24-01-2013, 10:07 AM
A steganographic method for digital images using side match
1A steganographic method.pdf (Size: 447.07 KB / Downloads: 27)
Abstract
In order to provide larger embedding capacity and to minimize the distortion of the stego-image, a novel steganographic
method using side information is presented in this paper. The method exploits the correlation between
neighboring pixels to estimate the degree of smoothness or contrast of pixels. If the pixel is located in edge area, then it
may tolerate larger changes than those in smooth areas. The two-sided, three-sided, and four-sided side match methods
are employed in our scheme. The experimental results show that our method provides a large embedding capacity
without making noticeable distortion. Besides, the embedded data can be extracted from the stego-image without
referencing the original image.
Introduction
Internet is a popular communication channel
nowadays. However, message transmissions over
the Internet still have to face some problems, such
as data security, copyright control, etc. Thus we
need secure secret communication methods for
transmitting message over the Internet. Encryption
is a well-known method for security protection,
which refers to the process of encoding secret
information in such a way that only the person
with the right key can successfully decode it.
The proposed method
The concept of side match is proposed by Kim
(1992). The side match vector quantization
(SMVQ) is an effective VQ coding scheme at low
bit rate. It utilizes the information of two neighboring
blocks (upper block and left block) to
predict the state codebook of an input vector. The
state codebook is partially selected from a common
codebook (the so-called super codebook).
Then the index size is reduced and higher compression
ratio is achieved.
Three-sided side match steganography
The three-sided side match utilizes not only the
upper and left pixels but also one of the other
neighboring pixels (right or bottom). A 3:1 sampling
arrangement is shown in Fig. 2 and around
33.3% of the pixels will be abandoned for embedding
data. This figure is an example of 88 image,
the shaded pixels remain unchanged; the blank
pixels are used for embedding the secret data;
the pixels marked with ‘‘’’ are pre-embedded
pixels.
Conclusions
In this paper, we have proposed a novel and
efficient steganographic method to embed secret
information into images without producing perceptible
distortions. There is no need of referencing
the original image when extracting the
embedded data from a stego-image. The method
utilizes the side information to estimate the
amount of data that can be embedded into an input
pixel of cover image. The pixels in edge areas
may embed more data than those in non-edge
areas. It is clear that our method performs better
than conventional LSBs substitution method in
both visual effect and security. Our experimental
results have shown that the proposed method
provides a better way for embedding large amount
of data into cover images without making noticeable
distortions.