Robust steganography method and device for large-size image with high quality factor

Abstract

The invention relates to the technical field of digital media processing, in particular to a robust steganography method and device for an image with a high quality factor and a large size, and the method comprises the steps: decoding a carrier image, obtaining a first spatial pixel value of the carrier image, and obtaining a second spatial pixel value of the carrier image based on the channel characteristics of a target network platform; carrying out pre-scaling processing on a spatial domain image corresponding to the carrier image, and obtaining a discrete cosine transform (DCT) coefficient; embedding the target secret information into the DCT coefficient to obtain an initial secret-carrying DCT coefficient, and adjusting an unstable coefficient of the initial secret-carrying DCT coefficient to obtain a final secret-carrying DCT coefficient; transforming the final density-carrying DCT coefficient into a spatial domain to obtain a second spatial domain pixel value, modifying the first spatial domain pixel value according to the second spatial domain pixel value until a spatial domain image corresponding to the modified first spatial domain pixel value meets a first preset scaling condition, and compressing the spatial domain image corresponding to the modified first spatial domain pixel value to obtain a compressed spatial domain image corresponding to the modified first spatial domain pixel value. And generating a final secret-carrying image meeting the preset size.

Description

Technical Field

[0001] The present application relates to the field of digital media processing technology, and in particular to a robust steganography method and device for images with high quality factors and large sizes. Background Art

[0002] In the related art, early image steganography algorithms assumed that the communication channel was lossless, and did not consider the possible lossy processing such as scaling and compression in the channel. They focused more on how to adaptively select the information embedding position according to the content of the image, so that the encrypted image after embedding the information has better anti-statistical detection performance. Therefore, in order to solve the problem of covert communication in lossy channels, robust steganography technology has gradually attracted the attention of researchers in the field of steganography. In the subsequent research on robust image steganography algorithms, only scaling or only considering a single lossy processing of recompression is often considered. Anti-scaling steganography algorithms often embed and modify in the spatial domain, usually assuming that the carrier file format is lossless, and are not robust to JPEG compression, so their scope of use is very limited; when using the anti-JPEG recompression scheme, the size needs to be adjusted to below the threshold of the communication channel to avoid triggering the scaling operation, which is behaviorally unsafe.

[0003] However, the image steganography algorithms in related technologies do not take into account the lossy processing in the channel. The generated secret images often cannot correctly extract the secret messages, or only consider single lossy processing such as scaling or heavy compression. The monitoring party can easily observe the quality factor and size of the images uploaded by the user, and then distinguish between communication behavior and public behavior. The scope of use is relatively limited and may lead to a series of behavioral security issues, which need to be solved urgently. Summary of the invention

[0004] The present application provides a robust steganography method, device, electronic device and storage medium for images with high quality factors and large sizes, so as to solve the problem that the image steganography algorithm in the related art does not take into account the lossy processing existing in the channel, and the generated secret image often cannot correctly extract the secret message, or only considers the single lossy processing of scaling or heavy compression, and the monitoring party can easily observe the quality factor and size of the image uploaded by the user, and then distinguish between communication behavior and public behavior. The scope of use is relatively limited and may lead to a series of behavioral security issues.

[0005] The first aspect of the present application provides a robust steganography method for images with high quality factors and large sizes, comprising the following steps: decoding a carrier image to obtain a first spatial domain pixel value of the carrier image, and pre-scaling the spatial domain image corresponding to the carrier image based on the channel characteristics of the target network platform to obtain a pre-scaled spatial domain image, and obtaining discrete cosine transform DCT coefficients of the pre-scaled spatial domain image; embedding target secret information into the DCT coefficients to obtain initial steganographic DCT coefficients, adjusting unstable coefficients in the initial steganographic DCT coefficients to obtain final steganographic DCT coefficients; transforming the final steganographic DCT coefficients into the spatial domain to obtain corresponding second spatial domain pixel values, and modifying the first spatial domain pixel values ​​according to the second spatial domain pixel values ​​until the spatial domain image corresponding to the modified first spatial domain pixel values ​​meets a first preset scaling condition, and compressing the spatial domain image corresponding to the modified first spatial domain pixel values ​​according to the quality factor of the carrier image to generate a final steganographic image that meets a preset size.

[0006] Optionally, in one embodiment of the present application, the embedding of the target secret information into the DCT coefficients to obtain initial encrypted DCT coefficients includes: performing stego quantization processing on the DCT coefficients to generate a stego carrier; embedding the target secret information into the stego carrier to obtain a encrypted sequence; modifying the DCT coefficients according to the carrier sequence of the stego carrier and the encrypted sequence to obtain the initial encrypted DCT coefficients.

[0007] Optionally, in one embodiment of the present application, before embedding the target secret information into the DCT coefficient, it also includes: obtaining quantization step information of the coefficient block composed of the DCT coefficient; determining the target position of embedding the target secret information into the DCT coefficient according to the quantization step information.

[0008] Optionally, in one embodiment of the present application, the adjusting of the unstable coefficients in the initial dense DCT coefficients to obtain the final dense DCT coefficients includes: performing inverse discrete cosine transform processing, spatial intake processing and stage processing on the initial dense DCT coefficients to obtain the corresponding first simulated spatial pixel values; based on the first simulated spatial pixel values, modifying the first spatial pixel values ​​to obtain second simulated spatial pixel values; obtaining a pre-sent intermediate picture that meets a second preset scaling condition according to the second simulated spatial pixel values; performing simulated channel processing on the pre-sent intermediate picture to adjust the unstable coefficients in the initial dense DCT coefficients to obtain the final dense DCT coefficients.

