CHAPTER 2 REVIEW OF LITERATUREshodhganga.inflibnet.ac.in/bitstream/10603/9089/6/06_chapter 2.pdf ·...
Transcript of CHAPTER 2 REVIEW OF LITERATUREshodhganga.inflibnet.ac.in/bitstream/10603/9089/6/06_chapter 2.pdf ·...
Information Hiding in Image and Audio Files 2007-2010
19
CHAPTER 2
REVIEW OF LITERATURE
Information hiding techniques can be classified into two categories,
namely irreversible information hiding schemes [26, 29, 30] and
reversible (lossless) information hiding schemes [32]. In irreversible
information hiding schemes, only secret data are extracted and no
restoration of cover objects is made. In contrast, in reversible
information hiding schemes, secret data are extracted and cover
objects can be completely restored. Reversible hiding schemes are
suitably used for some applications such as the healthcare industry
and online content distribution systems. In this thesis information
hiding is done using irreversible information hiding schemes.
2.1. Domain of Information Hiding
The existing schemes of information hiding in images and audio can
roughly be classified into the following three categories:
Spatial domain / Time domain
Transform domain
Compressed domain
2.1.1 Image files
(I)Spatial domain Information hiding
Data hiding in spatial domain [33-35] type directly adjust image
pixels in the spatial domain for data embedding. This technique is
simple to implement, offering a relatively high hiding capacity. The
quality of the Stego image can be easily controlled. Therefore, data
hiding of this type has become a well known method for image
steganography.
Information Hiding in Image and Audio Files 2007-2010
20
(a)Least Significant Bit Techniques
A large number of commercial steganographic programs use the
Least Significant Bit embedding (LSB) as the method of choice for
message hiding in 24-bit, and 8-bit images. The LSB insertion
method is a common, simple approach to embedding information in
a graphical image file. Unfortunately, it is extremely vulnerable to
attacks, such as any image manipulation.
The terminology LSB replacement/ LSB matching was discussed by
T.Sharp [14]. LSB substitution [27] algorithm is the simplest
scheme to hide message in a host image. It replaces the LSB of
each pixel with the encrypted message bit stream. Authenticated
receivers can extract the message by deciphering the LSB of every
pixel of the host image with a pre-shared key. Since only the LSB of
pixels is altered, it is visually imperceptible by human eye. The
capacity of the algorithm is 1 bit per pixel (bpp). Although this
algorithm is visually imperceptible, it can be statistically analyzed by
other entity without processing the pre-shared key.
The LSB insertion method is the most common and easiest method
for embedding messages in an image with high capacity, while it is
detectable by statistical analysis such as Regular and Singular (RS)
and Chi-Square analyses. The method proposed by Hong-juan
zhang et al., [37] is a novel LSB image steganography algorithm
that can effectively resist image Steganalysis[28,96,97,98,99]
based on statistical analysis. Every two sample‟s of LSB bits are
combined using addition modulo 2 (or m) to form the value which is
compared to the part of the secret message. If these two values are
equal, no change is made. Otherwise, the difference of these two
values is added to the second sample. Thus, the part of the secret
message can be embedded effectively. RS and Chi-Square analyses
are performed on stego-medium created using the steganography
Information Hiding in Image and Audio Files 2007-2010
21
technique. The experiment showed that the scheme has the same
insertion capacity and Signal to Noise Ratio (SNR) as the classic LSB
steganography and the proposed method can effectively resist
steganalysis based on RS and Chi-Square analyses.
A similar idea called LSB matching has been proposed by Ker [38]
to improve LSB substitution. Yet it is also vulnerable to other
designated detection algorithms. This is owing to the histogram of
the host image is changed. A revised version of LSB matching is
proposed by Mielikainen [39] in 2006. This method greatly improves
the above two methods by lowering the expected number of
modifications per pixel, from 0.5 to 0.375. Therefore, the histogram
affected by the scheme is less significant. Only a few detection
methods for LSB matching have been proposed.
In another data hiding technique by variable depth LSB
substitutions proposed by Smo-Huliiv et al., [40], pixels of cover-
image are grouped according to themselves by luminance values.
Then the occurrence of pixel-values in each group is counted and
sorted in monotone increasing. Finally a bit plane-wise data hiding
method is used to hide data in cover image. And the worst-square-
error between the stego-image and the cover-image is formulated;
Theory analysis indicates that image quality of the stego-image
hidden by this method can improve from 0 decibels (db) to 4db
against simple LSB substitution method [40].
