madbrainedmudlickΤεχνίτη Νοημοσύνη και Ρομποτική

20 Οκτ 2013 (πριν από 5 χρόνια και 2 μήνες)

122 εμφανίσεις



The wide spread communication of multimedia data has made the management of information more challenging,
especially in the field of digital imaging. Where large images are to be
transmitted and also stored.

Digital image watermarking is one of the latest techniques which is used for safeguarding the origins of the image
by protecting it against Piracy

This project the information is visible in the picture or video.Typically, the

information is text or a logo which
identifies the owner of the media. The image on the right has a visible watermark. When a television broadcaster
adds its logo to the corner of transmitted video.


Digital image watermarking is one of th
e latest techniques which is used for safeguarding the origins of the image
by protecting it against Piracy.

Watermark is a set of identifying information that is either visibly or invisibly imprinted on a host image in order to
identify the origin of the

host image.

Types of Watermark in terms of Fidelity

• Fragile Watermark

This watermark is mainly used for detecting modification of data. This watermark gets degraded even for a slight
modification of data in the image.

• Robust Watermark

This wa
termark has the ability to withstand various image attacks thus providing authentication.

• Semi
Fragile Watermark

It is an intermediate between fragile and robust watermarks. It is not robust against all possible image attacks.


In the chapter we discuss in detail about the RGB images, its model and the representations of such images in
decimal bit planes and binary bit planes as well. Then the representation of the character image using a bit map
is also discussed.

RGB co
lor model

The RGB color model is an additive model in which red, green and blue (often used in additive light models) are
combined in various ways to reproduce other colors. The name of the model and the abbreviation ‘RGB’ come
from the three primary
colors red, green and blue. These three colors should not be confused with the primary
pigments of red, yellow and blue known in the art world as ‘primary colors’.


Digital watermarking is a technique which follows an individ
ual to add hidden copyright notices or other
verification messages to digital audio, video, or image signals and documents. Such a message is a group of bits
describing information pertaining to the signal or to the author of the signal (name, place, etc.)
.The technique
takes its name from watermarking of paper or money as securtity measure. Digital watermarking can be a form of
stenography, in which data is hidden in the message without the end user’s Knowledge.

A simple example of a digital watermark wou
ld be a visible” seal” placed over an image to identify the copyright.
However the watermark might contain additional information including the identity of the purchaser of a particular
copy of the material. According to the human perception, the digital w
atermarks can be dividing into different
types as follows: visible and invisible. Visible watermark is a secondary translucent overlaid into the primary

Fundamentals of watermarking

This section describes the different approaches that watermarki
ng algorithms is based on. The main topics being
wether to embed watermaks in the spatial domain or a transform domain. Advantages and disadvanteges are
also discussed.

The ideal algorithm

Nothing is perfect, which also holds true for watermarking algo
rithms.But what if it was possible to construct the
perfect or ideal watermarking scheme? Listed here are a sum up of requirements to such an ideal algorithm,
serving as introduction to the topic,but also emphasizing the many contradictions one encounters
when dealing
with watermarking.


Since watermarking is primarily used for copyright protection and proving ownership, the embedded watermark
has to survive and be extractable after the marked image has been submitted to a variety of things, for e

● Scaling of the image

● Converting a color image to grayscale

● Blurring, sharpening and other image
effect algorithms

● Lossy compression, for example JPEG, used widely on the internet

There are some obvious reasons for wanting to embed the
watermark, without being able to see any difference
on the marked image contra the original. Not being able to see the watermark, may keep some people from trying
to remove it. If the image is used unrightfully, and your watermark can afterwards be extract
ed, you have a pretty
good case against the copyright violator. It is also desirable to preserve the quality of an image, even though a
watermark is embedded in it. Imagine for example that beautiful pictures promoting a tourist website are severely
ted by the watermarking. Then the algorithm would be practically unusable.

Tamper resistant

Tightly linked to robustness, since any effort made to remove or deteriorate the watermark should result in the
watermarked image being severely degraded in qua
lity. There are different approaches for achieving a good level
of robustness, which will be discussed later.

Cheap and easy implementation

For a watermarking algorithm to have success, it has to be relatively easy to implement, while not costing a
tune. An algorithm is of no use if it takes a day to mark a picture, and a day to extract the mark again. It has to
be usable in real life which of course is application dependant.

Robust and fragile watermarks

It seems that for most applications, it would be ideal to have a watermark that are able to survive transmission,
usage and attacks. Such a watermark is named robust. On the other hand, watermarks is also used to detect if
the image they are in, has been a
ltered. That is watermarks that cannot resist any alteration. Such watermarks is
called fragile. Finally watermarks have been proposed, trying to combine robustness and fragility. That is a
watermark that can survive some alterations, but would break if th
e image was cropped for example, or parts of
another image was inserted into it.

The idea is that the number of bits you wish to embed, the watermark size, is stored in the pixels of the image to
be marked. Let us for example want to embed 40 bits of wate
rmark information in a host image. Then we supply
a seed to a PRNG (Pseudo Random Number Generator) and uses it to provide us with 40sets of (x,y) with
respect to the illustration, in other words we select 40 pixels.Each of these pixels are now altered, ex
changing a
chossen bit of color information with our watermark bit. The choice of bit involves a trade

If we choose one of the lesser significant bits, our alteration can be held invisible in the marked picture, but image
compression and other things

might be attacking these less significant bits, which could mean that the watermark
is not very robust.