[0009] Optionally, in one embodiment of the present application, modifying the first spatial domain pixel value according to the second spatial domain pixel value includes: calculating the modification cost of each pixel in the spatial domain image using a target distortion function; calculating the interpolation block corresponding to each element in the second spatial domain pixel value on the spatial domain image, and obtaining a weight value of each element in the interpolation block; sorting the weight values ​​according to the modification cost to generate a modification sequence; and modifying the first spatial domain pixel value according to the second spatial domain pixel value based on the modification sequence.

[0010] The second aspect of the present application provides a robust steganographic device for images with high quality factors and large sizes, including: a first processing module, used to decode a carrier image to obtain a first spatial domain pixel value of the carrier image, and based on the channel characteristics of the target network platform, pre-scale the spatial domain image corresponding to the carrier image to obtain a pre-scaled spatial domain image, and obtain the discrete cosine transform DCT coefficient of the pre-scaled spatial domain image; a second processing module, used to embed target secret information into the DCT coefficient to obtain an initial load-carrying DCT coefficient, adjust the unstable coefficient in the initial load-carrying DCT coefficient to obtain a final load-carrying DCT coefficient; a steganographic module, used to transform the final load-carrying DCT coefficient into the spatial domain to obtain a corresponding second spatial domain pixel value, and modify the first spatial domain pixel value according to the second spatial domain pixel value until the spatial domain image corresponding to the modified first spatial domain pixel value meets the first preset scaling condition, and compress the spatial domain image corresponding to the modified first spatial domain pixel value according to the quality factor of the carrier image to generate a final load-carrying image that meets the preset size.

[0011] Optionally, in one embodiment of the present application, the second processing module includes: a first processing unit, used to perform stego quantization processing on the DCT coefficients to generate a stego carrier; an embedding unit, used to embed the target secret information into the stego carrier to obtain a stego sequence; and a first modification unit, used to modify the DCT coefficients according to the carrier sequence of the stego carrier and the stego sequence to obtain the initial stego DCT coefficients.

[0012] Optionally, in one embodiment of the present application, it also includes: an acquisition module, which acquires quantization step information of a coefficient block composed of the DCT coefficients before embedding the target secret information into the DCT coefficients; and a determination module, which is used to determine the target position of embedding the target secret information into the DCT coefficients based on the quantization step information.

[0013] Optionally, in one embodiment of the present application, the second processing module includes: a second processing unit, used to perform inverse discrete cosine transform processing, spatial intake processing and stage processing on the initial dense DCT coefficients to obtain corresponding first simulated spatial pixel values; a second modification unit, used to modify the first spatial pixel values ​​based on the first simulated spatial pixel values ​​to obtain second simulated spatial pixel values; a generation unit, used to obtain a pre-sent intermediate image that meets a second preset scaling condition based on the second simulated spatial pixel values; an adjustment unit, used to perform simulated channel processing on the pre-sent intermediate image to adjust the unstable coefficients in the initial dense DCT coefficients to obtain final dense DCT coefficients.

[0014] Optionally, in one embodiment of the present application, the steganography module includes: a first calculation unit, used to calculate the modification cost of each pixel in the spatial domain image using a target distortion function; a second calculation unit, used to calculate the interpolation block corresponding to each element in the second spatial domain pixel value on the spatial domain image, and obtain the weight value of each element in the interpolation block; a sorting unit, used to sort the weight values ​​according to the modification cost to generate a modification sequence; and a third modification unit, used to modify the first spatial domain pixel value according to the second spatial domain pixel value based on the modification sequence.

[0015] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the robust steganography method for images with high quality factors and large sizes as described in the above embodiments.

[0016] The fourth aspect embodiment of the present application provides a computer-readable storage medium, which stores a computer program, which, when executed by a processor, implements the above-mentioned robust steganography method for images with high quality factors and large sizes.

[0017] The fifth aspect of the present application provides a computer program product, including a computer program, which, when executed, is used to implement the above-mentioned robust steganography method for images with high quality factors and large sizes.

[0018] The embodiment of the present application can use the pre-scaled image of the original carrier image as a medium, embed the secret message based on the jitter modulation algorithm combined with the adaptive steganography framework, adjust the unstable DCT coefficients through analog channel processing, and finally generate a large-size, high-quality secret image through inverse interpolation processing. As a result, the generated secret image has the same quality factor and size as the original carrier image, can perform covert communication, and can resist the scaling and JPEG recompression attacks in the lossy social platform channel, and correctly extract the secret information in the secret image. As a result, the image steganography algorithm in the related art does not consider the lossy processing in the channel, and the generated secret image often cannot correctly extract the secret message, or only considers the single lossy processing of scaling or recompression. The monitoring party can easily observe the quality factor and size of the image uploaded by the user, and then distinguish between communication behavior and mass behavior. The scope of use is relatively limited and may lead to a series of behavioral security issues.

[0019] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0021] Figure 1 A schematic diagram of a social platform channel processing process according to an embodiment of the present application;

[0022] Figure 2 A schematic diagram of a framework of a robust steganographic method for images with high quality factors and large sizes according to an embodiment of the present application;

[0023] Figure 3 A flowchart of a robust steganography method for images with high quality factors and large sizes provided according to an embodiment of the present application;

[0024] Figure 4 A schematic diagram of fine jitter modulation according to an embodiment of the present application;

[0025] Figure 5 A schematic diagram of the structure of a robust steganographic device for images with high quality factors and large sizes provided according to an embodiment of the present application;

[0026] Figure 6 It is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present application.

