Steganalysis:Multi-Class Classification of Images Using Linear Support Vector Machine

Steganographic techniques have been used to embed covert messages inside a piece of unsuspicious media and sending it without anyone’s knowledge about the survival of the covert message. Steganalysis is the process of detecting the presence of concealed information from the stego image and it can lead to the prevention of terrible security incidents. Steganalysis consist of two stages, the first stage is to identify the existence of the hidden message and the second stage is to retrieve the content of the message. In the existing method, for identifying the existence of the message, two-class classification using Support Vector Machine is used to differentiate the cover and stego images. In this paper, a new technique called multi-class classification using Linear Support Vector Machine is used to differentiate the cover and different type’s stego images.