Image Steganography Scheme Research Articles

Sterilization of image steganography using self-supervised convolutional neural network

Background With the development of steganography technology, lawbreakers can implement covert communication in social networks more easily, exacerbating network security risks. Sterilization of image steganography methods can eliminate secret messages to block the transmission of illegal covert communication. However, existing methods overly rely on cover-stego image pairs and are unable to sanitize unknown image, which reduces stego image blocking rate in social networks. Methods To address the above problems, this paper proposes an effective sterilization of image steganography method using self-supervised convolutional neural network (SS-Net), which does not require any prior knowledge of image steganography schemes. SS-Net includes a purification module and a refinement module. Firstly, the pixel-shuffle down-sampling in purification module is adopted to reduce the spatial correlation of pixels in the stgeo image, and improve the learning mode from supervised learning to self-supervised learning. Secondly, centrally masked convolutions and dilated convolution residual blocks are merged to eliminate secret messages and avoid image quality degradation. Finally, a refinement module is employed to improve image texture details and boundaries. Results A series of experiments show that SS-Net from BOSSbase test sets is able to balance the destruction of secret messages with image quality, achieving 100% blocking rate of stego image. Meanwhile, our method outperforms the state-of-the-art methods in secret messages elimination ability and image quality preserving ability.

A novel approach for the enhancement of payload capacity in pixel value differencing image steganography schemes

A novel approach for the enhancement of payload capacity in pixel value differencing image steganography schemes

Domain Transformation of Distortion Costs for Efficient JPEG Steganography with Symmetric Embedding

Nowadays, most image steganographic schemes embed secret messages by minimizing a well-designed distortion cost function for the corresponding domain, i.e., the spatial domain for spatial image steganography or the JPEG (Joint Photographic Experts Group) domain for JPEG image steganography. In this paper, we break the boundary between these two types of schemes by establishing a theoretical link between the distortion costs in the spatial domain and those in the JPEG domain and thus propose a scheme for domain transformations of distortion costs for efficient JPEG steganography with symmetric embedding, which can directly convert the spatial distortion cost into its JPEG counterpart. Specifically, by formulating the distortion cost function for JPEG images in the decompressed spatial domain, a closed-form expression for a distortion cost cross-domain transformation is derived theoretically, which precisely characterizes the conversion from the distortion costs obtained by existing spatial steganographic schemes to those applied in JPEG steganography. Experimental results demonstrate that the proposed method outperforms other advanced JPEG steganographic schemes, e.g., JUNIWARD (JPEG steganography with Universal Wavelet Relative Distortion), JMiPOD (JPEG steganography by Minimizing the Power of the Optimal Detector), and DCDT (Distortion Cost Domain Transformation), in resisting the detection of various advanced steganalyzers.

StegaStyleGAN: Towards Generic and Practical Generative Image Steganography

The recent advances in generative image steganography have drawn increasing attention due to their potential for provable security and bulk embedding capacity. However, existing generative steganographic schemes are usually tailored for specific tasks and are hardly applied to applications with practical constraints. To address this issue, this paper proposes a generic generative image steganography scheme called Steganography StyleGAN (StegaStyleGAN) that meets the practical objectives of security, capacity, and robustness within the same framework. In StegaStyleGAN, a novel Distribution-Preserving Secret Data Modulator (DP-SDM) is used to achieve provably secure generative image steganography by preserving the data distribution of the model inputs. Additionally, a generic and efficient Secret Data Extractor (SDE) is invented for accurate secret data extraction. By choosing whether to incorporate the Image Attack Simulator (IAS) during the training process, one can obtain two models with different parameters but the same structure (both generator and extractor) for lossless and lossy channel covert communication, namely StegaStyleGAN-Ls and StegaStyleGAN-Ly. Furthermore, by mating with GAN inversion, conditional generative steganography can be achieved as well. Experimental results demonstrate that, whether for lossless or lossy communication channels, the proposed StegaStyleGAN can significantly outperform the corresponding state-of-the-art schemes.

A High-performance Image Steganography Scheme Based on Dual-adversarial Networks

A High-performance Image Steganography Scheme Based on Dual-adversarial Networks

Image steganography techniques for resisting statistical steganalysis attacks: A systematic literature review.

