Scholarly record

STATISTICAL ANALYSIS OF STEGANALZTICAL MOTHOD FOR STEGHIDE DETECTION

Ing. Radim Kuncicky, Prof. Lacezar Licev, Ing. Jakub Hendrych

Abstract

Steganography is part of security science that deals with art of data hiding. Modern steganography can serve as an effective tool for hidden data transfers or watermarking for various computer formats, like JPEG, PNG, and MPEG. But it can be misused for stealing data or information transfer in illegal activities too. Most of today available steganography tools are free for use with easy user interface. This significantly increases the risk of exploiting in business sector. Revealing this data is very difficult and nontrivial, many used methods are complicated and hard to analyze, especially the methods based on artificial intelligence. This article gives an idea about behavior of one of the modern method, which use analysis of inner properties of JPEG format. This method is based on changes, which are introduced in Huffman coding during data injection with steganography tool called Steghide. We reveal detection rate dependencies on different image resolutions and multiple message lengths. As a last thing we discus is the inner neural network success learning rate versus the networkпїЅs topology.

Publication Impact Profile

• Captures
• Mendeley - Readers: 1

PLUMX see details

DOI resolver: doi.org/10.5593/sgem2017/21/s07.078

Publication details

Back to publication list