5 In this section is explained how CIs can be used as an information hiding scheme.
7 \subsection{Most and Least Significant Coefficients}\label{sec:msc-lsc}
11 We first take into account the fact that, in a watermarking process, terms of the original content $x$ that may be replaced by terms issued
12 from the watermark $y$ are less important than other: they could be changed
13 without be perceived as such. More generally, a
14 \emph{signification function}
15 attaches a weight to each term defining a digital media,
16 depending on its position $t$ \cite{todo}.
18 \begin{definition}[Signification function]
19 A \emph{signification function} is a real sequence
20 $(u^k)^{k \in \mathds{N}}$. % with a limit equal to 0.
24 \begin{example}\label{Exemple LSC}
25 Let us consider a set of
26 grayscale images stored into portable graymap format (P3-PGM):
27 each pixel ranges between 256 gray levels, \textit{i.e.},
28 is memorized with eight bits.
29 In that context, we consider
30 $u^k = 8 - (k \mod 8)$ to be the $k$-th term of a signification function
31 $(u^k)^{k \in \mathds{N}}$.
32 Intuitively, in each group of eight bits (\textit{i.e.}, for each pixel)
33 the first bit has an importance equal to 8, whereas the last bit has an
34 importance equal to 1. This is compliant with the idea that
35 changing the first bit affects more the image than changing the last one.
38 \begin{definition}[Significance of coefficients]
40 Let $(u^k)^{k \in \mathds{N}}$ be a signification function,
41 $m$ and $M$ be two reals s.t. $m < M$.
43 \item The \emph{most significant coefficients (MSCs)} of $x$ is the finite
44 vector $$u_M = \left( k ~ \big|~ k \in \mathds{N} \textrm{ and } u^k
45 \geqslant M \textrm{ and } k \le \mid x \mid \right);$$
46 \item The \emph{least significant coefficients (LSCs)} of $x$ is the
48 $$u_m = \left( k ~ \big|~ k \in \mathds{N} \textrm{ and } u^k
49 \le m \textrm{ and } k \le \mid x \mid \right);$$
50 \item The \emph{passive coefficients} of $x$ is the finite vector
51 $$u_p = \left( k ~ \big|~ k \in \mathds{N} \textrm{ and }
52 u^k \in ]m;M[ \textrm{ and } k \le \mid x \mid \right).$$
56 For a given host content $x$,
57 MSCs are then ranks of $x$ that describe the relevant part
58 of the image, whereas LSCs translate its less significant parts.
59 These two definitions are illustrated on Figure~\ref{fig:MSCLSC}, where the significance function $(u^k)$ is defined as in Example \ref{Exemple LSC}, $M=5$, and $m=6$.
64 \begin{minipage}[b]{.98\linewidth}
66 \centerline{\includegraphics[width=6.cm]{img/lena512}}
67 %\centerline{\epsfig{figure=img/lena512.pdf,width=4cm}}
68 \centerline{(a) Original Lena.}
70 \begin{minipage}[b]{.49\linewidth}
72 \centerline{\includegraphics[width=6.cm]{img/lena_msb_678}}
73 %\centerline{\epsfig{figure=img/lena_msb_678.pdf,width=4cm}}
74 \centerline{(b) MSCs of Lena.}
77 \begin{minipage}[b]{0.49\linewidth}
79 \centerline{\includegraphics[width=6.cm]{img/lena_lsb_1234_facteur17}}
81 %\centerline{\epsfig{figure=img/lena_lsb_1234_facteur17.pdf,width=4cm}}
82 \centerline{(c) LSCs of Lena ($\times 17$).}
85 \caption{Most and least significant coefficients of Lena.}
100 \subsection{Presentation of the Scheme}
102 We have proposed in \cite{guyeux10ter} to use chaotic iterations as an information hiding scheme, as follows.
105 \item $(K,N) \in [0;1]\times \mathds{N}$ be an embedding key,
106 \item $X \in \mathbb{B}^\mathsf{N}$ be the $\mathsf{N}$ LSCs of a cover $C$,% $X$ be the initial state $X_0$,
107 \item $(S^n)_{n \in \mathds{N}} \in \llbracket 0, \mathsf{N-1}
108 \rrbracket^{\mathds{N}}$ be a strategy, which depends on the message to hide $M \in [0;1]$ and $K$,
109 \item $f_0 : \mathbb{B}^\mathsf{N} \rightarrow \mathbb{B}^\mathsf{N}$ be the vectorial logical negation.