[0027] Reference numerals:

[0028] 10-Robust steganographic device for images with high quality factor and large size: 100-first processing module, 200-second processing module and 300-steganographic module; 601-memory, 602-processor and 603-communication interface. DETAILED DESCRIPTION

[0029] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0030] The following describes the robust steganographic method and device for images with high quality factors and large sizes according to the embodiments of the present application with reference to the accompanying drawings. The image steganographic algorithm in the related art mentioned in the above background technology does not take into account the lossy processing in the channel, and the generated secret image often cannot correctly extract the secret message, or only considers the single lossy processing of scaling or recompression. The monitoring party can easily observe the quality factor and size of the image uploaded by the user, and then distinguish the communication behavior from the public behavior. The scope of use is relatively limited and may cause a series of behavioral security problems. The present application provides a robust steganographic method for images with high quality factors and large sizes. In this method, the pre-scaled image of the original carrier image can be used as a medium, and the secret message is embedded based on the jitter modulation algorithm combined with the adaptive steganographic framework, and the unstable DCT coefficients are adjusted by analog channel processing, and finally a large-size, high-quality secret image is generated by inverse interpolation processing. As a result, the generated secret image has the same quality factor and size as the original carrier image, can perform covert communication behavior, and can resist scaling and JPEG recompression attacks in lossy social platform channels, and correctly extract the secret information in the secret image. This solves the problem that the image steganography algorithm in the related technology does not take into account the lossy processing in the channel, and the generated secret image often cannot correctly extract the secret message, or only considers a single lossy processing of scaling or heavy compression. The monitoring party can easily observe the quality factor and size of the image uploaded by the user, and then distinguish between communication behavior and mass behavior. The scope of use is relatively limited and may lead to a series of behavioral security issues.

[0031] Before explaining the robust steganography method for images with high quality factors and large sizes in the embodiments of the present application, the system framework of the robust steganography method for images with high quality factors and large sizes in the embodiments of the present application is first explained.

[0032] Steganography is a technique that hides communication behavior under normal behavior to achieve covert communication. Steganography embeds secret information into a carrier sample in an imperceptible way to generate a carrier sample, and transmits the carrier sample to achieve covert transmission of secret information. Since images themselves contain rich information and the human eye is insensitive to subtle changes, coupled with the widespread popularity of image sharing on social networking platforms, images have become a high-quality carrier for covert communication on social networking platforms. With the increasing number of users on social networking platforms, a huge amount of data is uploaded to social networking platforms every day. Considering the network communication speed and data storage capacity, image files often undergo lossy operations such as scaling and compression to reduce the amount of data when they are uploaded to social networking platforms. These lossy processes usually do not affect the content information represented by the original data, but will cause irreversible damage to the image data, which may cause the secret information in the carrier image to be destroyed and cannot be correctly extracted.

[0033] In real life, people usually share photos taken with their mobile phones on social networking platforms. These images are usually JPEG images with large sizes and high quality factors. After uploading these images to social networking platforms, they are usually scaled to a smaller size and recompressed with a smaller quality factor. For example, Figure 1 The figure is a schematic diagram of the social platform channel processing process according to an embodiment of the present application.

[0034] The essence of steganography is to make communication behavior indistinguishable from mass behavior. However, normal users upload high-quality factor and large-size photos taken by mobile phones, while abnormal steganographic users upload low-quality factor or small-size images, so the monitoring party can easily observe the quality factor and size of the images uploaded by users. In contrast, steganography and uploading pictures with large size and high-quality factor is a more secure communication method in terms of behavior. Therefore, it is necessary to design a robust steganographic algorithm that is resistant to scaling and JPEG recompression for images with high-quality factor and large size.

[0035] In response to the problem that social platforms will scale and JPEG re-compress images with large size and high quality factor, and the relevant robust steganography algorithms cannot resist these two attacks at the same time, this application proposes a robust steganography algorithm that is resistant to scaling and JPEG re-compression for images with high quality factor and large size, aiming to solve the problem of covert communication of images with high quality factor and large size on social platforms with scaling and JPEG re-compression.

[0036] Figure 2 This is a schematic diagram of a framework of a robust steganographic method for images with high quality factors and large sizes according to an embodiment of the present application. Figure 2As shown, the framework mainly includes four modules: preprocessing, message embedding based on jitter modulation, coefficient adjustment based on analog channel processing, and secret image generation. For the convenience of explanation, this application mainly uses uppercase letters to represent matrices and vectors, and uses corresponding lowercase letters to represent elements in matrices or vectors. This application uses pre-scaled images as a medium, embeds secret messages based on a jitter modulation algorithm combined with an adaptive steganography framework, and adjusts unstable DCT coefficients through analog channel processing. Finally, a large-size, high-quality secret image is generated through an inverse interpolation algorithm, thereby realizing a robust steganography algorithm that is resistant to scaling and JPEG re-compression for images with high quality factors and large sizes.

[0037] Specifically, Figure 3 A flowchart of a robust steganography method for images with high quality factors and large sizes provided in an embodiment of the present application.

[0038] like Figure 3 As shown, the robust steganography method for images with high quality factors and large sizes includes the following steps:

[0039] In step S301, the carrier image is decoded to obtain the first spatial domain pixel value of the carrier image, and based on the channel characteristics of the target network platform, the spatial domain image corresponding to the carrier image is pre-scaled to obtain the pre-scaled spatial domain image, and the discrete cosine transform DCT coefficients of the pre-scaled spatial domain image are obtained.

[0040] It can be understood that the carrier image here refers to the original image that will carry the secret information. The spatial domain image here can be understood as a distribution representation of the carrier image in space. Each pixel in the spatial domain image corresponds to a point in the carrier image in space, and the spatial domain pixel value reflects the brightness or color information at that point. In the spatial domain, the processing of the image usually acts directly on these pixel values.

[0041] In some embodiments, in order to facilitate image operations, the present application can perform a decoding operation on the carrier image to obtain a first spatial pixel value, and can process the spatial image of the carrier image with the same scaling algorithm and scaling factor according to the channel characteristics of the target network platform to obtain a pre-scaled image, and obtain the DCT (Discrete Cosine Transform) coefficient value corresponding to the pre-scaled image.