Information hiding in images has gained popularity. As image steganography gains relevance, techniques for detecting hidden messages have emerged. Statistical steganalysis mechanisms detect the presence of hidden secret messages in images, rendering images a prime target for cyber-attacks. Also, studies examining image steganography techniques are limited. This paper aims to fill the existing gap in extant literature on image steganography schemes capable of resisting statistical steganalysis attacks, by providing a comprehensive systematic literature review. This will ensure image steganography researchers and data protection practitioners are updated on current trends in information security assurance mechanisms. The study sampled 125 articles from ACM Digital Library, IEEE Explore, Science Direct, and Wiley. Using PRISMA, articles were synthesized and analyzed using quantitative and qualitative methods. A comprehensive discussion on image steganography techniques in terms of their robustness against well-known universal statistical steganalysis attacks including Regular-Singular (RS) and Chi-Square (X2) are provided. Trends in publication, techniques and methods, performance evaluation metrics, and security impacts were discussed. Extensive comparisons were drawn among existing techniques to evaluate their merits and limitations. It was observed that Generative Adversarial Networks dominate image steganography techniques and have become the preferred method by scholars within the domain. Artificial intelligence-powered algorithms including Machine Learning, Deep Learning, Convolutional Neural Networks, and Genetic Algorithms are recently dominating image steganography research as they enhance security. The implication is that previously preferred traditional techniques such as LSB algorithms are receiving less attention. Future Research may consider emerging technologies like blockchain technology, artificial neural networks, and biometric and facial recognition technologies to improve the robustness and security capabilities of image steganography applications.

Image Steganography: A Review

Abstract: An image steganographic scheme is one kind of steganographic systems, where the secret message is hidden in a digital image with some hiding method. Someone can then use a proper embedding procedure to recover the hidden message from the image. The original image is called a cover image in steganography, and the message-embedded image is called a stego image. This Steganography technique is more popular in recent year than other steganography possibly because of the flood of electronic image information available with the advent of digital cameras and high-speed internet distribution. It can involve hiding information in the naturally occurred noise within the image. In this paper we provide a review on various image steganography techniques.

3D image steganography using cellular automata transform and depth estimation network

Traditional 3D image steganography is cumbersome and difficult owing to the special structure of 3D image, imposing a practical limit for its applications. Herein, we propose a 3D image steganography scheme utilizing cellular automata transformation (CAT). By using CAT with various rule numbers to hide images, the weakness of previous algorithms with only one transformation plane is overcome, and the hidden image quality can be improved by reducing energy loss. In the 3D reconstruction process, we apply a depth estimation algorithm based on a convolutional network to obtain corresponding depth maps from a single RGB image, simplifying the elemental image array (EIA) pickup process and even overcoming the phase difference problem caused by the traditional optical pickup process using lens arrays. Before hiding, we use the maximum length cellular automata (MLCA) to encrypt a single RGB image instead of encrypting EIA images with huge data, which greatly reduces the complexity of the encryption algorithm and improves efficiency. Finally, we reconstruct the decrypted image optically and obtain a full color and full parallax 3D image through the display device.

Deep learning approach and cover image transportation: a multi-security adaptive image steganography scheme

ABSTRACT In this era of information security and communication, a major priority is the achievement of a robust and secure steganography system when thinking about information concealment. The development of such an information-hiding scheme demands that the scheme be able to hide a secret message within the cover media. The most vexing issues in existing steganography protocols are imperceptibility, security, and capacity, and researchers have frequently emphasized a trade-off between these issues. Scholars have consistently ignored the balance between security and payload because resolving one problem has been shown to have an impact on the other, and vice versa. To overcome these problems, an effective method known as the Conventional Neural Network based Edge Detection Method (CNN-EDM) has been presented for image steganography in this study. The CNN-EDM is used to improve the contributions of the proposed scheme. Four main stages were used to achieve the objectives in this research, beginning with the cover image and secret image preparation, followed by embedding, and culminating in extraction. The last stage is the evaluation stage, which employs several evaluations to benchmark the obtained results. A standard database from the Signal and Image Processing Institute (SIPI) containing color and grayscale images with 512 × 512 pixels was utilized in this study. Different parameters were used to test the performance of the suggested scheme based on security and imperceptibility (image quality). The image quality was evaluated using three important metrics: histogram analysis, peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM). Furthermore, two metrics were used to evaluate the security properties of the proposed system: the Human Visual System (HVS) and Chi-square (X2) attacks. The evaluations showed that the proposed scheme can enhance the capacity, invisibility, and security features and address the already existing problems in this domain.