113 So the watermarked media is $C$ whose LSCs are replaced by $Y_K=X^{N}$, where:
119 \forall n < N, X^{n+1} = G_{f_0}\left(X^n\right).\\
124 Two ways to generate $(S^n)_{n \in \mathds{N}}$ are given in \cite{trx}, namely
125 Chaotic Iterations with Independent Strategy~(CIIS) and Chaotic Iterations with Dependent
127 In CIIS, the strategy is independent from the cover media $C$, whereas in CIDS the strategy will be dependent on $C$.
133 As we will use the CIIS strategy in this document, we recall it below.
134 Finally, MSCs are not used here, as we do not consider the case of authenticated watermarking.
136 \subsection{CIIS Strategy}
138 Let us firstly give the definition of the Piecewise Linear Chaotic Map~(PLCM, see~\cite{Shujun1}):
140 %\begin{definition}%[PLCM]
141 %The \emph{Piecewise Linear Chaotic Map} is defined by
145 x/p & \text{if} & x \in [0;p], \\
146 (x-p)/(\frac{1}{2} - p) & \text{if} & x \in \left[ p; \frac{1}{2} \right],
148 F(1-x,p) & \text{else,} & \\
154 \noindent where $p \in \left] 0; \frac{1}{2} \right[$ is a ``control parameter''.
156 The general term of the strategy $(S^n)_n$ in CIIS setup is defined by
157 the following expression: $S^n = \left \lfloor \mathsf{N} \times K^n \right \rfloor +
163 p \in \left[ 0 ; \frac{1}{2} \right] \\
165 K^{n+1} = F(K^n,p), \forall n \leq N_0, \end{array} \right.
170 \noindent in which $\otimes$ denotes the bitwise exclusive or (XOR) between two floating part numbers (\emph{i.e.}, between their binary digits representations).
173 \subsection{CIDS Strategy}
175 The same notations as above are used.
176 We define CIDS strategy as follows: $\forall k \leqslant N$,
178 \item if $k \leqslant \mathsf{N}$ and $X^k = 1$, then $S^k=k$,
181 In this situation, if $N \geqslant \mathsf{N}$, then only two watermarked contents are possible with the information hiding scheme proposed previously, namely: $Y_K=(0,0,\cdots,0)$ and $Y_K=(1,0,\cdots,0)$.
184 \subsection{PSNR evaluation}\label{section:psnr-ci-1}
186 To realize the evaluation of the PSNR of $CI_1$, we have used the same
187 architecture as described in
188 Section~\ref{section:architecture-presentation}~\vpageref{section:architecture-presentation}.
190 It has to be noted that in the $CI_1$ watermarking process, the embedding key
191 correspond to the strategy associated to the number of iterations.
193 Executing the $CI_1$ process on 1000 images, on the one hand with a constant
194 embedding key and one the other hand with a key generated randomly we have
195 obtain the results described in the
196 table~\ref{table:psnr-ci1}~\vpageref{table:psnr-ci1}, an average of and a standard deviation of has been obtained for the PSNR.\newline
203 \begin{tabular}{|c||c|c|c|}
205 \textbf{Strategy} & \textbf{PSNR average} & \textbf{PSNR standard
209 \textbf{Constant} & $21.6653$ & $0.03604$ \\
211 \textbf{Generated randomly} & $13.3909$ & $10.5690$ \\
216 \caption{Evaluation of the PSNR of the watermarking process $CI_1$.(Tests
217 realized on 1000 images)}
219 \label{table:psnr-ci1}
222 Let us now focus on some security aspects of information hiding schemes. In particular, we aim to explain why chaos is relevant in this field.