Steganalysis -Iterative Rule Learning to Discover Patterns

Abstract

In these years, everything is leading to new development and digitization. With these developments in technology, main challenge which exists is the threat of security. Steganography method usually embeds the sensitive messages in visually innocent cover images. The target of steganalysis is to determine the stego images from that of empty images. Every method depending on its hiding capacity of secret data in images place a unique markings or signature in stego images. To find this kind of markings in the images leads us to encorporate a classifier to be made for the purpose of finding the stego images which are usually the outcome of such steganography algorithm. In this, approach involves an evolutionary fuzzy rules to take out the markings of stego images in contrast to those empty images. Thus by using knowledge discovered, appropriate models for steganalysis can be involved and stego images can be found out and evolutionary algorithm can be optimized well. Thus the particular signature of steganographic method can be taken out well and also the kind of method used to produce stego image can be predicted.