POSTER PAPER

International Journal of Recent Trends in Engineering Vol. 1, No. 2, May 2009

Digital Watermarking using Image Fusion

Method
Prof. S. C. Tamane, Dr. R. R. Manza and Dr. R. R. Deshmukh,
MGMs, Jawaharlal Nehru Engineering College, IT Department, Aurangabad, India,
[email protected]
Dr. Babasaheb Ambedkar Marathwada University, Computer Department, Aurangabad, India
[email protected], [email protected]
Abstract-- With the growing popularity of digital Medias
through the WWW, intellectual property needs copyright
protection, prevention of illegal copying and verification of
content integrity. The new data hiding techniques need to be
developed that satisfy the requirements of imperceptibility,
robustness, capacity, or data hiding rate and security of the
hidden data in order to keep the distribution of digital
multimedia work both profitable for the document owner
and reliable for the customer.
Previous research [01] indicates that significant portions of
the host image, e.g. the low frequency components, have to
be modified in order to embed the information in a reliable
and robust way. This led to the development of
watermarking schemes embedding in the frequency domain.
The wavelet transform has the number of advantages over
other transforms such as the DCT that can be exploited for
both, image compression and watermarking applications.
Therefore it is imperative to consider the wavelet transform
domain for watermarking applications.

Key words: Digital Watermarking,

Transform, Robustness, Image Fusion.

Wavelet

I. INTRODUCTION
A digital watermark can be described as a visible or
preferably invisible identification code that is
permanently embedded in the data. It means that it
remains present within the data after any decryption
process. A general definition can be given as: "Hiding of
a secret message or information within an ordinary
message and the extraction of it at its destination".
Complementary to encryption, it allows some protection
of the data after decryption. The goal is to embed some
information in the image without affecting its visual
content. In the copyright protection context,
watermarking is used to add a key in the multimedia data
that authenticates the legal copyright holder and that
cannot be manipulated or removed without impairing the
data in a way that removes any commercial value.
Figure1 shows a general watermarking scheme in order to
give an idea of the different operations involved in the
process.
The first distinction that one needs to do in the study
of watermarking for digital images is the notion of visible
watermarks versus invisible ones. The first ones are used
to mark, obviously in a clearly detectable way, a digital
image in order to give a general idea of what it looks like

Original

i) Watermarker

ii) Encryption

+
watermark

iii) Transmission Storage

and Possible Corruption

vi) Comparison
Watermark
Extraction
v) Decoder

iv) Decryption

Figure1: General Watermarking Scheme

while preventing any commercial use of that particular

image. The purpose here is to forbid any unauthorized
use of an image by adding an obvious identification key,
which removes the images commercial value. On the
other hand, invisible watermarks are used for content
and/or author identification in order to be able to
determine the origin of an image. They can also be used
in unauthorized images copies detection either to prove
ownership or to identify a customer. The invisible
scheme does not intend to forbid any access to an image
but its purpose is to be able to tell if a specified image
has been used without the owners formal consent or if
the image has been altered in any way. It is possible to
differentiate two ways of embedding Multi-resolution
Watermarking for Digital Images:
The first watermarking scheme that was
introduced works directly in the spatial domain. By some
image analysis operations (e.g. Edge detection), it is
possible to get perceptual information about the image,
which is then used to embed a watermarking key, directly
in the intensity values of predetermined regions of the
image
Another way to produce high quality watermarked
image is by first transforming the original image into the
frequency domain by the use of Fourier, Discrete Cosine
or Wavelet transforms for example. With this technique,
the marks are not added to the intensities of the image but
to the values of its transform coefficients. Then inverse
transforming the marked coefficients forms the
watermarked image. The use of frequency based
transforms allows the direct understanding of the content
of the image; therefore, characteristics of the human

