1 For whole experiments, the whole set of 10000 images
2 of the BOSS contest~\cite{Boss10} database is taken.
3 In this set, each cover is a $512\times 512$
4 grayscale digital image in a RAW format.
5 We restrict experiments to
6 this set of cover images since this paper is more focussed on
7 the methodology than benchmarking.
8 Our approach is always compared to Hugo~\cite{DBLP:conf/ih/PevnyFB10}
9 and to EAISLSBMR~\cite{Luo:2010:EAI:1824719.1824720}.
15 \subsection{Adaptive Embedding Rate}
16 Two strategies have been developed in our scheme,
17 depending on the embedding rate that is either \emph{adaptive} or \emph{fixed}.
19 In the former the embedding rate depends on the number of edge pixels.
20 The higher it is, the larger the message length that can be inserted is.
21 Practically, a set of edge pixels is computed according to the
22 Canny algorithm with an high threshold.
23 The message length is thus defined to be half of this set cardinality.
24 In this strategy, two methods are thus applied to extract bits that
25 are modified. The first one is a direct application of the STC algorithm.
26 This method is further referred to as \emph{adaptive+STC}.
27 The second one randomly chooses the subset of pixels to modify by
28 applying the BBS PRNG again. This method is denoted \emph{adaptive+sample}.
29 Notice that the rate between
30 available bits and bit message length is always equal to 2.
31 This constraint is indeed induced by the fact that the efficiency
32 of the STC algorithm is unsatisfactory under that threshold.
33 In our experiments and with the adaptive scheme,
34 the average size of the message that can be embedded is 16,445 bits.
35 Its corresponds to an average payload of 6.35\%.
40 In the latter, the embedding rate is defined as a percentage between the
41 number of modified pixels and the length of the bit message.
42 This is the classical approach adopted in steganography.
43 Practically, the Canny algorithm generates
44 a set of edge pixels related to a threshold that is decreasing until its cardinality
45 is sufficient. If the set cardinality is more than twice larger than the
46 bit message length, a STC step is again applied.
47 Otherwise, pixels are again randomly chosen with BBS.
51 \subsection{Image Quality}
52 The visual quality of the STABYLO scheme is evaluated in this section.
53 For the sake of completeness, four metrics are computed in these experiments:
54 the Peak Signal to Noise Ratio (PSNR),
55 the PSNR-HVS-M family~\cite{psnrhvsm11},
56 %the BIQI~\cite{MB10},
58 the weighted PSNR (wPSNR)~\cite{DBLP:conf/ih/PereiraVMMP01}.
59 The first one is widely used but does not take into
60 account the Human Visual System (HVS).
61 The other ones have been designed to tackle this problem.
68 \begin{tabular}{|c|c|c||c|c|c|c|c|}
70 Schemes & \multicolumn{3}{|c|}{STABYLO} & \multicolumn{2}{|c|}{HUGO}& \multicolumn{2}{|c|}{EAISLSBMR} \\
72 Embedding & \multicolumn{2}{|c||}{Adaptive} & Fixed & \multicolumn{2}{|c|}{Fixed}& \multicolumn{2}{|c|}{Fixed} \\
74 Rate & + STC & + sample & 10\% & 10\%&6.35\%& 10\%&6.35\%\\
76 PSNR & 66.55 (\textbf{-0.8\%}) & 63.48 & 61.86 & 64.65 & {67.08} & 60.8 & 62.9\\
78 PSNR-HVS-M & 78.6 (\textbf{-0.8\%}) & 75.39 & 72.9 & 76.67 & {79.23} & 61.3 & 63.4\\
80 %BIQI & 28.3 & 28.28 & 28.4 & 28.28 & 28.28 & 28.2 & 28.2\\
82 wPSNR & 86.43(\textbf{-1.6\%}) & 80.59 & 77.47& 83.03 & {87.8} & & 80.6\\
86 \caption{Quality Measures of Steganography Approaches\label{table:quality}}
92 Results are summarized into the Table~\ref{table:quality}.
93 Let us give an interpretation of these first experiments.
94 First of all, the adaptive strategy produces images with lower distortion
95 than the one of images resulting from the 10\% fixed strategy.
96 Numerical results are indeed always greater for the former strategy than
98 These results are not surprising since the adaptive strategy aims at
99 embedding messages whose length is decided according to an higher threshold
100 into the edge detection.
101 Let us focus on the quality of HUGO images: with a given fixed
102 embedding rate (10\%),
103 HUGO always produces images whose quality is higher than the STABYLO's one.
104 However our appraoch always outperforms EAISLSBMR since this one may modify
105 the two least significant bits whereas STABYLO only alter LSB.
107 If we combine \emph{adaptive} and \emph{STC} strategies
108 (which leads to an average embedding rate equal to 6.35\%)
109 our approach provides equivalent metrics than HUGO.
110 The quality variance between HUGO and STABYLO for these parameters
111 is given in bold font. It is always close to 1\% which confirms
112 the objective presented in the motivations:
113 providing an efficient steganography approach with a lightweight manner.
116 Let us now compare the STABYLO approach with other edge based steganography
117 approaches, namely~\cite{DBLP:journals/eswa/ChenCL10,Chang20101286}.
118 These two schemes focus on increasing the
119 payload while the PSNR is acceptable, but do not
120 give quality metrics for fixed embedding rates from a large base of images.
125 \subsection{Steganalysis}
129 The quality of our approach has been evaluated through the two
130 AUMP~\cite{Fillatre:2012:ASL:2333143.2333587}
131 and Ensemble Classifier~\cite{DBLP:journals/tifs/KodovskyFH12} based steganalysers.
132 Both aims at detecting hidden bits in grayscale natural images and are
133 considered as the state of the art of steganalysers in spatial domain~\cite{FK12}.
134 The former approach is based on a simplified parametric model of natural images.
135 Parameters are firstly estimated and an adaptive Asymptotically Uniformly Most Powerful
136 (AUMP) test is designed (theoretically and practically), to check whether
137 an image has stego content or not.
138 This approach is dedicated to verify whether LSB has been modified or not.
139 In the latter, the authors show that the
140 machine learning step, which is often
141 implemented as support vector machine,
142 can be favorably executed thanks to an ensemble classifier.
144 %citer le second tableau, comparer avec EAISLSBMR
149 \begin{tabular}{|c|c|c|c|c|c|c|c|}
151 Schemes & \multicolumn{3}{|c|}{STABYLO} & \multicolumn{2}{|c|}{HUGO}& \multicolumn{2}{|c|}{EAISLSBMR}\\
153 Embedding & \multicolumn{2}{|c|}{Adaptive} & Fixed & \multicolumn{2}{|c|}{Fixed}& \multicolumn{2}{|c|}{Fixed} \\
155 Rate & + STC & + sample & 10\% & 10\%& 6.35\%& 10\%& 6.35\%\\
157 AUMP & 0.39 & 0.33 & 0.22 & 0.50 & 0.50 & 0.49 & 0.50 \\
159 Ensemble Classifier & \textbf{0.47} & 0.44 & 0.35 & 0.48 & 0.49 & 0.43 & 0.46 \\
165 \caption{Steganalysing STABYLO\label{table:steganalyse}}
169 Results show that our approach is more easily detectable than HUGO, which
170 is the most secure steganographic tool, as far as we know. However due to its
171 huge number of features integration, it is not lightweight, which justifies
172 in the authors' opinion the consideration of the proposed method.