A novel approach of image embedding was introduced by Rehab et
al., [41]. The method consists of three main steps. First, the edge
of the image is detected using Sobel mask filters. Second, the LSB
of each pixel is used. Finally, a gray level connectivity is applied
using a fuzzy approach and the American Standard Code for
Information Hiding in Image and Audio Files 2007-2010
22
Information Interchange (ASCII) code is used for information
hiding.
A novel data hiding technique is proposed by Dey et al., [42], as an
improvement over the Fibonacci LSB data-hiding technique. The
novel technique is based on decomposition of a number (pixel-
value) in sum of prime numbers. The particular representation
generates a different set of (virtual) bit-planes altogether, suitable
for embedding purposes. They not only allow one to embed secret
message in higher bit planes but also do it without much distortion,
with a much better stego-image quality, and in a reliable and
secured manner, guaranteeing efficient retrieval of secret message.
A comparative performance study between the classical LSB
method, the Fibonacci LSB data-hiding technique and these
proposed schemes has been done. Analysis indicates that image
quality of the stego-image hidden by the technique using Fibonacci
decomposition improves against that using simple LSB substitution
method, while the same using the prime decomposition method
improves drastically against that using Fibonacci decomposition
technique. Experimental results have shown that, the stego-image
is visually indistinguishable from the original cover-image.
Daniela et al., [43] used the LSB technique in YUV color space to
embed the secret message. The secret message was hidden in the V
plane and the image was converted back to RGB to obtain the stego
image.
A new method using the RBTC and LSB substitution to hide data is
proposed by Ching-Yu Yang [44]. Based on the RBTC, a simple LSB
substitution is employed for embedding secret data to the resulting
compressed images. Experiments show that the PSNR and hiding
Information Hiding in Image and Audio Files 2007-2010
23
rate for the proposed method are good while the perceived quality
is not bad. Since the perceptual quality of the mixed images are
close to that of compressed images generated by the RBTC without
hiding capability, the proposed method possesses a merit of being
attracted hardly by the grabbers. In other words, the grabbers do
not easily notice the existence of the embedded message.
An effective color image steganography method based on the
module substitutions is proposed by Ching-Yu Yang [45]. In
accordance with the base-value of the blocks, a variety of secret
bits is embedded to a RGB trichromatic. system by three types of
module substitutions More specifically, to alleviate further color
distortion and obtain a larger hidden capacity, the R-, G-. and B-
components are encoded by Mod u, Mod u-v, and Mod u-v-w
substitution, respectively. Experiments show that both PSNR and
hiding rate generated by the proposed method are better than
those generated by the reported schemes. In addition, the resulting
perceptual quality is good.
Block Truncation Coding (BTC) is an efficient compression technique
while offering good image quality. Nonetheless, the blocking effect
inherent in BTC causes severe perceptual artifact in high
compression ratio applications. In this proposed method by Jing-
Ming Guo [46], an Error-Diffused Block Truncation Coding (EDBTC)
is proposed to solve this problem. According to the EDBTC, the error
caused by the difference between the original grayscale pixel value
and the correspondingly high or low mean substitute is diffused to
the predefined neighborhood, and hence the average grayscale will
be maintained invariably. In addition, since the compressed data
are widely distributed in the internet transmission, the extra
message conveying in a secret way also highly raises attention.
Recently, in this method, the Complementary Steganography in
Information Hiding in Image and Audio Files 2007-2010
24
Error-Diffused Block Truncation Coding (CSEDBTC), which
cooperates with error diffusion to achieve the objective of secret
communication in BTC images, is proposed. As documented in the
experimental results, a low complexity with good image quality is
obtained.
A novel data-hiding methodology proposed by Chun-Hsiang Huang
et al., [47], denoted as digital invisible ink (DII), is proposed to
implement secure steganography systems. Like the real-world
invisible ink, secret messages will be correctly revealed only after
the marked works undergo certain pre-negotiated manipulations,
such as lossy compression and processing. Different from
conventional data-hiding schemes where content processing or
compression operations are undesirable, distortions caused by pre-
negotiated manipulations in DII-based schemes are indispensable
steps for revealing genuine secrets.