113
2009 ACADEMY PUBLISHER

Watermarked

POSTER PAPER
International Journal of Recent Trends in Engineering Vol. 1, No. 2, May 2009
visual system (HVS) can be taken into account more
easily when it is time to decide the intensity and position
of the watermarks to be applied to a given image.
II. APPLICATION AREAS
Digital watermarking is considered as an
imperceptible, robust and secure communication of data
related to the host signal, which includes embedding into
and extraction from the host signal. For copy protection
applications, the watermark must be recoverable even
when the watermarked signal undergoes a considerable
level of distortion, while for tamper assessment
applications, the watermark must effectively characterize
the modification that took place. Some applications of
invisible watermarks are listed here: Fingerprinting,
Indexing, Copyright Protection & Owner Identification,
Broadcast Monitoring, Copy Protection, Data
Authentication, Data Hiding (Covert Communications),
Medical Safeties. From the applications mentioned
above, one can divide watermarks into two distinct types:
Robust, for the first five applications and Fragile for the
last three.

IV. PROPOSED WATERMARKING TECHNIQUE

Image watermarking imperceptibly embeds data
into a host image [5]. The general process of image
watermarking is depicted in figure2. The original image
(host image) is modified using the signature data to create
the watermarked image. In this process some error or
distortion is introduced. To ensure transparency of the
embedded data, the amount of image distortion due to the
watermark embedding process has to be small. The
watermarked image is then distributed and may circulate
from legitimate to illegitimate customers. Thereby, it is
subjected to various kinds of image distortion. The
successive stages of the watermarking processes defined
in the figure2 are: The embedding stage (figure 3), the
distribution stage (figure 4), the extraction stage (figure
5), and the decision stage.
Original
Image
(Host)

Key
Decision,
(Comparison,
Correlation)

In many cases the most distinguished information

is hidden in the frequency content of the signal. The
frequency spectrum of a signal is basically the frequency
components (spectral components) of that signal i.e. it
shows what frequencies exist in the signal. With the help
of Fourier Transform we can measure frequency, or find
the frequency content of a signal. FT is a reversible
transform, i.e. it allows going back and forward between
the raw and processed (transformed) signals but, only
either of them is available at any given space (time).
When the space localization of the spectral components is
needed, a transform giving the good space (time)frequency representation of the signal is needed and for
this one has to go for many of the following transform:
Short Time Frequency Transform (STFT) and Wavelet
transform (CWT and DWT).
The Discrete wavelet transform has three
properties that make it difficult to use directly in the
continuous form. The 1st is the redundancy of the CWT.
In CWT, the wavelet transform is calculated by
continuously shifting a continuously scalable function
over a signal and calculating the correlation between the
two. These scaled functions will be nowhere an
orthogonal basis and obtained wavelet coefficients will
therefore be highly redundant. For most practical cases
this redundancy has to be removed.

Extracted
Signature

Distorted,
Attacked
Image
Extraction
Original
Image

Figure2: The Data Hiding Model, a General Overview

A. The Embedding Stage

The host image is first transformed to a domain
that facilitates data embedding. This work exclusively
considers the wavelet and wavelet packet transform
domains. The signature data can be some binary data or a
small image (a logo). Typically the signature data has to
be encrypted to de-correlate the information and
subjected to some error-correcting coding scheme.
Original
Image
(Host)

Forward
Transform
(DWT)
Adaptive
Model
(HVS)

Signature

Prepare Data
(Encryption,
Redundancy)

Transform
Coefficients

Embed data
Additive
Quantization

Key

Inverse
Transform
(DWT)

Watermarked
Image

Figure 3: Model of the Watermark Embedding Stage

Next, the subset of the transform coefficients is

modified with the prepared signature data. By choosing a
suitable frequency transform domain and selecting only
certain coefficients, a lot of HVS modeling can be done
implicitly. Finally, the inverse transformation is applied
on the transform domain coefficients to produce the
watermarked image.
B. Distribution
The watermarked image is then distributed, or
published on a web server or sold to a customer. During
transmission and distribution of watermarked image, not

114
2009 ACADEMY PUBLISHER

Embedding

Distortion,
Noise,
Attacks

Signature

III. WAVELET TRANSFORM

Mathematical transformations are applied to
signals to obtain further information from that signal that
is not readily available in the raw signal. For time domain
signal, the time amplitude representation is not always
the best representation of the signal for most signal
processing related applications. The same is true for twodimensional image. The pixel or space domain
representation is not always the best representation

