed to obtain B multiplied by B shadow image space domain blocks;

9. An electronic device, characterized in that the electronic device comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of a secret image sharing method against JPEG recompression as claimed in any of claims 1-7.

10. A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of a secret image sharing method for countering JPEG recompression of any one of claims 1-7.