Automatic detection of steganographic images to facilitate decoding

Abstract

Undeniably we are living in digital age. Digital gadgets are creating millions of tril-lion digital documents each day. Techniques and methodologies of hiding classified digital data have changed now. Terrorists and bad guys changed their techniques to hide their confidential data. Encryption converts plain text into cipher text, which is directly unreadable, but encrypted data gives an idea to investigator that this data have some important information. Steganography is used for data hiding and communica-tion of classified information un-noticeably. In image steganography, user hides the classified message behind an innocent or benign image. Sometimes user also modifies or removes file extension to further dodge the investigator. Investigation techniques are rated with its capacity (size) to hide secret message and imperceptibility. Several sophisticated techniques of image steganography are developed. Steganalysis is the method to recover secret message embedded in benign image. Several software tools are available that develop their own stenography and Steganal-ysis technique. Poor encoding techniques can be reversed with some virtual or statis-tical attacks. The purpose of this thesis is to develop a hybrid and sophisticated Steganalysis tool that decodes the embedded message encoded by common and best-practiced tech-nique. This hybrid algorithm has the advantage to scan a complete drive or folder to recover the modified or removed file extension and decode if there is any steno-graphic image found. The developed tool contains the decoding algorithm of most common stenographic technique.