[0042] The target network platform here can be understood as the network platform selected by the communication user when transmitting pictures, such as some social network platforms. The channel characteristics here can be understood as the influence and changes that the signal is subjected to when it is transmitted in the channel of the target network platform during the network communication process.

[0043] Furthermore, the carrier image decoding operation in the embodiment of the present application includes but is not limited to operations such as entropy decoding processing, inverse quantization processing, spatial rounding processing, and truncation processing.

[0044] Step S302: embed the target secret information into the DCT coefficients to obtain the initial encrypted DCT coefficients, and adjust the unstable coefficients in the initial encrypted DCT coefficients to obtain the final encrypted DCT coefficients.

[0045] In some other embodiments, in addition to preprocessing the carrier image, the present application also needs to process the target secret information loaded into the carrier image. Specifically, the target secret information can be embedded into the obtained DCT coefficients to obtain the initial encrypted DCT coefficients.

[0046] Furthermore, considering that there may be some unstable coefficients in the initial encrypted DCT coefficients, in order to ensure the security of the final transmitted image, the embodiment of the present application can also adjust the unstable coefficients in the initial encrypted image, thereby obtaining the final encrypted DCT coefficients required by the embodiment of the present application.

[0047] For example, the present application may record the DCT coefficient obtained in step S301 as D s Then, the present application can use the message embedding module based on jitter modulation to embed the target secret message into the DCT coefficient D of the pre-scaled image. s The initial dense DCT coefficient D is obtained y Finally, the coefficient adjustment module based on analog channel processing is used to adjust D y The unstable coefficients in are further adjusted to obtain the final dense DCT coefficients required by the embodiment of the present application.

[0048] This process is further explained below.

[0049] Optionally, in one embodiment of the present application, the target secret information is embedded into the DCT coefficients to obtain initial encrypted DCT coefficients, including: performing stego quantization processing on the DCT coefficients to generate a stego carrier; embedding the target secret information into the stego carrier to obtain a encrypted sequence; modifying the DCT coefficients according to the carrier sequence and the encrypted sequence of the stego carrier to obtain initial encrypted DCT coefficients.

[0050] Based on the relevant descriptions of other embodiments, it can be understood that the present application can embed the target secret information into the DCT coefficients, thereby obtaining the initial encrypted DCT coefficients. For example, the target secret message is embedded into the DCT coefficients D of the pre-scaled image using a message embedding module based on jitter modulation. s The initial dense DCT coefficient D is obtained y .

[0051] Among them, the message embedding module based on jitter modulation in the embodiment of the present application mainly but not limited to includes three steps: stego quantization, STC encoding and jitter modulation.

[0052] Specifically, the embodiment of the present application can first perform stego quantization processing on the DCT coefficients to generate a stego carrier, and then embed the target secret information into the stego carrier to obtain a stego sequence. Finally, the DCT coefficients can be modified according to the carrier sequence and the stego sequence of the stego carrier to obtain the initial stego DCT coefficients.

[0053] For example, since the loss caused by the quantization operation in the JPEG recompression operation is strongly correlated with the quantization step size, the quantization table Q used in the JPEG recompression operation of the channel of the target network platform can be used. c The coefficients at the corresponding positions in the 8*8 DCT block are adaptively quantized with the quantization step size in . The obtained robust stegoscopy vector is represented by X. The process can be expressed as follows:

[0054]

[0055] Combining the JPEG image scaling and recompression process, it can be known that the steganographic carrier is equal to the QDCT coefficient of the image obtained after the original carrier image is channel-scaled and recompressed.

[0056] After obtaining the stego carrier X, the embodiment of the present application can then use the distortion cost function to calculate the modification cost ρ of the elements in the stego carrier X, and then use the ternary STC embedding to adaptively embed the secret message M into the robust stego carrier X to obtain the encrypted sequence Y. The process can be expressed as follows:

[0057] Y=STC(X,ρ,M) (2)

[0058] Among them, when using the STC framework to perform adaptive embedding in the robust steganographic carrier X, the modification amplitude of each element in X can be set to +1, -1 or 0, etc.

[0059] Finally, according to the change from carrier sequence X to encrypted sequence Y, the DCT coefficient D of the spatial domain image after pre-scaling of the carrier image is s Modify it, and record the modified DCT coefficient, that is, the initial encrypted DCT coefficient as D y , the modified goal is to make D y The element corresponding to X in can be equal to Y after steganoquantization, that is, it satisfies the constraint shown in the formula:

[0060]

[0061] Furthermore, in order to ensure the security and imperceptibility of the secret image, the embodiment of the present application may also choose to use a fine dithering modulation scheme to modulate the D s The element d in s Revise. Figure 4 FIG. 1 is a schematic diagram of fine jitter modulation according to an embodiment of the present application. Figure 4 As shown, in the embodiment of the present application, the length q can be set according to the distance t from the boundary. c The interval is divided into two sub-intervals, A and B. The interval closer to the boundary and farther from the midpoint is the A interval, also known as the non-robust interval; the interval farther from the boundary and closer to the midpoint is the B interval, also known as the robust interval. When performing jitter modulation, the embodiment of the present application may not directly convert d s Instead of modifying it to the midpoint of the interval, it is modified to the distance d in the robust interval B. s The nearest point. That is, when modifying, only the position where x is modified by "+1" or "-1" is jitter modulated, and the position that has not changed after STC coding is not modified. It should be noted that even the elements in the B interval may not meet the constraints shown in formula (3) after channel processing, and the robustness needs to be further enhanced through subsequent coefficient adjustment.

