An overview of feature based steganalysis

Abstract

The field of cyber technology has grown tremendously over the years, which has made the transfer of data easy and fast. Open data can be of any form - text, image, audio, video and so forth. In order to transmit data through internet, it has to be converted into digital form. The advent of the internet has given people unlimited access to the digital world, which in turn may lead to huge security issues. Hence, the users need to protect their data from misuse. This has led to the birth of cyber security concepts like “data security in digital communication, copyright protection of digitized properties, and security of invisible communication over digital media”. Most of the research on steganalysis has been done separately on targeted and blind steganography with various steganographic schemes. Different steganalysis techniques have been used, but statistical steganalysis scheme proved to be more accurate than any other. For targeted steganalysis, the feature selection and extraction are easier since the steganographic algorithm is known and the features that would clearly project the changes can be identified. For blind steganalysis, the distinguishing features need to be identified, selected and extracted based on the image format and the transformation of images.

Keywords: Steganography, Steganalysis, Cryptography, Message Format, Feature Analysis, Statistical analysis, Targeted Steganography, Blind Steganography