Watermarked
Image

POSTER PAPER
International Journal of Recent Trends in Engineering Vol. 1, No. 2, May 2009
only compression adds distortion to the host data, but
also transmission errors and common image processing
tasks, such as contrast enhancements, re-sampling and
gamma correction, contribute errors to the watermarked
image. All manipulation of the watermarked image data
has to be seen as an attack on the embedded information.
Modifications that occur during normal image processing
are called as coincidental attack. The attacks that attempt
to weaken, remove or alter the watermark itself are
termed hostile or intentional attacks.
Public
Detection
Device

Original
Image
Host

Lossy
Image
Compression

JPEG

Transmission
Noise, Image
Processing

Hostile
Attacks,
(Collision etc.)

Known
Image
Enhancement, Algorithm
Cropping,
Rotation

JPEG2000
Progressive
Transmission

Distorted
Image

Similar
Watermarked
Image

In the decision stage, the watermarking system

analyzes the extracted data, depending on the type of the
application and the nature of the signature data, the
decision stage can produce a number of different outputs.
For image copy protection applications, the output
of the watermarking system can range from simple to
more complicated answers. In the simplest case, the result
is just a yes/no decision indicating if the copyright
holders mark has been found in the received image data.
More complex systems return the embedded logo image
or the textual copyright information that was placed into
the host image data. A widely used similarity measure
between the original watermark and the extracted
watermark sequence is the normalized correlation for
pseudo random sequences .
The extracted watermark yes/no answer can be
derived from the similarity measure with an appropriate
threshold , i.e. if >= then is watermark is detected
otherwise watermark could not be found in the image.
Image labeling and data hiding applications will
typically try return the message originally embedded.
Since message corruption can not be tolerated, the use of
error correcting codes is mandatory for this type of
application.

Key

Figure4: Model of the Distribution of the Watermarked

Image

C. Extraction Stage
Eventually, after the watermarked image has
undergone severe distortion, one would like to extract the
embedded signature from the host data this can be done
by the party that embedded the watermark, the customer
that received the image, a designated party- such as a web
crawler that scan the internet for illegal copies of the
protected work or a legal prosecution official- or by a
third party. In the first case, the secret key used to embed
the watermark as well as the original image might be
available. We call detection systems that have access to
the secret (private) key and original image non-oblivious,
non-blind or private watermarking systems.
The other extreme is the case where neither the
private key nor the original image is available during the
extraction process. These watermarking systems are
called public key watermarking systems.

Original
Image
(Host)

Forward
Transform
(DWT

Transform
Coefficients

Extract
Data

Adaptive
Model
(HVS)

Distorted
(Attacked)
Image

Forward
Transform
(DWT

Extracted
Signature
Data

Transform
Coefficients
(Distorted)
Key
Recover
Data

Figure5: Model of the watermark extraction stage

D. Decision Stage

V. WATERMARKING ALGORITHM
A. Image Fusion Algorithm
Watermarking
algorithms
which
embed
meaningful data in the form of a logo image instead of a
pseudo-random number sequence are called image-fusion
watermarking algorithms. The logo image is generally
smaller than the host image. Before being added to the
host signal, the logo image is encrypted (de-correlated)
and suitably transformed.
There are two important advantages of embedding
a logo image as watermark data. First, the extracted
image can be correlated with the originally embedded
image by a human observer, building on the superior
pattern-matching capabilities of the human brain. Second
the existence of a visual logo in the questionable image
might be much better proof of ownership than a high
statistical correlation value.
Watermark: The Watermark is a gray scale
image, with as least 25% of the host image size.
Decomposition: The algorithm proposes using 2-level
decomposition on both, the host and the logo image, with
the Haar wavelet filter. The wavelet domain
representation of the host image is denoted by f(m,n), the
DWT coefficients of the logo image by w(m,n).
Coefficient selection: Each approximation coefficient of
the host image whose value is greater than threshold
value (i.e. 250) is modified to embed the logo image.
Embedding: The host and logo image coefficients of
each subband are linearly scaled. Since the logo image is
smaller than the host image, the coefficients have to be
expanded. After adding the expanded logo image to a
scaled version of the host image, image representation is