The scheme proposed in [47] is carried out based on two important
data-hiding schemes: spread-spectrum [95] watermarking and
frequency-domain quantization watermarking. In some application
scenarios, the DII-based steganography system can provide
plausible deniability and enhance the secrecy by taking cover with
other messages. It is stated that DII-based schemes are indeed
superior to existing plausibly deniable steganography approaches in
many aspects. Moreover, potential security holes caused by
deniable steganography systems are discussed.
In the LSB technique the retrieval of the secret message can be
done by retrieving the LSB of the pixel value. From the security
aspect this technique is very fragile. To overcome this drawback, a
new robust technique is proposed by Kekre et al.,[48], in which
three different variations of LSB technique are proposed. These
Information Hiding in Image and Audio Files 2007-2010
25
proposed techniques do not directly replace the LSB but LSB is
modified by taking into consideration the magnitude of the Cover
image.
(b)Increasing Capacity Techniques
In the LSB technique, only the LSB of every pixel value is utilized
for secret message embedding, this results in a constraint to the
embedding capacity in the cover image. In this section, the
techniques proposing an increase in the embedding capacity are
reviewed.
A new and efficient steganographic method for embedding secret
messages into a gray-valued cover image was proposed by Wu et
al., [36]. In the process of embedding a secret message, a cover
image is partitioned into non-overlapping blocks of two consecutive
pixels. A difference value is calculated from the values of the two
pixels in each block. All possible difference values are classified into
a number of ranges. The selection of the range intervals is based on
the characteristics of human vision‟s sensitivity to gray value
variations from smoothness to contrast. The difference value then is
replaced by a new value to embed the value of a sub-stream of the
secret message. The number of bits which can be embedded in a
pixel pair is decided by the width of the range that the difference
value belongs to. The method is designed in such a way that the
modification is never out of the range interval. This method
provides an easy way to produce a more imperceptible result than
those yielded by simple LSB replacement methods. The embedded
secret message can be extracted from the resulting stego-image
without referencing the original cover image.
In order to improve the capacity of the hidden secret data and to
provide an imperceptible stego-image quality, a novel
Information Hiding in Image and Audio Files 2007-2010
26
steganographic method based on LSB replacement and Pixel Value
Differencing (PVD) method was presented by Wu et al., [36]. First,
a difference value from two consecutive pixels by utilizing the PVD
method is obtained. A small difference value can be located on a
smooth area and the large one is located on an edged area. In the
smooth areas, the secret data is hidden into the cover image by LSB
method while using the PVD method in the edged areas. Because
the range width is variable the area in which the secret data is
concealed by LSB or PVD method is hard to guess. The security
level is the same as that of a single LSB using the proposed PVD
method. From the experimental results, compared with the PVD
method being used alone, the proposed method can hide much
larger information and maintains a good visual quality of stego-
image.
A high quality steganographic technique is proposed by Wang et al.,
[50] with PVD and modulus function capable of producing a stego
image that is imperceptible from the original image by the human
eye. In addition, the method avoids the falling-off-boundary
problem by using PVD and the modulus function. First, a difference
value from two consecutive pixels is calculated using the PVD
technique. The hiding capacity of the two consecutive pixels
depends on the difference value. In other words, the smoother the
area is, the less secret data can be hidden; on the contrary, the
more edges an area has, the more secret data can be embedded.
This way, the stego-image quality degradation is more
imperceptible to the human eye. Second, the remainder of the two
consecutive pixels can be computed by using the modulus
operation, and then secret data can be embedded into the two
pixels by modifying their remainder. In this scheme, there is an
optimal approach to alter the remainder so as to greatly reduce the
image distortion caused by the hiding of the secret data. The values
Information Hiding in Image and Audio Files 2007-2010
27
of the two consecutive pixels are scarcely changed after the
embedding of the secret message by the proposed optimal
alteration algorithm. Experimental results have demonstrated that
the proposed scheme is secure against the RS detection attack.
A new adaptive LSB Steganography method using PVD that provides
a larger embedding capacity and imperceptible stego images was
proposed by Yang et al., [51]. The method exploits the difference
value of two consecutive pixels to estimate how many secret bits
will be embedded into the two pixels. Pixels located in the edge
areas are embedded by a k-bit LSB substitution method with a
larger value of k than that of the pixels located in smooth areas.
