Electrical & Computer Engineering, Department of
Date of this Version
2013
Citation
2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA)
Abstract
Video steganography is fast emerging as a next generation steganographic medium that offers many advantages over traditional steganographic cover media such as audio and images. Various schemes have recently emerged which take advantage of video specific properties for information hiding, most notably through the use of motion vectors. Although many steganographic schemes have been proposed which exploit several possible steganographic domains within video sequences, few attacks have been proposed to combat such schemes, and no current attacks have been shown to be capable of defeating multiple schemes at once.
In this paper, we will further expand upon our proposed Discrete Spring Transform (DST) steganographic attack. We will explore further applications of the transform and how it may be used to defeat multiple steganographic schemes, specifically current video steganography schemes. The effectiveness of the proposed algorithm will be shown by attacking a multi-dimensional steganographic algorithm embedded in video sequences, where the scheme operates in two different dimensions of the video. The attack is successful in defeating multiple steganographic schemes verified by determining the BER after DST attack which always remains approximately 0.5. Furthermore, the attack preserves the integrity of the video sequence which is verified by determining the PSNR which always remains approximately above 30dB.
Comments
©2013 IEEE