[0062] It should be noted that, in the embodiment of the present application, before performing STC encoding, it is necessary to calculate the modification cost of each element in X, and then adaptively select the position with the minimum total distortion for information embedding. The embodiment of the present application can combine the modification amplitude of the jitter modulation with the existing distortion function to design the distortion function of the present invention. Specifically, the existing distortion calculation method such as JUNIWARD (a digital steganography technology) can be used first to calculate the initial distortion of each element in X for forward modification and backward modification on the robust carrier and record them as σ respectively. + , σ - , then calculate as follows Figure 4 The modulation method shown in the figure is forward modified and backward modified to obtain the DCT coefficients after forward modification and backward modification and are recorded as The final distortion is defined as the product of the modified amplitude of the jitter modulation and the initial distortion, which can be expressed as follows:

[0063]

[0064] Next, the selection process of the embedding position in the embodiment of the present application is further explained.

[0065] Optionally, in one embodiment of the present application, before embedding the target secret information into the DCT coefficient, it also includes: obtaining quantization step information of a coefficient block composed of DCT coefficients; and determining a target position for embedding the target secret information into the DCT coefficient according to the quantization step information.

[0066] In the actual implementation process, before performing STC encoding, the present application needs to calculate the modification cost of each element in the steganographic carrier, and then adaptively select the position with the smallest total distortion to embed information.

[0067] Specifically, the embodiment of the present application can first obtain a coefficient block composed of DCT coefficients obtained from the spatial domain image of the original carrier image, and then obtain the quantization step information of the coefficient block, so as to determine the target position of the DCT coefficient to be embedded in the target secret information according to the quantization step information.

[0068] For example, the present application may select the embedding domain of the present application based on the quantization step corresponding to the elements in the 8*8 DCT coefficient block.

[0069] In view of the fact that the quantization step size corresponding to the high-frequency and medium-frequency coefficients of the robust steganography method based on jitter modulation is larger, it is more stable before and after JPEG compression, that is, the information embedding in the medium and high frequencies of the image has better robustness, and the robustness of information embedding in the low frequency is poor. Therefore, for the consideration of robustness, this application can choose D s The corresponding quantization step size q c Larger locations for information embedding.

[0070] However, for security reasons, when information is embedded in the robust carrier sequence X, the modification amplitude is "+1" or "-1", and the average modification amplitude of the DCT coefficients can reach q c (calculated according to the modulation to the midpoint of the interval), too large a modification will result in a decrease in image quality and a decrease in anti-detection performance, so the quantization step size q c Smaller locations are better for embedding.

[0071] Considering the robustness and security, this application sets two hyperparameters α and β. α is the threshold, which can be understood as D s The corresponding quantization step size q c Only elements greater than α will be selected as embedding coefficients, and β is the number of elements selected as the embedding domain in the 8*8 DCT block. That is, the embedding domain finally selected in the embodiment of the present application is the β elements with the smallest quantization step size among the elements with a quantization step size greater than α as embedding coefficients.

[0072] It should be noted that the hyperparameters α and β can be obtained by professionals in this field from existing data or simulation experiments according to specific circumstances and actual needs, and are not specifically limited here.

[0073] Optionally, in one embodiment of the present application, the unstable coefficients in the initial dense DCT coefficients are adjusted to obtain the final dense DCT coefficients, including: performing inverse discrete cosine transform processing, spatial intake processing and stage processing on the initial dense DCT coefficients to obtain the corresponding first simulated spatial pixel values; based on the first simulated spatial pixel values, modifying the first spatial pixel values ​​to obtain the second simulated spatial pixel values; obtaining a pre-sent intermediate image that meets the second preset scaling condition according to the second simulated spatial pixel values; performing simulated channel processing on the pre-sent intermediate image to adjust the unstable coefficients in the initial dense DCT coefficients to obtain the final dense DCT coefficients.

[0074] In some embodiments, since there may be unstable coefficients in the initial dense DCT coefficients, that is, the robustness is poor, the present application needs to adjust the unstable coefficients in the initial dense DCT coefficients to obtain final dense DCT coefficients with higher robustness.

[0075] Specifically, the present application can be adjusted through a coefficient adjustment module based on analog channel processing. First, the initial dense DCT coefficient is subjected to inverse discrete cosine transform processing, spatial intake processing and stage processing to obtain the corresponding first simulated spatial domain pixel value, and then the first spatial domain pixel value is modified based on the first simulated spatial domain pixel value to obtain the second simulated spatial domain pixel value. Then, a pre-sent intermediate image that meets the second preset scaling condition is obtained according to the second simulated spatial domain pixel value, and the pre-sent intermediate image is subjected to analog channel processing, thereby adjusting the unstable coefficients in the initial dense DCT coefficient to obtain the final dense DCT coefficient.

[0076] Among them, the coefficient adjustment module based on analog channel processing in the embodiment of the present application mainly but not limited to includes three steps: intermediate image generation, analog channel processing, and coefficient adjustment. The second preset scaling condition here can be understood as modifying the pixel value of the spatial domain image of the original carrier image through the second simulated spatial domain pixel value, and the simulated spatial domain image obtained after scaling is equal to the spatial pixel value corresponding to the initial carrier dense DCT coefficient after dithering modulation.

[0077] For example, the initial DCT coefficients after dithering modulation can be firstly y Perform inverse DCT transformation, spatial rounding and truncation to obtain the corresponding first simulated spatial pixel value R ′ m ; Then use the pixel value modification module based on inverse interpolation and the first simulated spatial domain pixel value R ′ m The pixel value S of the spatial domain image of the original carrier image u Modify to obtain the second simulated airspace pixel value so that the simulated airspace image S corresponding to the modified second simulated airspace value′ m The spatial pixel value after scaling is exactly equal to the initial dense DCT coefficient D after dither modulation. y The corresponding spatial pixel value R ′ m ; Then, the modified simulated spatial domain image is recompressed using the quality factor corresponding to the original carrier image to obtain a pre-sent intermediate image, and then the obtained pre-sent image is scaled and recompressed using the same processing parameters as the target channel to obtain a pre-received image.