115
2009 ACADEMY PUBLISHER

POSTER PAPER
International Journal of Recent Trends in Engineering Vol. 1, No. 2, May 2009
scaled back using the original minimal and maximal
coefficients values per subband. Finally the fused
(combined) image is produced via the IDWT.
Extraction: The extracted watermark sequence is
compared to the originally embedded watermark using
the normalized correlation of the sequences as a
similarity measure . The similarity measure varies in
the interval [-1, 1], a value above 0 and close to 1
indicates the extracted sequence matching the embedded
sequence and therefore we can conclude that the image
has been watermarked. Discussion: The proposed
method allows hiding surprisingly high amounts of image
data in a host image. The current implementation is
limited to logo images that are a quarter of the size of the
host image. However this constraint can easily be
removed by exploiting the multiresolution property of the
wavelet transform and performing more decomposition
steps.
VI. RESULTS
A. Original Image and B. Watermark resp.

C. Transformed Watermark
Watermarked Image resp.

Image.

And

D.

Watermarked image

50
50

100
100

150

REFERENCES
[1]
Ingemar J. Cox,Joe Kilian, Tom Leighton, Talal
G. Shamoon on Secure Spread Spectrum Watermarking
from Multimedia, IEEE, ICIP 97, volume 6, Pages 16731687, Santa Barbara, California, USA, October 1997.
[2]
Maryline Charrier, Diego Santa Cruz, and
Mathias Larsson, JPEG2000, the Next Millennium
Compression Standard for Still Images. In Proceedings of
the IEEE, ICMCS 99volume 1, pages 131,132, Florence
Italy, June 1999
[3]
Mahalingam Ramkumar, Ali N. Akansu, and A.
Aydn Alatan. A robust data hiding scheme for images
using DFT. In Proceedings of the 6th IEEE International
Conference on Image Processing, ICIP 99,pages 211215, Kobe, Japan, October 1999 .
[4]
Improved
Wavelet-Based
Watermarking
Through
Pixel-Wise Masking Mauro Barni, Member,
IEEE,
Franco Bartolini, Member, IEEE, and
Alessandro Piva, IEEE TRANSACTIONS ON IMAGE
PROCESSING, VOL. 10, NO. 5, MAY 2001

Watermark

original image

the image is sufficient to correctly guess the embedded

code. As Watermarking becomes more widely used in
computing there are issues that need to be resolved. There
are a wide variety of different techniques with their own
advantages and disadvantages.

[5]
Digital Image Watermarking in the wavelet
transform domain, Diplomarbeit, Peter Meerwald,
Salzburg,am , 11 Janner 2001.
[6]
A. G. Bors, "Watermarking Mesh-Based
Representations of 3-D Objects Using Local
Moments," IEEE Trans. on Image Processing, vol 15, no.
3, pp. 687- 701, Mar. 2006. IEEE

150

200

200

250

250

300

300

350

350

400

400

450

450

500

500

50

100

150

200

250

300

350

400

450

500

50

100

150

200

250

300

350

400

450

500

VII. CONCLUSION & FUTURE SCOPE

In this paper, a novel algorithm for image
watermarking has been presented.
The algorithm embeds the watermark code by
modifying the DWT coefficients of the image, and
exploits a model derived from image compression
techniques for adapting the watermark strength to the
characteristics of the HVS. The performances of the
novel algorithm are very good, experimental results, in
fact, supported the suitability of DWT watermarking
schemes for robustly hiding watermarks into images. In
particular, the behavior of the watermark detector with
respect to image cropping was surprisingly good. As a
matter of fact, DWT schemes do not spread the
watermark all over the image, but, the watermarking
energy can be kept so high that even a small portion of

116
2009 ACADEMY PUBLISHER