The range of difference values is adaptively divided into lower level,
middle level, and higher level. For any pair of consecutive pixels,
both pixels are embedded by the k-bit LSB substitution method.
However, the value k is adaptive and is decided by the level which
the difference value belongs to. In order to remain at the same
level where the difference value of two consecutive pixels belongs,
before and after embedding, a delicate readjusting phase is used.
When compared to the past study of Wu et al.'s PVD and LSB
replacement method, experimental results show that proposed
approach provides both larger embedding capacity and higher
image quality.
The main requirements for a steganographic scheme are visual
imperceptibility and statistical undetectability. Steganographic
techniques which exhibit these two qualities have lesser embedding
capacity. So to achieve statistical undetectability with higher
embedding efficiency, matrix embedding proposed by Arjun et al.,
[52] is preferred. The Embedding efficiency is defined as the ratio of
change density to embedding rate. The change density using matrix
embedding is defined as the number of modifications that are
Information Hiding in Image and Audio Files 2007-2010
28
needed per code sequence of length n, to embed a k bit message.
And embedding rate is defined as ratio of the number of message
bits k, which can be embedded in a code sequence n, with a single
modification. In [100] Cover is chosen such that it is correlated with
the covert messages thus reducing the biits needed to encode the
hidden message
A method of implementing image steganography in a color image
for applications such as covert communication via the Internet and
authentication of an employee carrying a picture identification card
is described by Kaliappan Gopalan [53]. By converting the color
image to a one-dimensional signal in red, green, or blue, audibly
masked frequencies in the one Dimensional (1-D) signal are
determined for each segment or block. Embedding of secure or
confidential data is carried out by modifying the spectral power at a
pair of commonly occurring masked frequencies. Compressing the
data-embedded image to standard Joint Photographic Experts
Group (JPEG) coding enables its transmission via the Internet.
Experimental results show that the technique is simple to
implement and causes barely noticeable distortion in the stego
image. Using an oblivious technique and a key consisting of the
frequencies where the spectrum is modified, successful data
retrieval even at low level noise levels and at low-loss compression
has been achieved. Higher payload of hidden data can be obtained
at a cost of perceptibility of embedding. Lossy JPEG compression,
however, leads to low payloads.
(II) Transform domain Information hiding:
In this method [54, 55] images are first transformed into transform
domain, and then data is embedded by modifying the transformed
coefficient
The spatial domain techniques embed the secret data within the
pixels of the cover image. The LSB (Least Significant Bit)
Information Hiding in Image and Audio Files 2007-2010
29
substitution is an example of spatial domain techniques. The other
type of hiding method is the transform domain techniques which
embed the secret data within the cover image that has been
transformed using some transforms such as DCT (Discrete Cosine
Transform), DWT (Discrete Wavelet Transform), FFT (Fast Fourier
Transform) etc. The DCT or DWT coefficients of the transformed
cover image will be quantized and then modified according to the
secret data to be embedded.
There are many transforms that can be used in data hiding. The
most commonly used is Discrete Cosine Transform ( DCT) which is
used in the common image compression format JPEG and MPEG
[56], The discrete wavelet transform (DWT) [94] now a days is
used in new JPEG2000.
(a)An Adaptive Steganography technique based on integer
wavelet transforms
An Adaptive steganography technique based on integer wavelet
transform is presented in [57]. This technique tries to optimize high
hiding capacity and imperceptibility by proposing a novel technique
for hiding data in digital images. This is done by combining the use
of adaptive hiding capacity function that hides secret data in the
integer wavelet coefficients of the cover image with the OPA
(Optimum pixel Adjustment) algorithm. The coefficients are selected
according to the pseudo random generator to increase the security
of the hidden data. The OPA algorithm is applied after embedding
the secret data to minimize the embedding error.
(b)Secure Blind Image Steganography Technique using
Discrete Fourier Transformation
Another Image Steganography technique has been proposed using
DFT in [58].The technique embeds the hidden information in the
Information Hiding in Image and Audio Files 2007-2010
30
DFT domain after permuting the image pixels in the spatial domain
using a key. The permutation process introduces randomness into
the cover image and results in a significant increase in the number
of transform coefficients that can be used to transmit the hidden
information. The information is embedded using quantization
technique.
