Active Steganalysis Research Articles

Test

Image steganalysis ranges from detecting the presence of covert information in an image (passive steganalysis) to extraction of the information from the stego image (active steganalysis). One of the steps in active steganalysis is determining the stego algorithm used to produce the stego image. In this paper, a low dimensional combination of textural features is adapted for steganalysis. Also a novel blind statistical steganalyser to determine the spatial domain Least Significant Bit (LSB) based algorithms using one against one multi class classification is proposed. The proposed steganalyser is a multi class ensemble Fisher Linear Discriminant (FLD) classifier that uses novel low dimensional textural features for steganalysis. The performed experiments on the Bossbase database for 5 different LSB based algorithms for 8 different payloads show that the results are much better than the state of art steganalyser

Low Dimensional Multi Class Steganalysis of Spatial LSB based Stego Images Using Textural Features

Image steganalysis ranges from detecting the presence of covert information in an image (passive steganalysis) to extraction of the information from the stego image (active steganalysis). One of the steps in active steganalysis is determining the stego algorithm used to produce the stego image. In this paper, a low dimensional combination of textural features is adapted for steganalysis. Also a novel blind statistical steganalyser to determine the spatial domain Least Significant Bit (LSB) based algorithms using one against one multi class classification is proposed. The proposed steganalyser is a multiclass ensemble Fisher Linear Discriminant (FLD) classifier that uses novel low dimensional textural features for steganalysis. The performed experiments on the Bossbase database for 5 different LSB based algorithms for 8 different payloads show that the results are much better than the state of art steganalyser.

An Ensemble Transfer Learning Model for Detecting Stego Images

As internet traffic grows daily, so does the need to protect it. Network security protects data from unauthorized access and ensures their confidentiality and integrity. Steganography is the practice and study of concealing communications by inserting them into seemingly unrelated data streams (cover media). Investigating and adapting machine learning models in digital image steganalysis is becoming more popular. It has been demonstrated that steganography techniques used within such a framework perform more securely than do techniques using hand-crafted pieces. This work was carried out to investigate and examine machine learning methods’ critical contributions and beneficial roles. Machine learning is a field of artificial intelligence (AI) that provides the ability to learn without being explicitly programmed. Steganalysis is considered a classification problem that can be addressed by employing machine learning techniques and recent deep learning tools. The proposed ensemble model had four models (convolution neural networks (CNNs), Inception, AlexNet, and Resnet50), and after evaluating each model, the system voted on the best model for detecting stego images. Since active steganalysis is a classification problem that may be solved using active deep learning tools and modern machine learning methods, this paper’s major goal was to analyze deep learning algorithms’ vital roles and main contributions. The evaluation shows how to successfully detect images that contain a steganography algorithm that hides data in images. Thus, it suggests which algorithms work best, which need improvement, and which are easier to identify.

Deep-Learning Steganalysis for Removing Document Images on the Basis of Geometric Median Pruning

The deep-learning steganography of current hotspots can conceal an image secret message in a cover image of the same size. While the steganography secret message is primarily removed via active steganalysis. The document image as the secret message in deep-learning steganography can deliver a considerable amount of effective information in a secret communication process. This study builds and implements deep-learning steganography removal models of document image secret messages based on the idea of adversarial perturbation removal: feed-forward denoising convolutional neural networks (DnCNN) and high-level representation guided denoiser (HGD). Further—considering the large computation cost and storage overheads of the above model—we use the document image-quality assessment (DIQA) as threshold, calculate the importance of filters using geometric median and prune redundant filters as extensively as possible through the overall iterative pruning and artificial bee colony (ABC) automatic pruning algorithms to reduce the size of the network structure of the existing vast and over-parameterized deep-learning steganography removal model, while maintaining the good removal effects of the model in the pruning process. Experiment results showed that the model generated by this method has better adaptability and scalability. Compared with the original deep-learning steganography removal model without pruning in this paper, the classic indicators params and flops are reduced by more than 75%.