Optical steganography with RVNCA-based keys for 3D object

Traditional 3D image steganography is cumbersome and difficult owing to the special structure of 3D image, imposing a practical limit for its applications. Herein, we propose a 3D image steganography scheme utilizing computational ghost imaging (CGI). Furthermore, triple keys are introduced in the steganography to increase the difficulty of being cracked greatly. With the help of backward ray tracing (BRT) technology, the element image array (EIA) of the 3D object is obtained as the target of ghost imaging. For encryption, we utilize restricted vertical neighborhood cellular automata (RVNCA) to encode random masks of the CGI system as the key, which greatly improves the security of the steganography scheme and facilitates the management and transmission of the key. Finally, we reconstruct the decrypted image optically and obtain a full color and full parallax 3D image through the display device.

An RGB Color Image Steganography Scheme by Binary Lower Triangular Matrix

Steganography is an area where researchers focus on how to develop novel hiding techniques that are aided by the mathematical or computational model. In order to conceal the secret bits in a cover image, the imperceptible property and embedding payload capacity must be maintained. In most of the developed methods, some selective least significant bits (LSBs) of each pixel of the cover image is modified to hide the data which may lead to substantial distortion in the stego-image especially when the payload size is large. Therefore, to attain the trade-off between payload capacity and image quality, the authors have devised a binary lower triangular matrix-based secret message embedding process in which data can be hidden in the color pixel. In this secret message embedding process, the secret data is embedded aided by the binary lower triangular matrix. The formulated mathematical model for the secret message embedding process has ensured a secure embedding technique with reduced aberration on the stego-image. The proposed steganography scheme is preserving significant visual quality while emphasizing the security aspect by adopting an indirect data-hiding mechanism in contrast to the conventional mechanism. Furthermore, for the betterment of the visual properties of the stego-image, we have generically extended the formulation and tested it to check the visual quality. The empirical results of the proposed steganography scheme depict outcomes satisfactorily for the RGB color images. Researchers can explore and apply this novel data-hiding scheme in different data communication applications.

A secure controlled quantum image steganography scheme based on the multi-channel effective quantum image representation model

A secure controlled quantum image steganography scheme based on the multi-channel effective quantum image representation model

An efficient image steganographic scheme for a real-time embedded system and its hardware implementation on AMD Xilinx Zynq-7000 APSoC platform

An efficient image steganographic scheme for a real-time embedded system and its hardware implementation on AMD Xilinx Zynq-7000 APSoC platform

Steganographic secret sharing via AI-generated photorealistic images

Steganographic secret sharing is an access control technique that transforms a secret message into multiple shares in a steganographic sense. Each share is in a human-readable format in order to dispel suspicion from a malicious party during transmission and storage. Such a human-readable format can also serve to facilitate data management. The secret can be reconstructed only when a sufficient number of authorized shareholders collaborate. In this study, we use neural networks to encode secret shares into photorealistic image shares. This approach is conceptually related to coverless image steganography in which the data are transformed directly into an image rather than concealed into a cover image. We further implement an authentication mechanism to verify the integrity of the image shares presented in the decoding phase. All coverless image steganography schemes can be used to achieve steganographic secret sharing, but our detection mechanism can further improve the robustness of these schemes. Experimental results confirm the robustness of the proposed scheme against various steganalysis and tampering attacks.

Identification of Content-Adaptive Image Steganography Using Convolutional Neural Network Guided by High-Pass Kernel

Digital images are very popular and commonly used for hiding crucial data. In a few instances, image steganography is misused for communicating with improper data. In this paper, a robust deep neural network is proposed for the identification of content-adaptive image steganography schemes. Multiple novel strategies are applied to improve detection performance. Two non-trainable convolutional layers is used to guide the proposed CNN with fixed kernels. Thirty-one kernels are used in both non-trainable layers, of which thirty are high-pass kernels and one is the neutral kernel. The layer-specific learning rate is applied for each layer. ReLU with customized thresholding is applied to achieve better performance. In the proposed method, image down-sampling is not performed; only the global average pooling layer is considered in the last part of the network. The experimental results are verified on BOWS2 and BOSSBase image sets. Content-adaptive steganography schemes, such as HILL, Mi-POD, S-UNIWARD, and WOW, are considered for generating the stego images with different payloads. In experimental analysis, the proposed scheme is compared with some of the latest schemes, where the proposed scheme outperforms other state-of-the-art techniques in the most cases.

