Steganalysis Approach Research Articles

Undetectable video steganography by considering spatio-temporal steganalytic features in the embedding cost function

The basic requirement of a steganography approach is security against steganalysis attacks. In other words, a steganography method is reliable as long as it withstands all of the known steganalysis approaches. In order to preserve the security of a steganography method, the statistical features of the embedded and the original media must be as close as possible. To achieve this goal, in this paper steganography is applied by introducing the following contributions. Firstly, a new method is suggested to find the most impalpable embedded motion vector (MV). Also, a novel modification cost function with respect to the MVs’ intra-frame and inter-frame statistical differences before and after embedding is proposed for the syndrome-trellis coder. Furthermore, a pseudo-random generator is introduced for altering the arrangement of motion vectors which are used by the syndrome-trellis coder to improve its efficiency. Experimental results show that the proposed method is the most secure MV-based steganography scheme against the state-of-the-art video steganalysis methods as well as preserving other steganography measurements including imperceptibility and compression ratio. Moreover, the computational cost of the proposed scheme is far less than its main rival.

A Novel Quantum Audio Steganography–Steganalysis Approach Using LSFQ-Based Embedding and QKNN-Based Classifier

In this paper, a steganography–steganalysis model for quantum audio signals is presented. This model consists of two separate parts of steganography and steganalysis. In the steganography part, to increase the imperceptibility and enhance the undetectability of the least significant bit method, first, the quantum host audio signal is divided into quantum frames of the specified length, and then, using the triangular number sequence, some samples of each frame are selected for the embedding operation. Finally, the embedding process is carried out within the least significant fractional qubit of the amplitude information of the selected samples. The steganalysis part also includes a quantum universal steganalyzer for detecting the embedding operations of the steganography part and a feature extraction module for calculating the power feature of quantum audio signal frames. The recognition operation is based on the quantum K-nearest neighbor algorithm and the Hamming distance criterion. All quantum circuits of the steganography and steganalysis parts of the proposed model were simulated and then tested and evaluated using different audio files. Detection accuracy of over 90% indicates the accuracy and efficiency of the proposed method in quantum audio steganography detection in the context of quantum communication networks.

A Video Steganalytic Approach Against Quantized Transform Coefficient-Based H.264 Steganography by Exploiting In-Loop Deblocking Filtering

This article presents a method for detection of H.264 steganographic algorithms that modulate quantized transform coefficients (QTCs) for data hiding. The in-loop deblocking filter specified in the H.264 standard is used for attenuating the discontinuities at coded block boundaries. However, our investigation indicates that the deblocking filtering process would probably be affected by QTC modification. Therefore, we suggest performing steganalysis by considering the influence of QTC modification on filtering decisions and filtering operations. In the filtering of luma block edges, for each luma sample, we try to obtain the cumulative amount of modification applied to it caused by filtering operations, and determine the frequency of the modification made to its sample value. Based on this, a 768-dimensional feature set is designed for training and classification. Extensive experiments were conducted under a range of setups to evaluate the effectiveness of the proposed method. The corresponding experimental results demonstrate that, compared with the prior art, our proposed method is generally capable of providing superior performance on detecting QTC modification in H.264 video, even at low embedding strengths.

A New Convolutional Neural Network-Based Steganalysis Method for Content-Adaptive Image Steganography in the Spatial Domain

Convolutional neural network-based methods are attracting increasing attention in steganalysis. However, steganalysis for content-adaptive image steganography in the spatial domain is still a difficult problem. In this paper, a new convolutional neural network-based steganalysis approach was proposed with two contributions. 1) By adding more convolutional layers in the lower part of the model, we proposed a new arrangement of convolutional layers and pooling layers, which can process the local information better than the existing CNN models in steganalysis. 2) By adding the global average pooling layer before the softmax layer instead of using global average pooling before the fully connected layer, the global average pooling was placed in a better position for steganalysis. Two state-of-the-art steganographic algorithms in the spatial domain, namely, WOW and S-UNIWARD, were used to evaluate the effectiveness of our model. The experimental results on BOSSbase showed that the proposed CNN could obtain better steganalysis performance than YeNet across all tested algorithms when the payloads were 0.2, 0.3, and 0.4 bpp.

