Steganalysis of spread spectrum data hiding exploiting cover memory

Abstract

In this paper we study steganalysis, the detection of hidden data. Specifically we focus on detecting data hidden in grayscale images with spread spectrum hiding. To accomplish this we use a statistical model of images and estimate the detectability of a few basic spread spectrum methods. To verify the results of these findings, we create a tool to discriminate between natural "cover" images and "stego" images (containing hidden data) taken from a diverse database. Existing steganalysis schemes that exploit the spatial memory found in natural images are particularly effective. Motivated by this, we include inter-pixel dependencies in our model of image pixel probabilities and use an appropriate statistical measure for the security of a steganography system subject to optimal hypothesis testing. Using this analysis as a guide, we design a tool for detecting hiding on various spread spectrum methods. Depending on the method and power of the hidden message, we correctly detect the presences of hidden data in about 95% of images.

Steganalysis based on Markov Model of Thresholded Prediction-Error Image

“…In other words, it is the overall correct classification rate for all test images. We have implemented the Markov chain based method in [5] and applied it to the same set of images and the same data hiding methods. The same training and testing procedures are used.…”

“…Inspired by [5], we propose in this paper a steganalysis system based on Markov chain model of thresholded prediction-error image. Image pixels are predicted with the neighboring pixels.…”

“…In [5], a steganalysis method based on Markov model is proposed. The empirical transition matrix of a test image is formed.…”

“…Because the size of the empirical transition matrix is very large, e.g., the 65536 elements for a grey level image with bit depth of 8, it cannot be used as features directly. The authors of [5] select several largest probabilities along the main diagonal together with their neighbors, and randomly select some other probabilities along the main diagonal as features. It is obvious that some useful information might be ignored due to the random fashion of feature formulation.…”

“…It is obvious that some useful information might be ignored due to the random fashion of feature formulation. The data hiding methods addressed in [5] are restricted to spread spectrum (SS) data hiding methods. Although it may not carry as many information bits as the LSB methods in general, the SS methods can still serve for the covert communication purpose.…”