[0078] After the generated intermediate image is processed through the analog channel, the QDCT (Quaternion Discrete Cosine Transform) coefficients Y of the pre-received image in the embedded domain are compared. ′ Is it equal to the element in the encrypted sequence Y? If so, no modification is made. For unequal positions, if the corresponding element d ′ y In the non-robust region A ′ Then, d ′ y Adjust to the robust zone B away from d ′ y The nearest position; if d ′ y In the robust region B ′ In the ′ y Move one position toward the midpoint of the interval. Because the coefficients in the robust region are adjusted with a step size of 1 during coefficient adjustment, the adjustment amplitude is small, and the adjusted coefficients may still not meet the robustness requirements. In addition, the impact of spatial rounding will also change after each coefficient modification. Therefore, multiple rounds of coefficient adjustment are required to ensure the robustness of the final dense DCT coefficients. The number of coefficient adjustment processing rounds is denoted as γ, and the process can be expressed as follows:

[0079]

[0080] Where S(·) is the sign function.

[0081] Step S303, transform the final dense DCT coefficients into the spatial domain to obtain the corresponding second spatial domain pixel values, and modify the first spatial domain pixel values ​​according to the second spatial domain pixel values, until the spatial domain image corresponding to the modified first spatial domain pixel values ​​meets the first preset scaling condition, and compress the spatial domain image corresponding to the modified first spatial domain pixel values ​​according to the quality factor of the carrier image to generate a final dense image that meets the preset size.

[0082] As a possible implementation method, after obtaining the final dense DCT coefficient, the present application can transform the final dense DCT coefficient into the spatial domain to obtain the corresponding second spatial domain pixel value, and modify the first spatial domain pixel value according to the second spatial domain pixel value until the spatial domain image corresponding to the modified first spatial domain pixel value meets the first preset scaling condition, and compress the large-size spatial domain image corresponding to the modified first spatial domain pixel value according to the quality factor of the carrier image to generate a final dense image that meets the preset size.

[0083] The first preset scaling condition can be understood as the spatial image corresponding to the modified first spatial pixel value is exactly equal to the spatial scaled image of the classified image after scaling. The preset size can be understood as the size that meets the requirements of the target network platform for sending classified images.

[0084] The secret image generated by this method has the same quality factor and size as the original carrier image, and can resist scaling and JPEG recompression attacks in lossy social platform channels. After the generated secret image is transmitted through the communication channel of the social platform containing interpolation scaling and JPEG recompression, the receiver will receive the scaled and recompressed secret image. When extracting the secret message M, the received image can first be entropy decoded to obtain the QDCT coefficient D of the received image. r Then, follow the same embedding domain as when embedding the message in D r Select the encryption coefficients in the decoder, which together constitute the encryption carrier for extracting the secret message. Finally, the secret message M can be extracted using the STC decoding algorithm.

[0085] Optionally, in one embodiment of the present application, the first spatial domain pixel value is modified according to the second spatial domain pixel value, including: using the target distortion function to calculate the modification cost of each pixel in the spatial domain image; calculating the interpolation block corresponding to each element in the second spatial domain pixel value on the spatial domain image, and obtaining the weight value of each element in the interpolation block; sorting the weight values ​​according to the modification cost to generate a modification sequence; based on the modification sequence, modifying the first spatial domain pixel value according to the second spatial domain pixel value.

[0086] Based on the relevant descriptions of other embodiments, in order to ensure the robustness of the final dense DCT coefficients, analog processing is adopted, using a coefficient adjustment module based on analog channel processing and a pixel value modification module based on inverse interpolation. Among them, the main function of the pixel value modification module based on inverse interpolation is to modify the first spatial domain pixel value according to the second spatial domain pixel value, that is, to modify the pixel value S of the spatial domain image corresponding to the original carrier image. u Modify so that the modified spatial image S m After scaling, it is exactly equal to the spatial pixel value R corresponding to the final dense DCT coefficient after dither modulation.m .

[0087] Specifically, the present application can first use the distortion function to calculate S u The modification cost of each pixel in R m Every element R m (u,v), calculate its corresponding u The interpolation block C on the interpolation block C and the weight value W corresponding to each element in the interpolation block C; then sort the elements in C in the modification order according to the calculated modification cost, the elements with small modification cost are modified in the first order, and the elements with large modification cost are modified in the last order; then modify the elements in C in the modification order, according to the original scaled image pixel value R u (u,v) and the modified scaled image pixel value R m The difference between (u,v) and the original spatial image S u Make modifications, and the calculation formula of the modification process can be expressed as follows:

[0088] d=S(R m (u,v)-R u (u,v)) (6)

[0089] C ′ (x,y)=C(x,y)+d (7)

[0090] Where (x, y) is the position of the pixel to be modified in the interpolation block. m (u,v) and R u The difference between (u, v) determines the modification direction. At the same time, the sign function can be used to limit the modification step to 1. After modifying each element in the interpolation block, the modified interpolation block is checked to determine whether the current interpolation block is exactly equal to R after calculation. m (u,v), that is, check whether the pixel value in the interpolation block satisfies the following formula:

[0091]

[0092] If it is satisfied, the current interpolation block modification ends, otherwise the modification continues. m After all the interpolation blocks corresponding to the elements in are modified, the modified spatial image S m After scaling, it is exactly equal to R m .