Enhancing the performance of CNN-based blind image steganalysis approach using multi-GPU TESLA P100

Blind image Steganalysis is the binomial classification problem of determining if an image contains hidden data or not. Classification problems have two main steps: i) feature extraction step and ii) classification step. Traditional blind image steganalysis approaches use handcrafted filters for the first step and use classifiers such as support vector machine (SVM) for the second step. The rapid development of steganographic techniques makes it harder to design new effective handcrafted filters, which negatively affect the feature extraction step. Recently, Convolutional Neural networks (CNNs) are introduced as an auspicious solution for this problem. CNN-based steganalysis can automatically extract features from the input images without using handcrafted filters. Although considerable success has been achieved with CNNs, CNN-based applications are considered as time consuming applications. Accordingly, it is important to quicken the CNN-based steganalysis approaches training in order to make them more applicable. This paper suggested an implementation technique of the improved Gaussian-Neuron CNN (IGNCNN) steganalysis approach on GPUs. In this paper data parallelism concept is applied to the convolutional layers while model parallelism concept is applied to the fully connected layers. Results show that the proposed method provides better performance as compared with IGNCNN [1] by an average speed up factor of 1.4 X.

Comparative Study of Three Steganographic Methods Using a Chaotic System and Their Universal Steganalysis Based on Three Feature Vectors.

In this paper, we firstly study the security enhancement of three steganographic methods by using a proposed chaotic system. The first method, namely the Enhanced Edge Adaptive Image Steganography Based on LSB Matching Revisited (EEALSBMR), is present in the spatial domain. The two other methods, the Enhanced Discrete Cosine Transform (EDCT) and Enhanced Discrete Wavelet transform (EDWT), are present in the frequency domain. The chaotic system is extremely robust and consists of a strong chaotic generator and a 2-D Cat map. Its main role is to secure the content of a message in case a message is detected. Secondly, three blind steganalysis methods, based on multi-resolution wavelet decomposition, are used to detect whether an embedded message is hidden in the tested image (stego image) or not (cover image). The steganalysis approach is based on the hypothesis that message-embedding schemes leave statistical evidence or structure in images that can be exploited for detection. The simulation results show that the Support Vector Machine (SVM) classifier and the Fisher Linear Discriminant (FLD) cannot distinguish between cover and stego images if the message size is smaller than 20% in the EEALSBMR steganographic method and if the message size is smaller than 15% in the EDCT steganographic method. However, SVM and FLD can distinguish between cover and stego images with reasonable accuracy in the EDWT steganographic method, irrespective of the message size.

A new steganalysis approach with an efficient feature selection and classification algorithms for identifying the stego images

The detection of stego images by using the steganalysis approach is one of the demanding task in the recent days. Because, it is used as an importer for the immoral activities by hiding the secrets in the messages. For this secret identification, the traditional works develop various steganographic techniques for steganalysis. But, it has some important drawbacks such as low detection percentage, inefficient results, and increased complexity. In order to solve these issues, this paper introduces a new steganalysis approach with an efficient feature selection and optimization techniques. The aim of this paper is to accurately detect the stego and clean images by implementing an efficient classification algorithm. Initially, a novel Coefficient based Walsh Hadamard Transform along with the Gray Level Co-occurrence Matrix (GLCM) is used for extracting the features of the image. Then, an efficient feature selection technique, namely, Pine Growth Optimization (PGO) is developed to select the optimal features from the extracted features. Finally, the Cross Integrated Machine Learning (CIML) classifier is implemented to classify the stego and clean images. The newness is provided during the feature extraction, selection and classification processes. In experiments, the performance results of the proposed steganalysis is evaluated and compared with the existing approaches by using different measures.

PFA-based feature selection for image steganalysis

This paper presents a universal image steganalysis method based on feature selection by principal feature analysis (PFA). The goal of this paper is to increase the performance of existing image steganalysis approaches using PFA-based feature selection method and reduce the high dimensionality of the features used in state-of-the-art steganalysis methods. Principal component analysis (PCA) is widely used in pattern recognition applications. However, PCA has disadvantage that, all the generated features are transformed features. While, PFA selects the subset of preliminary features which contains necessary information. PFA is applied on spatial domain subtractive pixel adjacency matrix features and in case of transform domain, CHEN features (intrablock and interblock Markov-based features) and CC-PEV features (PEV features enhanced by Cartesian calibration). The experimental results show that PFA is effective and efficient in eliminating redundant features. Experimental results prove that the use of PFA method in steganalysis is superior in terms of dimensionality reduction of features and increases the classification performance.

PFA-based feature selection for image steganalysis

This paper presents a universal image steganalysis method based on feature selection by principal feature analysis (PFA). The goal of this paper is to increase the performance of existing image steganalysis approaches using PFA-based feature selection method and reduce the high dimensionality of the features used in state-of-the-art steganalysis methods. Principal component analysis (PCA) is widely used in pattern recognition applications. However, PCA has disadvantage that, all the generated features are transformed features. While, PFA selects the subset of preliminary features which contains necessary information. PFA is applied on spatial domain subtractive pixel adjacency matrix features and in case of transform domain, CHEN features (intrablock and interblock Markov-based features) and CC-PEV features (PEV features enhanced by Cartesian calibration). The experimental results show that PFA is effective and efficient in eliminating redundant features. Experimental results prove that the use of PFA method in steganalysis is superior in terms of dimensionality reduction of features and increases the classification performance.