Steganographic Tool Detection using Specific Composite Feature Set and Weighted Decision Function

Steganographic tools available in the internet and other commercial steganographic tools are preferred than customized steganographic tools developed from scratch by unlawful groups. Hence a clue regarding the steganographic tool deployed in the covert communication process can save time for the steganalyst in the crucial active steganalysis phase. Signature analysis can lead to success in targeted steganalysis but tool detection needs to be taken forward from a point with a suspicious stego image in hand with no additional details available. In such scenarios, statistical steganalysis comes to rescue but with issues to be addressed like huge dimensionality of feature sets and complex ensemble classifiers. This work accomplishes tool detection with a specific composite feature set identified to distinguish one stego tool from the others with a weighted decision function to enhance the role of the specific feature set when it votes for a particular class. A tool detection accuracy of 85.25% has been achieved simultaneously addressing feature set dimensionality and complexity of ensemble classifiers and a comparison with a benchmark procedure has been made

Universal secret payload location identification in spatial LSB stego images

Locating steganographic payload is an important aspect of active steganalysis which deals with finding the pixel locations of the embedded secret message (payload) in the stego image. The residual-based payload location detectors are well suited for this purpose. In this paper, an improved universal, blind method of precisely identifying locations in the spatial domain using enhanced weighted residuals with only a few known stego images is proposed. This is done by estimating the cover from the available stego image using the proposed novel locally weighted bivariate shrinkage function in the transform domain. The enhanced weighted residuals ensure that the proposed method is universally applicable for detecting payload locations in spatial least significant bit stego images. The added advantage of the proposed method is that it does not require any prior knowledge of the cover source or the embedding algorithm to determine the exact locations. Experiments conducted on five spatial LSB domain algorithms show that the payload locations can be precisely estimated with a minimum accuracy of 90% with 100 known stego images.

A new fast and efficient active steganalysis based on combined geometrical blind source separation

We have presented a new active steganalysis method in order to break the block-discrete cosine transform (DCT) coefficients steganography. Our method is based on a combination of the Blind Source Separation (BSS) technique and Maximum A posteriori (MAP) estimator. We have additionally introduced a new geometrical BSS method based on the minimum range of mixed sources which reduces the computational cost of the proposed steganalysis. The high efficiency of this new combined method has been confirmed by enough experiments. These experiments show that, compared to the previous active steganalysis methods our active steganalysis method not only reduces the error rate but also causes a low computational cost.

Active Steganalysis Based on Pixels Classification in the Image

Active Steganalysis Based on Pixels Classification in the Image

Active steganalysis for interpolation-error based reversible data hiding

The interpolation-error based (IEB) reversible data hiding combines difference expansion and histogram-shifting techniques to embed secret data into interpolation-error histograms of a stego-image with high payload and low distortion has been proposed recently. In this paper, an active steganalysis scheme is proposed by analyzing and modeling histogram abnormality in the interpolation-error domain of sub-sampled images with the general Gaussian distribution (GGD) features. A support vector machine (SVM) classifier is trained by estimated parameters of GGD, and then a quantitative algorithm is presented to estimate the embedding length and locations. The experimental results show that the proposed active steganalysis scheme is effective in not only detecting the IEB method but also estimating its message length and embedding locations.

Nonparametric Steganalysis of QIM Steganography Using Approximate Entropy

This paper proposes an active steganalysis method for quantization index modulation (QIM)-based steganography. The proposed nonparametric steganalysis method uses irregularity (or randomness) in the test image to distinguish between the cover image and the stego image. We have shown that plain quantization (quantization without message embedding) induces regularity in the resulting quantized object, whereas message embedding using QIM increases irregularity in the resulting QIM-stego. Approximate entropy, an algorithmic entropy measure, is used to quantify irregularity in the test image. The QIM-stego image is then analyzed to estimate secret message length. To this end, the QIM codebook is estimated from the QIM-stego image using first-order statistics of the image coefficients in the embedding domain. The estimated codebook is then used to estimate secret message. Simulation results show that the proposed scheme can successfully estimate the hidden message from the QIM-stego with very low decoding error probability. For a given cover object the decoding error probability depends on embedding rate and decreases monotonically, approaching zero as the embedding rate approaches one.

