Advanced methods of detection of the Steganography content

Abstract

Steganography can be used for illegal activities. It is essential to be prepared. To detect steganography images, we have a counter-technique known as steganalysis. There are different steganalysis types, depending on if the original artifact (cover work) is known or not, or we know which algorithm was used for embedding. In terms of practical use, the most important are “blind steganalysis” methods that can be applied to image files because we do not have the original cover work for comparison. This philosophiæ doctor thesis describes the methodology to the issues of image steganalysis.In this work, it is crucial to understand the behavior of the targeted steganography algorithm. Then we can use it is weaknesses to increase the detection capability and success of categorization. We are primarily focusing on breaking the steganography algorithm OutGuess2.0. and secondary on breaking the F5 algorithm. We are analyzing the detector's ability, which utilizes a calibration process, blockiness calculation, and shallow neural network, to detect the presence of steganography message in the suspected image. The new approach and results are discussed in this Ph.D. thesis.