Doctor of Philosophy (PhD), Wright State University, 2009, Computer Science and Engineering PhD

Security is an important issue related to the storage and communication of data and information. In data and information security, cryptography and steganography are two of the most common security techniques. On one hand, there is cryptography, which is the secret communication between two parties by message scrambling on the sender's side and message unscrambling on the receiver's side so that only the intended receiver gets the secret message. On the other hand, there is steganography, which is the hiding of information in a medium in such a way that no one other than the sender or the intended receiver realizes there is a hidden message. Successful reverse engineering of cryptography and steganography give cryptanalysis and steganalysis respectively. Cryptography and cryptanalysis constitute cryptology (or crypto) while steganography and steganalysis make up steganology (or stegano). This dissertation consists of three parts needed for a scientific study of a cryptanalysis problem. The first part lays out a comparative survey of various cryptology and steganology techniques by analyzing and comparing different methodologies using a set of predefined parameters. This part offers valuable knowledge on the state of the art techniques used on cryptanalysis. The second part proposes a new lossless synthetic stegano-crypto methodology that blends together five cryptography, steganography and compression techniques to form a single methodology for mutual information encryption and hiding in images. The methods that compose the synthetic methodology are SCAN Encryption, SCAN Compression, SCAN Steganography, Least Significant Bit (LSB) Steganography and Regional Steganography with Segmentation. The synthetic methodology plays the role of a complex and difficult technique that we have to work on in an attempt to break it by using a reverse engineering approach. In the third part, a cryptanalysis attack against the proposed synthetic stegano-crypto methodology is presented (open full item for complete abstract)

Committee: Nikolaos Bourbakis PhD (Advisor); Nikolaos Bourbakis PhD (Committee Chair); Soon Chung PhD (Committee Member); Yong Pei PhD (Committee Member); Arnab Shaw PhD (Committee Member); Monish Chatterjee PhD (Committee Member) Subjects: Computer Science

Master of Science, University of Akron, 2009, Computer Science

The purpose of steganography is to hide the presence of a message. Modern day techniques embed pictures and text inside computer files. Steganalysis is a field devoted to detecting steganography in files and possibly extracting the hidden image or text. This thesis introduces a new idea for steganalysis, that of training an artificial neural network to identify images that have another image embedded in them. Two different types of artificial neural networks, a standard and a shortcut type, are trained for two different types of data sets. One data set contains images with and without hidden images embedded in them. The other data set is derived from calculating the luminance values of the files in the first data set. The experimental results show that the shortcut artificial neural network performs better than the standard trained network, but still does not yield good results. We compare these results to two well known steganalysis tools. To date, no steganalysis technique has shown much promise, but this is highly experimental research. Many questions remain unanswered, and this thesis forms the basis for future experiments with using an artificial neural network as a useful steganalysis tool.

Committee: Kathy Liszka PhD (Advisor) Subjects: Computer Science