This paper introduces a deep learning-based Steganography method for hiding secret information within the cover image. For this, we use a convolutional neural network (CNN) with Deep Supervision based edge detector, which can retain more edge pixels over conventional edge detection algorithms. Initially, the cover image is pre-processed by masking the last 5-bits of each pixel. The said edge detector model is then applied to obtain a gray-scale edge map. To get the prominent edge information, the gray-scale edge map is converted into a binary version using both global and adaptive binarization schemes. The purpose of using different binarization techniques is to prove the less sensitive nature of the edge detection method to the thresholding approaches. Our rule for embedding secret bits within the cover image is as follows: more bits into the edge pixels while fewer bits into the non-edge pixels. Experimental outcomes on various standard images confirm that compared to state-of-the-art methods, the proposed method achieves a higher payload.

Image steganography using deep learning based edge detection

Modern research in Steganography focuses on increasing the capacity without keeping any perceptible signature into the stego-media. In this work, we present a pixel value difference (PVD) based high capacity methodology. Embedding is not restricted only to high contrast pixel pairs. Low contrast pairs are also considered to enhance the capacity. Embedding process varies depending on the local contrast of the pair. Secret message is encrypted before embedding and pixels pairs are chosen in a non-sequential manner, rendering the methodology more secure. Experiment is performed on more than two hundred images. It is observed that embedding by proposed technique is visually imperceptible and can withstand various types of attacks. The benchmark test like StirMark analysis also indicates the strength of the work. Comparison of performance with a number of other methods shows that proposed methodology maintains high PSNR with enhanced capacity.

A PVD based high capacity steganography algorithm with embedding in non-sequential position

Steganography, the hiding technique used to secure sensitive data (i.e., images, audio ) while communication takes place. In this paper, the message is embedded in color image in frequency domain exploiting Genetic Algorithm (GA) which provides the robustness i.e., the algorithm can withstand against any rigorous testing and brutal attack except destruction of the stego image. The fragmentation of 8-bit binary stream of each color component to 4-bit and applying Genetic Algorithm (GA) to increase the robustness of the scheme. Random Multiple bits are chosen to embed secret message which increases security along with the payload. The use of transform domain, hash function based random pixel and bit selections for data hiding, secrete data encryption and more over use of Genetic Algorithm (GA) for optimization is the novelty of the proposed work. Perspective and meticulous statistical analysis has been done to immune the algorithm from any attack. The algorithm proposed here is tested with benchmark tools like StirMark 4.0, Confusion Matrix (Receiver Operating Curve Characteristic (ROC) curve) along with steganalysis and statistical tools. Visual disturbance and distortion in stego image is also very insignificant here.

Random selection based GA optimization in 2D-DCT domain color image steganography

Digital image steganography: A literature survey

Digital image steganography survey and investigation (goal, assessment, method, development, and dataset)

Steganography is a very useful technique which aims at preventing loss of privacy during the process of data communication, especially over the internet. It can involve different forms of media like image, video (i.e., image sequence), audio etc. We propose a novel steganographic approach in spatial domain using pixel value differencing (PVD) or sample value differencing (SVD) technique and Galois field (GF (28)) operations in order to provide a two layered security for hiding message bits. Our method not only has a very high embedding capacity, but is also capable of withstanding statistical attacks. The proposed method embeds from 2 to 6 bits of the message per pixel in each image component, whereas it can embed a minimum of 6 bits and a maximum of 13 bits of message per sample in audio component at the expense of no perceivable distortion and loss of the cover media quality.

An efficient multi-bit steganography algorithm in spatial domain with two-layer security

Two-point FFT-based high capacity image steganography using calendar based message encoding

For years, people have devised different techniques for encrypting data while others have attempted to break these encrypted codes. Steganography is a technique that allows one to hide data within an image, audio, or video. Imperceptibility of noise is an important issue for any steganographic technique. For this reason, we have proposed a new audio steganographic approach without considering the sample value less than 50 Hz and above 5000 Hz, because this range is highly sensible by ear as per biological phenomenon. The novelty of our work is that we have embedded the matching position value (determined from audio sample value itself) in 4 least significant bits rather than direct message bits flipping. Rather than performing sequential embedding, samples are selected nonsequentially based on audio feature itself that helps to withstand statistical attacks. Blind audio steganalysis techniques are hardly found—most of them are either algorithm specific or useful for bit-flipping data hiding scheme. In this context, unlike most common message bit flipping techniques, a novel bit matching and position concealment procedure has been proposed. Satisfactory theoretical and experimental outcomes have been achieved.

Nabanita Mukherjee

Papers

A PVD based high capacity steganography algorithm with embedding in non-sequential position

An efficient multi-bit steganography algorithm in spatial domain with two-layer security

Two-point FFT-based high capacity image steganography using calendar based message encoding

A Novel Position Concealment Audio Steganography in Insensible Frequency