A Linguistic Steganalysis Approach Base on Source Features of Text and Immune Mechanism

Linguistic steganalysis is a technique that discovering potentially hidden information embedded through using linguistically in plain text using. Varieties of syntax and multi-meanings of semantics for linguistics augment the difficulty of linguistic steganalysis intensely, thereby it is a challenge area. In this paper, we propose a novel steganalysis method for linguistics based on immune. This method has two attributions: i). basis statistical features of text are employed for blind steganalysis ii). immune technique is chosen to build a two-level detection mechanism to detect two categories of stego text respectively, one of which is Success-Stego-text and another is False-Stego-text. Appropriate detections are generated and preferable features are signed. Experiments prove the approach has higher accuracy than current steganalysis algorithms. Especially when the segment size of text is greater than 3kB, the accuracies of detecting for natural text and stego text are both more than 95%.

Adaptive Image Steganalysis

Image steganography is the process of sending messages secretly by hiding the message in AN image content. Although steganography hides information in any kinds of digital mediums, digital images are the most popular carrier because of the frequency of usage in the internet. Adaptive steganography methods do not embed a secret message uniformly in images. Steganalytic techniques are used to detect whether an image contains a hidden message by analyzing various image features extracted from stego images which contain hidden messages and cover images with no hidden messages. In this paper, an Adaptive Image Steganalysis (AIS) algorithm is proposed to breakdown adaptive steganography methods. In this method, a suspicieuse block of an image is selected intelligentlly. Afterward, a set of selected features is extracted from this chosen image block. As a result, a feature vector is generated which improves accuracy of steganalysis. Experimental results show that AIS algorithm outperforms other classic steganalysis approaches.

A Steganalytic Approach to Detect Motion Vector Modification Using Near-Perfect Estimation for Local Optimality

This paper presents a steganalytic approach against motion vector-based video steganography that does not depend on the detailed knowledge of embedding algorithms. In most state-of-the-art video coding standards, the motion vector is the result of block-based motion estimation using rate-distortion optimization. That is to say, each motion vector is locally optimal in a rate-distortion sense, and any modification will inevitably shift the motion vector from locally optimal to non-optimal. As a consequence, it is a very strong evidence of steganography if some motion vectors are found to be locally non-optimal. Based on this fact, the core of our method is an estimator to check the local optimality of motion vectors in a rate-distortion sense. We try to recover the necessary information used for motion vector decision that is lost during lossy compression, based on which a 36-D feature set is formed for training and classification. To demonstrate the effectiveness of the proposed approach, experiments are carried out in different settings. The corresponding results show that our approach has a wide applicability even at low embedding strengths. Particularly, the problem of cover source mismatch is largely alleviated, which indicates that the proposed approach is suitable to be used in situations where a very limited priori knowledge is available.

Segmentation Based Video Steganalysis to Detect Motion Vector Modification

This paper presents a steganalytic approach against video steganography which modifies motion vector (MV) in content adaptive manner. Current video steganalytic schemes extract features from fixed-length frames of the whole video and do not take advantage of the content diversity. Consequently, the effectiveness of the steganalytic feature is influenced by video content and the problem of cover source mismatch also affects the steganalytic performance. The goal of this paper is to propose a steganalytic method which can suppress the differences of statistical characteristics caused by video content. The given video is segmented to subsequences according to block’s motion in every frame. The steganalytic features extracted from each category of subsequences with close motion intensity are used to build one classifier. The final steganalytic result can be obtained by fusing the results of weighted classifiers. The experimental results have demonstrated that our method can effectively improve the performance of video steganalysis, especially for videos of low bitrate and low embedding ratio.

Deep learning is a good steganalysis tool when embedding key is reused for different images, even if there is a cover sourcemismatch

Since the BOSS competition, in 2010, most steganalysis approaches use a learning methodology involving two steps: feature extraction, such as the Rich Models (RM), for the image representation, and use of the Ensemble Classifier (EC) for the learning step. In 2015, Qian et al. have shown that the use of a deep learning approach that jointly learns and computes the features, was very promising for the steganalysis. In this paper, we follow-up the study of Qian et al., and show that in the scenario where the steganograph always uses the same embedding key for embedding with the simulator in the different images, due to intrinsic joint minimization and the preservation of spatial information, the results obtained from a Convolutional Neural Network (CNN) or a Fully Connected Neural Network (FNN), if well parameterized, surpass the conventional use of a RM with an EC. First, numerous experiments were conducted in order to find the best "shape" of the CNN. Second, experiments were carried out in the clairvoyant scenario in order to compare the CNN and FNN to an RM with an EC. The results show more than 16% reduction in the classification error with our CNN or FNN. Third, experiments were also performed in a cover-source mismatch setting. The results show that the CNN and FNN are naturally robust to the mismatch problem. In Addition to the experiments, we provide discussions on the internal mechanisms of a CNN, and weave links with some previously stated ideas, in order to understand the results we obtained. We also have a discussion on the scenario "same embedding key".