(c) High Capacity Steganographic Method Based Upon JPEG
JPEG technique divides the input image into non-overlapping blocks
of 8x8 pixels and uses the DCT. The method discussed in [56]
divides the cover image into non-overlapping blocks of 16x16
pixels. For each quantized DCT block, the least two significant bits
of each middle frequency coefficients are modified to embed two
secret bits. The hiding technique proposed in [56] achieves better
hiding capacity than Jpeg-Jsteg methods which are based on the
conventional blocks of 8x8 pixels.
(d)An Effective Image Steganographic Scheme Based on
Wavelet Transformation and Pattern - Based Modification
This scheme [160] also hides a secret message in a digital image.
The scheme is called Pattern Based Image Steganography (PBIS).
First PBIS apply Discrete Wavelet Transform on the cover image;
separates the transformation result into non-overlapping blocks.
PBIS uses a secret key and a pseudo random number generator to
select some blocks. The scheme analyzes the pattern types of the
coefficients of the selected blocks and change the pattern types of
these selected blocks according to the secret message Finally the
DWT is applied to transform the wavelet coefficients to their spatial
domain and obtain the stego. Advantage of this scheme is that it
can survive under JPEG lossy compression.
Information Hiding in Image and Audio Files 2007-2010
31
(e) High Capacity Image Steganography using Wavelet-
Based Fusion
The main idea of the proposed algorithm presented in [161] is the
wavelet based fusion. It involves merging of the wavelet
decomposition of the normalized versions of both the cover image
and the secret image into a single fused result. In a normalized
image the pixel components take on values that span a range
between 0.0 and 1.0 instead of integer ranges of [0-255]. Hence
the corresponding wavelet coefficients also range between 0.0 to
1.0. Before the embedding process, the system does a
preprocessing step on the cover image. This step involves shrinking
the range of the pixel color components in order to avoid any range
violation due to the embedding process. This ensures that the
embedding message will be recovered correctly. The extraction
process involves subtracting the original cover image from the stego
image in the wavelet domain to get the coefficients of the secret
message. Then the embedded message is retrieved by applying
inverse discrete wavelet transform (IDWT).
(III) Compressed domain Information hiding: Information
hiding in compressed domain [32, 59] is obtained by modifying the
coefficients of the compressed code of a cover image. Since most
images transmitted over Internet are in compressed format,
embedding secret data into the compressed domain would provoke
little suspicion.
(a)Vector Quantization for Information Hiding:
One of the most commonly studied image compression techniques
is Vector Quantization (VQ) [117-133] [60], which is an attractive
choice because of its simplicity and cost-effective implementation.
Indeed, a variety of VQ techniques have been successfully applied
in real applications such as speech and image coding [61, 62].
Information Hiding in Image and Audio Files 2007-2010
32
Vector quantization has been very popular in variety of research
fields such as video-based event detection [133], speech data
compression [132], image segmentation [134-45], face recognition
[146, 147, 148], iris recognition [149, 150, 151], CBIR [152-155],
Colorization [156, 157] data hiding [158,159] etc. Vector
Quantization not only has faster encode/decode time and a simpler
framework than JPEG/JPEG2000 but it also requires limited
information during decoding, and those advantages cost VQ a little
low compression ratio and visual quality. VQ works best in
applications in which the decoder has only limited information and a
fast execution time is required [63].
The key point to the design of a perfect VQ scheme is to generate a
perfect codebook from the training images. The LBG algorithm,
proposed by Linde, Buzo and Gray in 1980 [64], gives a good
answer and is probably the most famous codebook design
algorithm. However, VQ still has its limitations. It usually generates
visible boundaries between blocks since the current block is coded
independently of its neighboring blocks. To deal with the above
problem, side match vector quantization (SMVQ) was proposed by
Kim in 1992 [65]. Kim successfully reduces the blocking effect by
using local edge information and provides better visual quality and
compression ratio than VQ does. Then, to make data hiding more
convenient, some researchers have tried to hide secret data in
cover images already compressed by VQ or SMVQ [66].
In [67], Lin et al. presented a method of embedding that was based
on VQ compressed images. The approach involves reducing the size
of the codebook and placing data in the remaining spaces of index
values. A codebook is first partitioned into two sub-codebooks such
that all pairs of the corresponding code vectors between sub
codebooks are as similar as possible. Any modification of the least
Information Hiding in Image and Audio Files 2007-2010
33
significant bits of index values does not markedly distort the
reconstructed image, because the two sub codebooks have similar
content. Accordingly, secret data can be placed into the LSB of all
indices. In [68], Lu and Sun presented a similar method, but
extended to partition the codebook into 2k sub-codebooks for
embedding k bits into a single index.
