An Artificial Neural Network for Wavelet Steganalysis

Hiding messages in image data, called steganography, is used by criminals and noncriminals alike to send information over the internet. The detection of hidden messages in image data stored on websites and computers, called steganalysis, is of prime importance to cyber forensics. Automated detection of hidden messages is a requirement, since the shear amount of image data available online makes it impossible for a person to investigate each image separately. The purpose of this project is to develop a prototype software system that automatically classifies an image as having hidden information or not, using a powerful classifier called an artificial neural network (ANN). The novelty of this ANN is its ability to detect messages hidden with wavelet embedding algorithms, in addition to other transforms. 1 Project Description Steganalysis is of increasing importance to cyber security. While the number of freeware packages available for steganography is increasing each year, the detection of most of these methods is neither satisfactory nor fully automated. While it is possible to hide messages within a variety of data file types, image data is likely to be the medium of choice for cyber criminals for several reasons. First, because of the high level of redundancy in image data, it is possible to embed a great deal of hidden information. Second, innocuous-looking images are commonplace on every computer and arouse little suspicion. Virtually all computer-users keep digital photos of friends and family, vacations, special events, etc. on their hard drives. Many web sites use images as a way to add interest and break up the monotony of text. By contrast, audio or video files posted on web sites are prone to be examined for copyright infringement. 1 Final Report for "An artificial neural network for wavelet steganalysis," supported by the Midwest Forensic Resource Center, Grant No. 2002–LP–R–083, 2004-6.