Professional Documents
Culture Documents
Amita Dhankhar
Assistant Professor
UIET, MD University Rohtak, Haryana
Abstract
Steganography is an important area of research in recent years. Steganography usually deals with the ways of hiding the
existence of the communicated data in such a way that it remains confidential. Steganography is used some cover media like
image file, audio file, video file, text file to hide the data. In image steganography, secrecy is achieved by embedding data into
cover image and generating a stego-image. Bits of information are placed in pixel values of image so that intruder cannot find
out where data is hidden. There are different types of steganography techniques each have their strengths and weaknesses. In this
paper, we review the different technique of Image steganography - LSB, PVD (least significant bit, pixel value differencing), PIT
(pixel indicator technique), of Image steganography techniques - LSB, PVD (least significant bit, pixel value differencing), PIT
(pixel indicator technique),
Keywords: LSB, PVD, edge, PIT, Steganography, Stego-image, cover-image
_______________________________________________________________________________________________________
I. INTRODUCTION
Two techniques are available to those wishing to transmit secrets using unprotected communications media. One is
cryptography, where the secret is scrambled and can be reconstituted only by the holder of a key. When cryptography is used, the
fact that the secret was transmitted is observable by anyone. The second method is steganography. The word steganography in
Greek means Covered Writing. Steganography is the art of information hidden into cover media. Cover media can be a text, or
an image, an audio or video etc.
Main objective of steganography is to communicate securely in such a way that the true message is not visible to the observer.
Today steganography is mostly used on computer with digital data being the carriers and networks being the high speed delivery
channel. Most steganographic utilities hide information inside image, as it is relatively easy to implement. Images are mostly
used in the process or of steganoraphy because it is hard to break.
During the process of hiding the information three factors must be considered that are capacity it includes amount of
information that can be hidden in the cover medium. Security implies to detect hidden information and Robustness to the amount
of modification the stego medium can withstand before an adversary can destroy hidden information.
The importance of steganography lies in the fact that it hides the very existence of the secret it is protecting. Attackers are
threats because they attempt to damage or gain access to assets by taking advantage of these assets vulnerabilities.
Steganography makes the job of the attacker more difficult because the very existence of the asset is hidden. The importance of
steganography in maintaining confidentiality can be illustrated with a simple example. Imagine two coworkers, Alice and Bob,
are communicating with each other over the internet. Eve, an attacker, has access to this communication link, so she eavesdrops
on Alice and Bobs communications. If Alice is asking Bob if he is free for lunch, then Alice probably does not mind if Eve
reads this message. Thus, Alice can send her query to Bob along the communication link in plain text. However, if Alice is
sending Bob confidential information, such as specifications for their companys latest project, then she probably does not want
Eve to be able to read these messages. Therefore, Alice will likely encrypt her messages. A problem arises because the encrypted
text is likely garbled, nonsensical data. Thus, Eve, even though she cannot read the encrypted messages, will know that Alice has
a secret that she is sending to Bob. Eve can then take the encrypted message and attempt to crack it. This is a very real problem
because as computational power increases, encryption is becoming easier to break. However, if Alice uses steganography, and
hides her secret message in a generic image file, then she can transmit her secret message to Bob without evoking Eves
suspicion. For instance, Alice can hide her secret message in a picture of her garden. She can then send the image, with the secret
message hidden inside it, to Bob. Eve will think Alice is just sending Bob a harmless picture, so she will ignore that
communication between Alice and Bob. Thus, Alice and Bob defeat Eve.
243
Information can be hidden in many different ways. In order to hide information, straight message insertion may encode every
bit of information in the cover data or it may selectively embed messages in noisy areas that draw less attention. Messages may
also be scattered randomly throughout the cover data. There are a number of ways to hide information; the most common
methods are the least significant bit (LSB) insertion, masking and filtering and algorithms and transformations.
This paper describes different technique used in image steganography, performance, analysis & comparison on each technique.