High-capacity coverless image steganographic scheme based on image synthesis

In this paper, a coverless image steganographic scheme with high capacity is proposed. It can hide a color secret image into a color stego image with the same size. Furthermore, the stego image is synthesized without the aid of any cover images. Since there are no cover modification operations, it reduces the risk of being detected by steganalysis tools. The proposed scheme comprises a stego image synthesis sub-network (SynthesisNet) and a secret image recovery sub-network (RevealNet). In SynthesisNet, we apply the image-to-image translation to implement cross-domain image synthesis. It accepts source and reference images as the inputs, which can be used to customize the semantic and style of the stego image. A mapping between source and reference images is also designed to enhance the authenticity of the stego images. Experiments show that the proposed scheme provides high steganographic undetectability as well as good quality on stego and recovered secret images. Moreover, it can generate stego images with specified contents different from any existing images in the public channel.

Quantum Image Steganography Schemes for Data Hiding: A Survey

Quantum steganography plays a critical role in embedding confidential data into carrier messages using quantum computing schemes. The quantum variant of steganography outperforms its classical counterpart from security, embedding efficiency and capacity, imperceptibility, and time-complexity perspectives. Considerable work has been carried out in the literature focusing on quantum steganography. However, a holistic view of available schemes is missing. This paper provides an overview of latest advances in the field of quantum-steganography and image-steganography schemes. Moreover, the paper includes discussion of improvements made in the aforementioned fields, a brief explanation of the methodologies used for each presented algorithm, and a comparative study of existing schemes.

Refining Medical Image Steganography Scheme Based on Pixels Disparity Value and Huffman Coding

The steganography is a new and challenging method for transmitting confidential information over the internet utilizing a hidden cover object. Developing an efficient picture steganography system is currently fraught with difficulties because of the limited capacity, weak resilience, and imperceptibility of current steganography systems under development. This problem can only be solved by increasing the picture steganography's capability and security levels. Methods: Based on these variables, this work aims to present the state-of-the-art approach known as the Pixels Disparity Value (PDV) that hides patient information reports in the distinct medical photos for the second party. It's recommended that the procedure be broken down into three primary phases, each with its own set of instructions. Results: The statistical and economic evaluations of the suggested approach.

A novel color image steganography algorithm based on bit counting and multiple-base system

This paper proposes a novel color image steganography scheme using a compound sinusoidal discrete memristor map and bit counting method. Firstly, a new 2D compound sinusoidal discrete memristor map is designed to generate chaotic sequences with good randomness. Secondly, a new steganography algorithm combining bit counting and multiple-base system is proposed to reduce the distortion of the stego image. Finally, a new steganography framework based on the HSV model is designed to improve the security of the steganography scheme. Experimental and test results show that the proposed algorithm has good imperceptibility and can embed a large number of secret information.

An Efficient Image Steganography Scheme Using Bit-plane Slicing with Elliptic Curve Cryptography and Wavelet Transform

Information security is indispensable in the transmission of multimedia data. While accumulating and distributing such multimedia data, the access of data from a third person is the real security challenging issue. Information hiding plays an important role. Scramble the data before hiding it in carrier media gives enhanced security level for the data. In this paper, bit plane slicing is used to represent an input image with eight planes at bit-level instead of pixel-level. As the least significant bit contains noisy information, only the most significant bit plane can be used to represent an image. At the first level, an input image is processed through the spatial domain. Transform domain techniques are used to process the image at the middle level. Elliptic curve cryptography is used to scramble and descramble the MSB plane image. A logistic chaotic sequence of the input image is added to the most significant bit plane image to generate the final scrambled image. The discrete wavelet transform is used to embed the scrambled image in its high-frequency sub-bands. At the last level, a least significant bit technique, a spatial domain is used to embed the scrambled image in the carrier image. Message integrity is also verified by finding the hash of an input image. The performance of the proposed method is evaluated through various security measures. It gives good results as number of pixel change rate is closer to 100% and unified average changing intensity is 33.46.