[0093] According to the robust steganography method for images with high quality factors and large sizes proposed in the embodiment of the present application, the pre-scaled image of the original carrier image can be used as a medium, and the secret message can be embedded based on the jitter modulation algorithm combined with the adaptive steganography framework, and the unstable DCT coefficients can be adjusted by analog channel processing, and finally a large-size, high-quality secret image can be generated by inverse interpolation processing. As a result, the generated secret image has the same quality factor and size as the original carrier image, and can perform covert communication behavior, and can resist scaling and JPEG recompression attacks in lossy social platform channels, and correctly extract the secret information in the secret image. As a result, the image steganography algorithm in the related art does not consider the lossy processing in the channel, and the generated secret image often cannot correctly extract the secret message, or only considers the single lossy processing of scaling or recompression. The monitoring party can easily observe the quality factor and size of the image uploaded by the user, and then distinguish between communication behavior and mass behavior. The scope of use is relatively limited and may lead to a series of behavioral security issues.

[0094] Next, a robust steganographic device for images with high quality factors and large sizes proposed according to an embodiment of the present application is described with reference to the accompanying drawings.

[0095] Figure 5 It is a structural schematic diagram of a robust steganographic device for images with high quality factors and large sizes according to an embodiment of the present application.

[0096] like Figure 5 As shown, the robust steganographic device 10 for images with high quality factors and large sizes includes: a first processing module 100 , a second processing module 200 and a steganographic module 300 .

[0097] The first processing module 100 is used to decode the carrier image to obtain a first spatial domain pixel value of the carrier image, and based on the channel characteristics of the target network platform, pre-scale the spatial domain image corresponding to the carrier image to obtain a pre-scaled spatial domain image, and obtain a discrete cosine transform DCT coefficient of the pre-scaled spatial domain image;

[0098] The second processing module 200 is used to embed the target secret information into the DCT coefficients to obtain the initial encrypted DCT coefficients, and adjust the unstable coefficients in the initial encrypted DCT coefficients to obtain the final encrypted DCT coefficients;

[0099] The steganographic module 300 is used to transform the final steganographic DCT coefficients into the spatial domain to obtain the corresponding second spatial domain pixel values, and to modify the first spatial domain pixel values ​​according to the second spatial domain pixel values ​​until the spatial domain image corresponding to the modified first spatial domain pixel values ​​meets the first preset scaling condition, and to compress the spatial domain image corresponding to the modified first spatial domain pixel values ​​according to the quality factor of the carrier image to generate a final steganographic image that meets the preset size.

[0100] Optionally, in one embodiment of the present application, the second processing module 200 includes: a first processing unit, an embedding unit and a first modifying unit.

[0101] The first processing unit is used to perform stego quantization processing on the DCT coefficients to generate a stego carrier;

[0102] An embedding unit, used for embedding the target secret information into the stego carrier to obtain a stego sequence;

[0103] The first modification unit is used to modify the DCT coefficients according to the carrier sequence and the encrypted sequence of the steganographic carrier to obtain initial encrypted DCT coefficients.

[0104] Optionally, in one embodiment of the present application, it also includes: an acquisition module and a determination module.

[0105] The acquisition module acquires quantization step information of a coefficient block composed of DCT coefficients before embedding the target secret information into the DCT coefficients;

[0106] The determination module is used to determine the target position where the target secret information is embedded into the DCT coefficient according to the quantization step information.

[0107] Optionally, in one embodiment of the present application, the second processing module 200 includes: a second processing unit, a second modifying unit, a generating unit and an adjusting unit.

[0108] The second processing unit is used to perform inverse discrete cosine transform processing, spatial intake processing and stage processing on the initial dense DCT coefficients to obtain corresponding first simulated spatial domain pixel values;

[0109] A second modifying unit, configured to modify the first spatial domain pixel value based on the first simulated spatial domain pixel value to obtain a second simulated spatial domain pixel value;

[0110] A generating unit, configured to obtain a pre-sent intermediate picture satisfying a second preset scaling condition according to a second simulated spatial domain pixel value;

[0111] The adjustment unit is used to perform analog channel processing on the pre-sent intermediate picture to adjust the unstable coefficients in the initial dense DCT coefficients to obtain the final dense DCT coefficients.

[0112] Optionally, in one embodiment of the present application, the steganography module 300 includes: a first calculation unit, a second calculation unit, a sorting unit and a third modification unit.

[0113] Wherein, the first calculation unit is used to calculate the modification cost of each pixel in the spatial domain image using the target distortion function;

[0114] A second calculation unit is used to calculate the interpolation block corresponding to each element in the second spatial domain pixel value on the spatial domain image, and obtain a weight value of each element in the interpolation block;

[0115] A sorting unit, used for sorting the weight values ​​according to the modification costs to generate a modification sequence;

[0116] The third modification unit is used to modify the first spatial domain pixel value according to the second spatial domain pixel value based on the modification sequence.

[0117] It should be noted that the aforementioned explanation of the embodiment of the robust steganography method for images with high quality factors and large sizes is also applicable to the robust steganography device for images with high quality factors and large sizes in this embodiment, and will not be repeated here.

[0118] According to the robust steganographic device for images with high quality factors and large sizes proposed in the embodiment of the present application, the pre-scaled image of the original carrier image can be used as a medium, and the secret message can be embedded based on the jitter modulation algorithm combined with the adaptive steganographic framework, and the unstable DCT coefficients can be adjusted by analog channel processing, and finally a large-size, high-quality secret image can be generated by inverse interpolation processing. As a result, the generated secret image has the same quality factor and size as the original carrier image, and can perform covert communication behavior, and can resist scaling and JPEG recompression attacks in lossy social platform channels, and correctly extract the secret information in the secret image. As a result, the image steganographic algorithm in the related art does not consider the lossy processing in the channel, and the generated secret image often cannot correctly extract the secret message, or only considers the single lossy processing of scaling or recompression. The monitoring party can easily observe the quality factor and size of the image uploaded by the user, and then distinguish between communication behavior and mass behavior. The scope of use is relatively limited and may lead to a series of behavioral security issues.