Steganalysis of JPEG images using enhanced neighbouring joint density features

In this study, a blind steganalysis approach which accurately discloses low rate data hiding schemes is proposed. The absolute values of neighbouring joint density (absNJ) are used for feature extraction. In this way, the intra- and inter-block situations are employed providing a variety of different features. Aside from the absolute values of discrete cosine transform (DCT) coefficients, the differential DCT coefficients are also exploited to extract the features. Moreover, the absNJ features will be extended and used over differential DCT coefficients. It is shown that applying the Pth power of the DCT coefficients instead of their first power gains more discriminative features. Then, using the ensemble classifier, the cover image is discriminated from the stego one. Experimental results indicate the efficiency of the new scheme in comparison with the previously presented schemes.

Detection of Stego-Images in Communication among the Terrorist Boko-Haram Sect in Nigeria

Nigeria was listed as a part of terrorist states by United States of America as a result of Islamic group (Boko Haram Sect) attacks and other activities in the nation. It has also been discovered that the group employs “steganographic” schemes as a secure means for transmitting their hidden information to each other via Internet and social networks. The group has killed thousands of people since their increased insurgency in July, 2009. These challenges have affected the nation’s foreign policies, political and social economic developments. This research addresses the challenges by employing forensic technique using blind steganalysis approach to detect the presence of the hidden messages in images. Image Quality Metric is employed for extracting the features, and logistic regression is trained as the classifier to predict the stego-images. We show the effectiveness of the method by conducting test and analysis with 319 images varying in size and style. The result shows that the performance of the method is better than other steganalysis methods.

A new metaheuristic feature subset selection approach for image steganalysis

Processing a huge amount of information takes extensive execution time and computational sources most of the time with low classification accuracy. As a result, it is needed to employ a phase of pre-analysis processing, which can influence the performance of execution time and computational sources and classification accuracy. One of the most important phases of pre- processing is Feature selection, which can improve the classification accuracy of steganalysis. The experiments are accomplished by using a large and important data set of 686 features vectores named SPAM. One of the promising application domains for such a feature selection process is steganalysis. In this paper, we propose a new metaheuristic approach for image steganalysis method for detecting stego images from the cover images in JPEG images using a feature selection technique based on an improved artificial bee colony. Within the ABC structure the k-Nearest Neighbor (kNN) method is employed for fitness evaluation. ABC and kNN have been adjusted together to make an operative dimension reduction method Experimental results demonstrate the effectiveness and accuracy of the proposed technique compared to recent ABC-based feature selection methods and other existing techniques.

JPEG Steganalysis Based on Feature Fusion by Principal Component Analysis

Aiming to the problems in the existing JPEG steganalysis schemes, such as high redundancy in features and failure to make good use of the complementarities among them, this study proposed a JPEG steganalysis approach based on feature fusion by the principal component analysis (PCA) and analysis of the complementarities among features. The study fused complementary features and isolated redundant components by PCA, and finally used RBaggSVM classifier for classification. Experimental results show that this scheme effectively improves the detection rate of steganalysis in JPEG images and achieves faster speed of image classification.

A New Approach For Image Hiding Based On Contourlet Transform

A new image hiding method based on the contourlet transform is proposed in this paper. This strategy is based on storing information in high frequency subbands of contourlet transform. The embedding approach is in direction that the contourlet sub-bands have the least statistical disorder. As a result, the proposed algorithm has a higher robustness against to common steganalysis approaches. In addition, the quality of stegano image has considerably improved in comparison with related state of the art methods, with the extracted secret image having an acceptable quality. Furthermore, the experimental results show robustness respect to Gaussian noise and other attacks such as JPEG compression. DOI: http://dx.doi.org/10.11591/ijece.v2i5.1433 Full Text: PDF

Mixing high-dimensional features for JPEG steganalysis with ensemble classifier

This paper proposes a JPEG steganalysis scheme based on the ensemble classifier and high-dimensional feature space. We first combine three current feature sets and remove the unimportant features according to the correlation between different features parts so as to form a new feature space used for steganalysis. This way, the dependencies among cover and steganographic images can be still represented by the features with a reduced dimensionality. Furthermore, we design a proportion mechanism to manage the feature selection in two subspaces for each base learner of the ensemble classifier. Experimental results show that the proposed scheme can effectively defeat the MB and nsF5 steganographic methods and its performance is better than that of existing steganalysis approaches.