Universal steganalysis using IQM and multiclass discriminator for digital images

As the modern world is gradually becoming internet and computer savvy with huge amount of information stored and exchanged over the internet through digital images for covert communication. There are different methods to hide information in digital images in different domain. First part of this paper gives concise information about different approaches for exiting steganalysis techniques for information detection in images. The objective of this paper is to know limitation of exiting Steganalysis techniques and proposed a Universal Steganalysis technique with the help of prominent statistical features of digital images through which the presence of information will be detected. There are lot of techniques for features extraction and classification. All techniques treat different images as same, therefore the different statistical properties of different images are considered for prominent feature selection. Second part consist of proposed model for universal Steganalysis.