Performance analysis of image steganalysis against message size, message type and classification methods

Abstract

Image steganalysis finds its application in the field of digital investigation. Performance of any image steganalysis algorithm depends on sensitivity of features and amount of data hidden in an image. The goal of this paper is to evaluate the performance of DWT feature based steganalysis algorithms against various state-of-art steganography methods and variable message embedding rates. Feature selection and classification are the two main steps of any image steganalysis algorithm. This paper also presents the comparative performance of individual algorithms against different classification methods. The images used for quantitative evaluation are taken from image database BSDS500 which contains images of different types and textures. All the algorithms are implemented in MATLAB and they are evaluated against stego images generated by steganography tools available for data hiding methods like F5, BlindHide, HideSeek, DBS, DFF and LSB.