Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: Ph.D. Dissertation
- Language: Farsi
- Document No: 46608 (05)
- University: Sharif University of Technology
- Department: Electrical Engineering
- Advisor(s): Ghaem Maghami, Shahrokh
- Abstract:
- The goal of this thesis is using sparse decomposition to design secure steganography algorithms to insert secret data into the different cover signals. We propose data embedding in higher semantic levels of the cover signal to reach this goal. Sparse decomposition represents a signal as a linear combination of its structural elements. Data insertion into the sparse coefficients slightly changes the effect of structural elements in the signal representation. So, quality of the stego signal is preserved and such a steganography method leads to higher imperceptibility. In addition, data insertion into the higher semantic levels leads to the higher undetectability. Steganalyzers use statistical properties of signals to determine whether a suspicious signal contains hidden data or not while semantic contents of the signal are usually disregarded in the statistical analysis. Thus, we embed secret bits into the semantic contents of the cover signal by using its sparse representation. This method is more secure comparing to previous methods because it leads to less detectable statistical changes in the cover signal.
In this work, we propose a transform domain steganography method and also a spatial domain method for data embedding in the color images. Transform domain method is called SDW-steg which uses wavelet transform domain for data embedding. The dictionaries are estimated from four wavelet sub-bands of a random color band of the cover image. Sparse representation of wavelet sub-bands of two other color bands over the estimated dictionary are used for data insertion. Bit error rate of hidden data extraction is reduced to zero by introducing a novel refinement procedure in the proposed algorithm. The errors are caused by the rounding process, the overflows and the nature of approximation in sparse decomposition. Capacity of the SDW-steg method is calculated using necessary conditions for uniqueness of the sparse representations and validated by means of experimental results. Imperceptibility of the method is at least 11 dB higher that the former methods. It is also more undetectable against five well-known steganalyzers comparing to other methods. The second steganography method is a spatial domain method. It is called SSRSteg-color method and uses matching pursuit method for sparse decomposition. It uses one random color band of cover image for dictionary estimation and inserts hidden bits into the signs of the sparse coefficients of two other color bands. We prove that using signs of the sparse coefficients for data embedding instead of their values, solves two main problems of sparse based steganography methods. Conducted experiments show the superiority of the proposed SSRSteg-color method in the senses of imperceptibility and undetectability comparing to former methods. It is also 260 times faster than the previous sparse based steganography method. Finally, we propose the DWT-SD method which is a sparse based speech steganography method. It uses high frequency components of the cover signal for dictionary estimation and inserts secret bits into the sparse coefficients of low frequency components which results in higher undetectability. The frames with lower energy than a predefined threshold are discarded from data embedding which leads to higher imperceptibility. It is proved that stego and cover signals are perceptually indistinguishable by using SNR and PESQ measures. Conducted experiments show superiority of the proposed SSRSteg-color method comparing to former methods in the senses of imperceptibility and capacity. Three proposed steganography methods prove efficiency of sparse decomposition for data hiding - Keywords:
- Watermarking ; Wavelet Transform ; Sparse Decomposition ; Steganography ; Steganalysis
Digital Object List
-
محتواي کتاب
- view
Bookmark