244
2) PVD Method
The pixel value differencing (PVD) method proposed by Wu and Tsai [15] can successfully provide both high embedding
capacity and outstanding imperceptibility for the stego-image. The pixel value differencing (PVD) method segments the cover
image into non overlapping blocks containing two connecting pixels and modifies the pixel difference in each block (pair) for
data embedding. All possible difference values are classified into a number of ranges. A larger difference in the original pixel
values allows a greater modification.
Kim, Jung and Yoo [16] proposed A high capacity data hiding using PVD and LSB Replacement Method. This method
calculates the difference value between two consecutive pixels. The LSB substitution method is used when the difference value
is small (i.e smooth areas) and PVD is used when it is large (i.e edge areas) of cover image pixels. The pixels belong to the edge
areas could embed more data than the smooth area of image thus LSB substitution is used to embed more data on smooth area
without distortion to the human visual system. This method is useful for the grayscale images.
3) Parity Checker Method
In this method, Rajkumar et al [17] gives the concept of odd and even parity. According to this method, 0 can be inserted at a
pixel location if that pixel has odd parity i.e. the number of 1s in the binary value of the pixel should be odd. Similarly, 1 can be
inserted at a pixel location if that pixel has even parity i.e. the number of 1s in the binary value of pixel should be even. If the
corresponding parity does not exist at a pixel location either for 0 or 1, then we make corresponding parity at that pixel location
(odd parity for 0 and even parity for 1) by adding or subtracting 1 to the pixel location such that the change in the image quality
should not be visible to the human visual system (HVS). For Retrieval of message, again we used the parity checker. If odd
parity is present at the selected location then, 0 is message bit, else message bit is 1. Retrieval process was repeated for all
locations where message bits were hidden. In this way, we retrieved the message bits from all the locations where the message
bit were inserted.
4) Edge based Method
Yang, Weng, Wang (2008) [18] proposed the new adaptive LSB technique using Pixel Value differencing with spatial LSB
domain technique. It gives high embedding capacity and imperceptible Stego image. This method is used to distinguish between
edge areas and smooth areas. The edge areasare used for the higher embedding capacity for data hiding .This technique comes
under the spatialdomain image processing and this method applied on grayscale images. The quality of stego imagescalculated
from the peak signal to noise ratio (PSNR).
5) PIT Method
Adnan Abdul-Aziz Gutub[19] proposed a new improved technique Pixel indicator that takes the advantage of the 24 bits in each
pixel in the RGB images using the two least significant bits of one channel to indicate existence of data in the other two
channels. The technique uses least two significant bits of one of the channels Red, Green or Blue as an indicator of secret data
existence in the other two channels. The indicator channel is chosen in sequence from R, G and B, i.e. RGB, RBG, GBR, GRB,
BRG and BGR. However the indicator LSB bits are naturally available random, based on image profile and its properties. The
stego method does not depend on a separate key to take out the key management overhead. Instead, it is using the size of the
secret data as selection criteria for the first indicator channel to insert security randomness.
6) 6th, 7th Bit Method
Parvinder et al [20] obtained an algorithm to hide the message in 6th and 7th bit of pixel value. Here instead of LSB bits 6th and
7th bit of pixel values are used to hide the secret data within an image. This novel proposed method overcomes the all
disadvantages of LSB insertion method .However it has its own disadvantage that is the chance that the message bit will be
inserted at pseudorandom location at first instance is less as compared to LSB.
7) 6th, 7th & 8th Bit Method
Rajkumar et al [21] proposed a novel approach to hide the data 6 th, 7th and 8th bit of pixel values. In this method 6th, 7 th and 8th
bits of the image pixels are used to hide the message.
Since this method involves 8th bit for hiding the message, intruder can easily change 8th bit of all image pixels and this may
result in the loss of message. To avoid this, time factor has been introduced, i.e. at some time t1, sender sends the cover object
with message and at some other time t2 sender sends the cover object without message. Sender and recipient agree on this time
factor initially before starting any communication. The advantage of introducing time factor (slot) is that if least significant bits
of all pixels are changed by the intruder even then the message can be retrieved by comparing the two cover objects, i.e. one
containing the message and the other not containing the message.
