Steganography and Steganalysis of JPEG Images: A Statistical Approach to Information Hiding and Detection

Steganography is the art of secret communication between two parties that not only conceals the contents of a message, but also its existence. Steganography has advantages over cryptography since message can be communicated to the desired recipient without raising any suspicion. Steganalysis attempts to detect the existence of embedded data in a steganographically altered cover file. This book describes three JPEG steganography algorithms and one steganalysis algorithm. The steganography algorithms are primarily based on restoring the statistics of a JPEG image after embedding data in them which results in low detection rate as compared to other existing algorithms. The steganalysis algorithm uses second order statistical features along with machine learning to train and predict if an image contains embedded message. Steganography can be used in many applications. In commercial sector, data hiding techniques can be used to enforce copyright on a digital medium. The access control information can be embedded inside the digital medium which can eventually be verified by the media player before playing the file.