Quantized Gaussian JPEG Steganography and Pool Steganalysis (original) (raw)
Related papers
In this paper, we introduce a new feature-based steganalytic method for JPEG images and use it as a benchmark for comparing JPEG steganographic algorithms and evaluating their embedding mechanisms. The detection method is a linear classifier trained on feature vectors corresponding to cover and stego images. In contrast to previous blind approaches, the features are calculated as an L 1 norm of the difference between a specific macroscopic functional calculated from the stego image and the same functional obtained from a decompressed, cropped, and recompressed stego image. The functionals are built from marginal and joint statistics of DCT coefficients. Because the features are calculated directly from DCT coefficients, conclusions can be drawn about the impact of embedding modifications on detectability. Three different steganographic paradigms are tested and compared. Experimental results reveal new facts about current steganographic methods for JPEGs and new design principles for more secure JPEG steganography. system is considered broken. For a more exact treatment of the concept of steganographic security, the reader is referred to [1,2].
Pooled Steganalysis in JPEG: how to deal with the spreading strategy?
2019 IEEE International Workshop on Information Forensics and Security (WIFS), 2019
In image pooled steganalysis, a steganalyst, Eve, aims to detect if a set of images sent by a steganographer, Alice, to a receiver, Bob, contains a hidden message. We can reasonably assess that the steganalyst does not know the strategy used to spread the payload across images. To the best of our knowledge, in this case, the most appropriate solution for pooled steganalysis is to use a Single-Image Detector (SID) to estimate/quantify if an image is cover or stego, and to average the scores obtained on the set of images. In such a scenario, where Eve does not know the spreading strategies, we experimentally show that if Eve can discriminate among few well-known spreading strategies, she can improve her steganalysis performances compared to a simple averaging or maximum pooled approach. Our discriminative approach allows obtaining steganalysis efficiencies comparable to those obtained by a clairvoyant, Eve, who knows the Alice spreading strategy. Another interesting observation is that DeLS spreading strategy behaves really better than all the other spreading strategies. Those observations results in the experimentation with six different spreading strategies made on Jpeg images with J-UNIWARD, a state-of-the-art Single-Image-Detector, and a discriminative architecture that is invariant to the individual payload in each image, invariant to the size of the analyzed set of images, and build on a binary detector (for the pooling) that is able to deal with various spreading strategies.
Performance study of common image steganography and steganalysis techniques
Journal of Electronic Imaging, 2006
We investigate the performance of state of the art universal steganalyzers proposed in the literature. These universal steganalyzers are tested against a number of well-known steganographic embedding techniques that operate in both the spatial and transform domains. Our experiments are performed using a large data set of JPEG images obtained by randomly crawling a set of publicly available websites. The image data set is categorized with respect to size, quality, and texture to determine their potential impact on steganalysis performance. To establish a comparative evaluation of techniques, undetectability results are obtained at various embedding rates. In addition to variation in cover image properties, our comparison also takes into consideration different message length definitions and computational complexity issues. Our results indicate that the performance of steganalysis techniques is affected by the JPEG quality factor, and JPEG recompression artifacts serve as a source of confusion for almost all steganalysis techniques. © 2006 SPIE and IS&T.
Effect of JPEG Quality on Steganographic Security
Proceedings of the ACM Workshop on Information Hiding and Multimedia Security, 2019
This work investigates both theoretically and experimentally the security of JPEG steganography as a function of the quality factor. For a fixed relative payload, modern embedding schemes, such as J-UNIWARD and UED-JC, exhibit surprising non-monotone trends due to rounding and clipping of quantization steps. Their security generally increases with increasing quality factor but starts decreasing for qualities above 95. In contrast, old-fashion steganography, such as Jsteg, OutGuess, and model-based steganography, exhibit complementary trends. The results of empirical detectors closely match the trends exhibited by the KL divergence computed between models of cover and stego DCT modes. In particular, our analysis shows that the main reason for the complementary trends is the way modern schemes attenuate embedding change rates with increasing spatial frequency. Our model also provides guidance on how to adjust the embedding algorithm J-UNIWARD to substantially improve its security for high quality factors.
Side-Informed Steganography for JPEG Images by Modeling Decompressed Images
arXiv (Cornell University), 2022
Side-informed steganography has always been among the most secure approaches in the field. However, a majority of existing methods for JPEG images use the side information, here the rounding error, in a heuristic way. For the first time, we show that the usefulness of the rounding error comes from its covariance with the embedding changes. Unfortunately, this covariance between continuous and discrete variables is not analytically available. An estimate of the covariance is proposed, which allows to model steganography as a change in the variance of DCT coefficients. Since steganalysis today is best performed in the spatial domain, we derive a likelihood ratio test to preserve a model of a decompressed JPEG image. The proposed method then bounds the power of this test by minimizing the Kullback-Leibler divergence between the cover and stego distributions. We experimentally demonstrate in two popular datasets that it achieves state-of-the-art performance against deep learning detectors. Moreover, by considering a different pixel variance estimator for images compressed with Quality Factor 100, even greater improvements are obtained.