VQ consists of three phases:
Codebook design
Encoding and
Decoding
Vector quantization can be defined as a mapping function that maps
k-dimensional vector space to a finite set CB = {C1, C2, C3, ..….,
CN}. The set CB is called codebook consisting of N number of
codevectors and each codevector Ci = {ci1, ci2, ci3, ……, cik} is of
dimension k. The key to VQ is the good codebook. Codebook can be
generated in spatial domain by using clustering.algorithms
Following codebook generation algorithms are used for data hiding.
Linde Buzo Gray (LBG)[64]
Kekre‟s Proportionate Error algorithm (KPE) [123]
Kekre‟s Median Codebook Generation algorithm (KMCG) [123]
Kekre‟s Fast Codebook Generation algorithm (KFCG) [125,
127]
Since the codebook is used for hiding data security is at a higher
level.
Information Hiding in Image and Audio Files 2007-2010
34
2.1.2 Audio Files
(I)Overview of Human Auditory System (HAS)
Data hiding in audio signals is especially challenging as compared to
data hiding in digital images because the human auditory system
(HAS) operates over a wide dynamic range in comparison with
human visual system (HVS). Sensitivity to additive random noise is
also acute [31]. On the other hand, opposite to its large dynamic
range, HAS contains a fairly small differential range, i.e., loud
sounds generally tend to mask our weaker sounds [108].
Additionally, the HAS is unable to perceive absolute phase, only
relative phase [31]. Finally, there are some environmental
distortions so common as to be ignored by the listener in most of
the cases [31]. Such are the weaknesses of HAS that can be
exploited for hiding data in audio signals.
The effects of human auditory system (HAS) relative to
Steganography are temporal masking and frequency masking. In
temporal masking, a weaker audible signal on either side (pre and
post) of a strong masker becomes imperceptible. Similarly, in
frequency masking, if two signals occurring simultaneously are close
together in frequency, the stronger masking signal may make the
weaker signal inaudible [49].
(II)Methods of Audio Steganography
In the past few years, several algorithms for the embedding and
extraction of messages in audio sequences have been proposed. All
of the developed algorithms take advantage of the perceptual
properties (characteristics) of the human auditory system (HAS) in
order to hide data into the host signal in a perceptually transparent
manner. Some commonly used methods are [102]:
Least Significant Bit (LSB) Coding
Information Hiding in Image and Audio Files 2007-2010
35
Parity Coding
Phase Coding
Spread Spectrum
Echo Data Hiding
(a)Least Significant Bit (LSB) Coding:
One of the earliest techniques studied in the information hiding of
digital audio (as well as other media types) is LSB coding. In this
technique, LSB of binary sequence of each sample of digitized audio
file is replaced with binary equivalent of secret message [102].
For example the letter „A‟ (binary equivalent 1000001) is to be
hidden into a digitized audio file where each sample is represented
with 16 bits, then LSB of 7 consecutive samples (each of 16 bit size)
is replaced with each bit of binary equivalent of the letter „A‟ [112]
as illustrated in Table 2.1.
Table 2.1 Example of Least Significant Bit (LSB) Coding
Sampled Audio Stream (16 bits)
‘A’ in binary
Audio stream with encoded message
1001 1000 0011 1100 1 1001 1000 0011 1101
1101 1011 0011 1000 0 1101 1011 0011 1000
1011 1100 0011 1101 0 1011 1100 0011 1100
1011 1111 0011 1100 0 1011 1111 0011 1100
1011 1010 0111 1111 0 1011 1010 0111 1110
1111 1000 0011 1100 0 1111 1000 0011 1100
1101 1100 0111 1000 1 1101 1100 0111 1001
Information Hiding in Image and Audio Files 2007-2010
36
The number of LSB‟s for data hiding can be increased, but it also
increases the amount of resulting noise in the audio file as well.