[0119] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0120] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0121] When the processor 602 executes the program, the robust steganographic method for images with high quality factors and large sizes provided in the above embodiments is implemented.

[0122] Furthermore, the electronic device further comprises:

[0123] The communication interface 603 is used for communication between the memory 601 and the processor 602 .

[0124] The memory 601 is used to store computer programs that can be executed on the processor 602 .

[0125] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0126] If the memory 601, the processor 602 and the communication interface 603 are implemented independently, the communication interface 603, the memory 601 and the processor 602 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0127] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 603 are integrated on a chip, the memory 601, the processor 602 and the communication interface 603 can communicate with each other through an internal interface.

[0128] The processor 602 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0129] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned robust steganographic method for images with high quality factors and large sizes.

[0130] An embodiment of the present application also provides a computer program product, including a computer program, which can run computer instructions. When the computer instructions are executed by a processor, the robust steganography method for images with high quality factors and large sizes provided by the embodiment of the present application is implemented.

[0131] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0132] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0133] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0134] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.

[0135] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented by hardware, as in another embodiment, it can be implemented by any one or a combination of multiple of the following technologies known in the art: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0136] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0137] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0138] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A robust steganographic method for images with high quality factors and large sizes, characterized in that The following steps are involved: Decoding the carrier image to obtain a first spatial domain pixel value of the carrier image, and pre-scaling the spatial domain image corresponding to the carrier image based on the channel characteristics of the target network platform to obtain a pre-scaled spatial domain image, and obtaining a discrete cosine transform DCT coefficient of the pre-scaled spatial domain image; Embedding target secret information into the DCT coefficients to obtain initial encrypted DCT coefficients, and adjusting unstable coefficients in the initial encrypted DCT coefficients to obtain final encrypted DCT coefficients; The final dense DCT coefficient is transformed into the spatial domain to obtain a corresponding second spatial domain pixel value, and the first spatial domain pixel value is modified according to the second spatial domain pixel value until the spatial domain image corresponding to the modified first spatial domain pixel value meets the first preset scaling condition, and the spatial domain image corresponding to the modified first spatial domain pixel value is compressed according to the quality factor of the carrier image to generate a final dense image that meets the preset size.

2. The method according to claim 1, characterized in that The step of embedding the target secret information into the DCT coefficients to obtain initial encrypted DCT coefficients includes: Performing stego quantization processing on the DCT coefficients to generate a stego carrier; Embedding the target secret information into the stego carrier to obtain a stego sequence; The DCT coefficients are modified according to the carrier sequence of the steganographic carrier and the encrypted sequence to obtain the initial encrypted DCT coefficients.

3. The method according to claim 1, characterized in that Before embedding the target secret information into the DCT coefficient, the method further includes: Obtaining quantization step information of a coefficient block composed of the DCT coefficients; The target position where the target secret information is embedded in the DCT coefficient is determined according to the quantization step information.

4. The method according to claim 1, characterized in that The step of adjusting the unstable coefficients in the initial dense DCT coefficients to obtain the final dense DCT coefficients includes: Performing inverse discrete cosine transform processing, spatial intake processing and stage processing on the initial dense DCT coefficients to obtain corresponding first simulated spatial domain pixel values; Based on the first simulated spatial domain pixel value, modify the first spatial domain pixel value to obtain a second simulated spatial domain pixel value; Obtaining a pre-sent intermediate picture that meets a second preset scaling condition according to the second simulated spatial domain pixel value; The pre-sent intermediate picture is subjected to analog channel processing to adjust unstable coefficients in the initial dense DCT coefficients to obtain final dense DCT coefficients.

5. The method according to claim 1, characterized in that: The modifying the first spatial domain pixel value according to the second spatial domain pixel value comprises: Calculating the modification cost of each pixel in the spatial domain image using a target distortion function; Calculate the interpolation block corresponding to each element in the second spatial domain pixel value on the spatial domain image, and obtain the weight value of each element in the interpolation block; sorting the weight values ​​according to the modification costs to generate a modification sequence; Based on the modification sequence, the first spatial domain pixel value is modified according to the second spatial domain pixel value.

6. A robust steganographic device for images with high quality factor and large size, characterized in that include: A first processing module is used to decode the carrier image to obtain a first spatial domain pixel value of the carrier image, and based on the channel characteristics of the target network platform, pre-scale the spatial domain image corresponding to the carrier image to obtain a pre-scaled spatial domain image, and obtain a discrete cosine transform DCT coefficient of the pre-scaled spatial domain image; A second processing module is used to embed the target secret information into the DCT coefficient to obtain an initial encrypted DCT coefficient, and adjust the unstable coefficient in the initial encrypted DCT coefficient to obtain a final encrypted DCT coefficient; The steganographic module is used to transform the final steganographic DCT coefficient into the spatial domain to obtain the corresponding second spatial domain pixel value, and modify the first spatial domain pixel value according to the second spatial domain pixel value until the spatial domain image corresponding to the modified first spatial domain pixel value meets the first preset scaling condition, and compress the spatial domain image corresponding to the modified first spatial domain pixel value according to the quality factor of the carrier image to generate a final steganographic image that meets the preset size.

7. The device according to claim 6, characterized in that The second processing module comprises: A processing unit, configured to perform a stego quantization process on the DCT coefficients to generate a stego carrier; An embedding unit, used for embedding the target secret information into the stego carrier to obtain a stego sequence; The modification unit is used to modify the DCT coefficients according to the carrier sequence of the steganographic carrier and the encrypted sequence to obtain the initial encrypted DCT coefficients.

8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the robust steganography method for images with high quality factors and large sizes as described in any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the robust steganography method for images with high quality factors and large sizes as described in any one of claims 1 to 5.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed, it is used to implement the robust steganography method for images with high quality factors and large sizes as described in any one of claims 1 to 5.

Images