245
B. Secret Communication
In many situations, transmitting a cryptographic message draws unwanted attention. The use of cryptographic technology may be
restricted or forbidden by law. However, the used steganography does not advertise covert communication and therefore, avoid
scrutiny of the sender, message and recipient.
C. Tamper Proofing
The objective of tamper-proofing is to answer the question, Has this image been modified? Tamper-proofing techniques are
related, but distinct from the other data-hiding technologies. What differentiates them is the degree to which information is
secured from the host signal. Most data-hiding techniques attempt to secure data in the face of all modifications. Tamperproofing techniques must be resilient to small modifications (e.g., cropping, tone scale or gamma correction for images or
balance or equalization for sounds) but not to large modifications (e.g., removing or inserting people from an image or taking
words out of context in an audio recording [24].
D. Digital Watermark
The objective of a digital watermark is to place an indelible mark on an image. Usually, this means encoding only a handful of
bits, sometimes as few as one. This signature could be used as a means of tracing the distribution of images for an on-line
news service and for photographers who are selling their work for digital publication. One could build a digital camera that
places a watermark on every photograph it takes. Theoretically, this would allow photographers to employ a web-searching
agent to locate sites where their photographs appear [21].
E. Smart Ids:
In smart IDs the information about the person is embedded into their image for confidential information. For an organization,
the authentication of the resources is accessed by the people. So identifying the theft related to prevention of crimes.
2552
PSNR 10log10
MSE
where:
I is the dynamic range of pixel values, or the maximum value that a pixel can take, for 8-bit images: I=255.
MSE is the mean square error.
B. RMSE
The Root mean square error measures the difference between these two images. Since the computing of these two metrics is very
easy and fast, they are widely-used and very popular [13].
RMSE
X
N M
N
i 1 j 1
ij
Yij
( X
i 1
X )2
Here,
N, is the number of rows in the cover image.
M, is the number of column in the cover image
Xij, intensity of pixel ij in cover image
Yij, intensity of pixel ij in stego image
Assuming pixel values in the range [0,255], the following observations are mentioned as:
RMSE of zero which means an identical image results in infinite or undefined PSNR
RMSE of 255 result in PSNR of zero
RMSE greater than 255 results in negative PSNR.
246
C. NCC
Normalized cross correlation (NCC) has been commonly used as a metric to evaluate the degree of similarity (or dissimilarity)
between two compared images. The main advantage of the normalized cross correlation over the cross correlation is that it is less
sensitive to linear changes in the amplitude of illumination in the two compared images.
Furthermore, the NCC is confined in the range between 1 and 1 to evaluate the performance the Normalized CrossCorrelation (NCC) which is given by the following equation
x
N
NCC
i 1
ij
j 1
N
M
yij
x
i 1
j 1
ij
Authors
Year
Features
1.
Neil F. Johnson
& Sushil
Jajodia
1998
3.
M. Chaumont et
al.,
2007
Simple to implement
100 % chances of
insertion.
less prone to attacks,
high compression
ratio
4.
Yang, Weng,
Wang
2008
More secure
2010
Take advantage of
24 bits
Computationally Complex
Yadav et.al
2010
Parvinder at.al
2010
Yadav et.al
2011
5.
6.
7.
8.
Advantages
Easy to implement
Disadvantages
Not immune to noise and compression
technique Not Secure
Not robust against rotation, cropping and
translation.
Chances of insertion
are 50%
V. CONCLUSION
In this paper we have reviewed some steganographic techniques, different parameters and different methods for analyzing of
original image and stego image. We have reviewed techniques like LSB, parity checker method, pixel value indicator, 6th, 7th
bit method, 6th, 7th & 8th bit method, etc. with their advantages and disadvantages. In future we will try to develop a new image
steganography technique which will cover the limitations of these techniques.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
247
248