Universal Steganography Detector Based on an Artificial Immune System for JPEG Images

Steganography is a hiding information technique heavily used nowadays. Though initially it was used to establish hidden communication channels, modern steganography has been found useful to hide code inside multimedia objects, mostly images. Its goal is to infiltrate malware into organizations or personal devices. This kind of malware is called stegomalware. As countermeasure, modern steganalysis methods employing different Computational Intelligence techniques such as Support Vector Machine, Machine Learning, Fisher Linear Discriminant, and others have been utilized. In this work we present a new stegananalysis method based on an Artificial Immune System (AIS), to detect JPEG images modified with three well known steganographic tools: F5, Outguess, or Steghide. It is also proposed the usage of Haar Wavelets to extract a feature vector that best describes the analyzed image, this is due the Haar Wavelets fast calculation and information synthesis. Our experimentation results are competitive against techniques representative of the state of the art.