Active steganalysis for histogram-shifting based reversible data hiding

This paper presents an innovative active steganalysis algorithm for reversible data hiding schemes based on histogram shifting. These schemes use histogram shifting to embed secret data in cover-images. However, some histogram patterns originating during the embedding procedure may be recognized readily by a steganalyst. The proposed algorithm analyzes the characteristics of histogram changing during the data embedding procedure, and then models these features into reference templates by using a 1×4 sliding window. A support vector machine is trained as the classifier for discriminating between cover-images and stego-images by adopting the template matching techniques. The hidden messages located at the histogram peak of the cover-image were further estimated by measuring the feature of adjacent histogram differences. Experimental results indicate that the proposed active steganalysis algorithm can effectively detect stego-images at low bit rates and estimate the hidden messages locations.

An active steganalysis approach for echo hiding based on Sliding Windowed Cepstrum

This paper presents an active steganalysis technique for echo hiding (EH). This approach can differentiate stego audios (with hidden message) from nature ones (without hidden message), and then extract sequential EH messages in stego audios without prior knowledge about the EH technique. The procedure of this algorithm can be simplified as follows: (i) Sliding Windowed Cepstrum (SWC) is obtained by employing a Sliding Window (SW) to move over the audio signal sample by sample. (ii) The echo detector is designed based on statistical analysis of the Cepstrum Peak Location Aggregation Degree (CPLAD). (iii) For stego signals, both echo delay and the range of segment length are estimated by exploring the changing rule of the SWC. (iv) The accurate estimation of segment length and synchronism position is acquired by the grid search algorithm. (v) Finally, the binary secret message can be extracted using the estimated parameters in the previous steps. The proposed method has been implemented and tested to extract the message which is embedded into audio clips by using EH techniques with different kernels. The experimental results show that the accuracy of extraction is no less than 90% when attenuation is greater than 0.45.

Message estimation for universal steganalysis using multi-classification support vector machine

To prevent misusing of the steganography from the terrorists, effective steganalysis schemes which discriminate the stego-images from suspicious images are necessary. Some steganalysis methods can accurately estimate the length of embedded messages but they are only useful in the pre-defined condition. Active steganalysis methods are powerful in length estimation such as regular singular (RS) and sample pairs analysis (SPA) steganalysis schemes, but they would become invalid in frequency domain. Passive steganalysis methods may discriminate stego-images from suspicious images in spatial and frequency domains such as Lyu and Fraid's steganalysis scheme, but they could not estimate the length of hidden messages. Although length estimation has been discussed in the active steganalysis methods for a while, it is a novel study in passive steganalysis method. We improve the Lyu and Fraid's universal steganalysis scheme and design an efficient length estimation policy in passive steganalysis methods. Experimental results demonstrate the efficiency and practicability of the proposed universal steganalysis scheme.

A mathematical framework for active steganalysis

A mathematical framework for steganalysis is presented in this Paper, with linear steganography being the main focus. A mathematically formal definition of steganalysis is given. Then active steganalysis, defined as the extraction of a hidden message with little or no a priori information, is formulated as a blind system identification problem within this framework. Conditions for identifiability (i.e., successful steganalysis) are derived. A procedure to systematically exploit any available spatial and temporal diversity information for efficient steganalysis is also discussed. Experimental results are given for steganalysis of Gaussian distributed, spread spectrum image steganography and watermarking. The proposed technique is observed to produce impressive results for a variety of performance measures. Based on the results we conclude that a common belief, namely, spread spectrum steganography/watermarking, is secure because the low strength, noise-like message carrier is no longer valid within the current context. Therefore, new questions regarding steganography security that differ from the standard information theoretic notion are raised and some answers are provided.