Thus, one should consider the signal content before deciding on the
LSB operation to use. For example, a sound file that was recorded
in a bustling subway station would mask low-bit encoding noise. On
the other hand, the same noise would be audible in a sound file
containing a piano solo [102],[118]. To extract a secret message
from an LSB encoded sound file, the receiver needs access to the
sequence of sample indices used in the embedding process.
Advantages: (i) It is the simplest way to embed information in a
digital audio file. It allows large amount of data to be concealed
within an audio file, use of only one LSB of the host audio sample
gives a capacity equivalent to the sampling rate which could vary
from 8 kbps to 44.1 kbps (all samples used) [113]. (ii) It is widely
used as modification to LSB‟s usually does not create audible
changes to the sound files.
Disadvantages: (i) The Human ear is very sensitive and can often
detect even the slightest bit of noise introduced into a sound file.
(ii) It has considerably low robustness against attacks. For example,
if a sound file embedded with a secret message using LSB coding is
resampled, the embedded information would be lost [118].
(b)Parity Coding:
Instead of breaking a signal down into individual samples, the parity
coding method [102] breaks a signal down into separate regions of
samples and encodes each bit from the secret message in a sample
region's parity bit. If the parity bit of a selected region does not
match the secret bit to be encoded, the process flips the LSB of one
of the samples in the region. Thus, the sender has more of a choice
Information Hiding in Image and Audio Files 2007-2010
37
in encoding the secret bit. Figure 2.1 illustrates how the first three
bits of the message 'HEY' are encoded with even parity.
Figure 2.1 Example of Parity Coding
The decoding process extracts the secret message by calculating
and lining up the parity bits of the regions used in the encoding
process. This method also has disadvantages similar to the LSB
coding method although the parity coding method does come much
closer in making the introduced noise inaudible.
(c)Phase coding:
Phase coding [102] is much more complex method than the
simplistic LSB encoding. Phase encoding “works by substituting the
phase of an initial audio segment with a reference phase that
represents the data. The phase of subsequent segments is adjusted
in order to preserve the relative phase between segments”. Phase
Information Hiding in Image and Audio Files 2007-2010
38
coding relies on the fact that the phase components of sound are
not as perceptible to the human ear as noise is. Rather than
introducing perturbations, the technique encodes the message bits
as phase shifts in the phase spectrum of a digital signal, achieving
an inaudible encoding in terms of signal-to-perceived noise ratio.
Original and encoded signal are as shown in Figure 2.2. This method
is used when only a small amount of data, such as a watermark,
needs to be concealed.
Figure 2.2 The signals before and after Phase coding procedure
(d)Spread Spectrum (SS):
It attempts to spread out the encoded data across the available
frequencies as much as possible. This is analogous to a system
using an implementation of the LSB coding that randomly spreads
the message bits over the entire sound file. However, unlike LSB
coding, the Spread S method spreads the secret message over the
sound file‟s frequency spectrum, using a code that is independent of
the actual signal. As a result, the final signal occupies a bandwidth
in excess of what is actually required for transmission [102].
The SS method has the potential to perform better in some areas
than LSB coding, parity coding and phase coding techniques in that
it offers a moderate transmission rate while also maintaining a high
Information Hiding in Image and Audio Files 2007-2010
39
level of robustness against attacks. However, it still has a
disadvantage that it can introduce noise into a sound file [102].
(e)Echo Data Hiding:
Text can be embedded in audio data by introducing an echo to the
original signal. the data is then hidden by varying three parameters
of the echo: initial amplitude, decay rate, and offset (delay time)
from the original signal [102]. If only one echo was produced from
the original signal, only one bit of information could be encoded.
Therefore, the original signal is broken down into blocks before the
encoding process begins. Once the encoding process is completed,
the blocks are concatenated back together to create the final signal.
For example, the original signal is divided up into blocks, and each
block is assigned a one or a zero based on the secret message. In
this case, the message is the binary equivalent of 'HEY' as shown in
Figure 2.3. First an echo signal is created from the entire original
signal using the binary zero offset value. Then a second echo signal
is created from the entire original signal using the binary one offset
value. Thus the "one" echo signal only contains ones, and the "zero"
echo signal only contains zeros. To combine the two echoes
together to get the final encoding, two mixer signals are used.
Figure 2.4 summarizes the implementation of the echo hiding
process.
Information Hiding in Image and Audio Files 2007-2010
40
Figure 2.3 An Example of Echo Hiding
Figure 2.4 Echo Hiding process