Quantitative steganalysis of digital images: estimating the secret message length
Multimedia Systems, 2003
The objective of steganalysis is to detect messages hidden in cover objects, such as digital images. In practice, the steganalyst is frequently interested in more than whether or not a secret message is present. The ultimate goal is to extract and decipher the secret message. However, in the absence of the knowledge of the stego technique and the stego and cipher keys, this task may be extremely time consuming or completely infeasible. Therefore, any additional information, such as the message length or its approximate placement in image features, could prove very valuable to the analyst. In this paper, we present general principles for developing steganalytic methods that can accurately estimate the number of changes to the cover image imposed during embedding. Using those principles, we show how to estimate the secret message length for the most common embedding archetypes, including the F5 and OutGuess algorithms for JPEG, EzStego algorithm with random straddling for palette images, and the classical LSB embedding with random straddling for uncompressed image formats. The paper concludes with an outline of ideas for future research such as estimating the steganographic capacity of embedding algorithms.
An Experimental Investigation of Statistical Model based Secure Steganography for JPEG Images
Indian Journal of Science and Technology 10(27):1-11, 2017
Objectives: This paper intends to propose a secure steganography approach in JPEG compressed domain by providing more possibilities to analyzing the DCT coefficients in lower frequency area by modifying the primary Quantization Table (QT) with generating random data hiding patterns. Methods/Statistical analysis: The upper left part of the primary QT extracted from gray scale image dataset is modified by multiplying the factors ¼, ½ and ¾ to produce secondary QTs for investigating randomly generated data hiding patterns in lower frequency area of quantized DCT coefficients by Least Significant Bit (LSB) method. We create a pool of QTs and those tables are cross checked with randomly generated hiding patterns to find best QT with appropriate data hiding pattern by assessing Peak Signal to Noise Ratio (PSNR). Our method can be used to attain trade-off between some parameters such as image features, QT, data hiding pattern. Further, statistical features of a given image data set are extracted and analyzed with the selected QT and appropriate hiding pattern by using R software. Findings: The Experimental results revealed that our proposed method can embed high capacity data (57 bits per block) without noticeable visual artifacts by considering lower frequency coefficients for data hiding by assessing the image steganographic requirements. The maximum PSNR value 48 and the minimum PSNR value 32 are found among the fifty jpeg gray images based on their contents. Although, the embedding capacity and PSNR fluctuate among images, our method can be used to attain trade-off between some parameters such as image features, QT, data hiding pattern. Further, statistical features of a given image data set are extracted and analyzed with the selected QT and appropriate hiding pattern using R. The hypothesis test was deployed among the QTs, hiding patterns and image features. The following five P-Values, 0.04128, 0.02486, 0.02241, 0.04898, 0.01966 less than 0.05 show the good relationship between the QTs, hiding patterns and image features among the cover images and also the following four P-Values, 0.00685, 0.03017, 0.001085, 4.568e-12, less than 0.05, show the good relationship between those above mentioned factors among the stego images. Application/Improvements: Finally, we present a secure model to explore the relationship between QT, hiding pattern and image contents. The found model is stego invariant for sender and receiver that will enable them to identify QT and pattern by extracting the image features without fully decoding the stego jpeg image. This method is very practical and adaptable for extending QTs and hiding patterns.
An Optimal Steganalysis Based Approach for Embedding Information in Image Cover Media with Security
World Academy of Science, Engineering and Technology, International Journal of Electrical, Computer, Energetic, Electronic and Communication Engineering, 2016
This paper deals with the study of interest in the fields of Steganography and Steganalysis. Steganography involves hiding information in a cover media to obtain the stego media in such a way that the cover media is perceived not to have any embedded message for its unintended recipients. Steganalysis is the mechanism of detecting the presence of hidden information in the stego media and it can lead to the prevention of disastrous security incidents. In this paper, we provide a critical review of the steganalysis algorithms available to analyze the characteristics of an image stego media against the corresponding cover media and understand the process of embedding the information and its detection. We anticipate that this paper can also give a clear picture of the current trends in steganography so that we can develop and improvise appropriate steganalysis algorithms. Keywords—Optimization, heuristics and metaheuristics algorithms, embedded systems, low-power consumption, Steganalys...
ANALYSIS OF IMAGE STEGANALYSIS TECHNIQUES TO DEFEND AGAINST STATISTICAL ATTACKS – A SURVEY.pdf
IJRET, 2012
Steganography is the art concealing information to transmit it in such a way that nobody but the intended receiver knows the existence of the message. Steganalysis techniques work on eliminating suspicion about the existence of a message. If suspicion is raised, then the message cannot be passed covertly. One of the ways to detect the hidden message is to view the statistical properties of the image or medium in which the message is hidden. This is called a statistical attack. In this paper, we explain the nature of such attacks and present our conclusions based on reviews of existing methods of defense against statistical attacks.
Benchmarking steganographic and steganalysis techniques
Security, Steganography, and Watermarking of Multimedia Contents VII, 2005
There have been a number of steganography embedding techniques proposed over the past few years. In turn the development of these techniques have led to an increased interest in steganalysis techniques. More specifically Universal steganalysis techniques have become more attractive since they work independently of the embedding technique. In this work, our goal is to compare a number of universal steganalysis techniques proposed in the literature which include techniques based on binary similarity measures, wavelet coefficients' statistics, and DCT based image features. These universal steganalysis techniques are tested against a number of well know embedding techniques, including Outguess , 1 F5 , 2 Model based , 3 and perturbed quantization . 4 Our experiments are done using a large dataset of JPEG images, obtained by randomly crawling a set of publicly available websites. The image dataset is categorized with respect to the size and quality. We benchmark embedding rate versus detectability performances of several widely used embedding as well as universal steganalysis techniques. Furthermore, we provide a framework for benchmarking future techniques.