If a more significant bit is chossen, we would achieve much better robustness, but our alteration would be visible.
It would look weird if the sunset
picture had 40 black spots in it. When you would attempt to extract the
watermark at a given time, the PRNG would then be started with the same seed, and again produce the 40
pixels, from where we canextract our bits. This method is easy to use, but has ma
ny drawbacks. Even small
alterations to our image would destroy at least some of our watermark information. Therefore the method could
be used for embedding fragile watermarks.corresponds to the siblings of chosen coordinate, the last four to the
of the chosen coordinates parent in the tree. The four outputs are the coefficients corresponding to the
chosen coordinates four children.

The trained neural network can now be used to embed the watermark. The eight input vectors for each selected
ate is given to the neural network, resulting in four outputvectors.The watermark information is then
embedded by replacing the original coefficients with the output from the neural network, adjusted by a
constant..To obtain the watermarked image, an inver
se DWT is performed.

Extraction phase

1. Transform the watermarked image using DWT.

2. Build the tree representation.

3. Use the seed k to start a sequence with the PRNG. This gives a set of coordinates.

4. The corresponding vectors are fed to the n
eural network, which results in four output

vectors. The difference between the expected output and the actual output of the neural

network provides information of the hidden watermark

Proposed Watermarking Technique

In this section, we give a descri
ption about the technique we propose, including the embedding and extraction
processes. Our proposed watermarking technique uses biorthogonal wavelets (biorthogonal splines in specific)
for watermarking. We decompose the image using DWT. The watermark we u
se is a binary image. We embed
the watermark, with the use of a key ( a small binary image), in the detail wavelet coefficients in order to make
the technique robust against several attacks, and to preserve imperceptibility.Our technique for embedding and
extraction can be summarized as follows:

The Embedding Process

The general embedding process is summarized in Fig. 1. Some of the conventional wavelet watermarking
techniques embeds the watermark in the components of the first level DWT. Other techniqu
es, that perform a
second level decomposition, obtain the approximation sub
band from the first level decomposition, and then
decompose it further as shown in Fig. 2. In our proposed embedding process, we use a localized decomposition,
meaning that the sec
ond level decomposition is performed on the detail sub
band resulting from the first level
decomposition as shown in Fig.

The proposed watermark embedding method. The host image is taken as the input to decompose it and embed
the watermark in its coeffic
ients to produce the watermarked image.


Invisible watermarks do not change the signal to a perceptually great extent, i.e., there are only minor variations
in the output signal. An example of an invisible watermark is when some bits are added to an image modifying
only its latest significant bits. Invisible watermarks that are unknown to the end user are stegnatography. While
the addition of the hidden messag
es to the signal does not restrict that signals use, it provides a mechanism to
track the signal to the original owner.

Another application is to protect digital media by fingerprinting each copy with the purchaser’s information. If the
purchaser makes il
legitimate copies, these will contain his name. Fingerprints are an extension to watermarking
principle and can be both visible and invisible.

There are various spatial and frequency domain techniques used for adding watermarks to and removing them
from s
ignals. Purely spatial techniques are not robust to some attacks to the signal like cropping and zooming,
where as most frequency domain techniques and mixed domain techniques are quite robust to such attacks.

Functional requirements of watermarking:


watermark conveys as much information as possible. This implies that the watermark data should be high. a
watermark should be a secret and be accessible to authorized parties only. This requirement is known as security
of the watermark. This is achieved b
y use of cryptographic keys.

A watermark is an integral part of data. This must persist even after single processing and data manipulation.
This also includes manipulation that attempts to remove the watermark. This requirement is known as robustness

A watermark though being irremovable should also be impercepitable. It should not modify or alter the quality of
content. Normally, the degradation in quality is well below watermark recovery process may or may not be
allowed to use the original
contents of the digital watermark one percent.

Generic model of watermarking scheme:

The generic model shows how a digital watermarking scheme functions. The first part shows how to create
watermarked data. The second part shows how to recover the wate
rmark from a test data.

A content owner approaches a neutral registration authority. Depending on the nature of multimedia content, the
authority allots a unique registration number. It also archives content and unique registration number for future

A content owner generates suitable water mark, which can be embedded with in data. Such a watermark should
be unobtrusive and secure. To ensure that watermark is imperceptible, the watermark signal amplitude should be
relatively small compared to t
he average amplitude of content.

To ensure security of embedded digital watermark, one or several secret and crypto logically secured keys have
to be used.

To ensure robustness against data manipulation and processing, it is helpful to have very small di
watermarks and ensure that it is redundantly disturbed in the host data. Thus, while extracting a digital
watermark, a small sample of watermark data is enough.

The digital watermark, public/private key and host data is processed using watermarking
algorithm to generate
the watermarked data.To extract the watermark, the authorized agency requires watermark and/or original host
data, secure or public key and test data.

All these inputs are processed by watermark recovery program to extract watermark
or confidence measure. The
confidence measure indicates the degree of closeness of original watermark and recovered watermark.

Requirements of digital watermarking:

Digital watermarking has to meet the following requirements:

1. Perceptual
transparency: The algorithm must embed data without affecting the perceptual quality of underlying
host signal.

2. Security: A secure data embedding procedure can not be broken unless the unauthorized user access to a
secret key that controls the inserti
on of data in host signal.

3. Robustness: Watermarking must survive attacks by lossy data compression and image manipulation like cut
and paste, filtering etc.

4. Unambiguous: Retrieval of watermark should unambiguously identify the owner

5. Universal:
Same watermark algorithm should be applicable to all multimedia under consideration.

6. Imperceptibility: The watermark should not be visible by human visual system (HVS) and should not degrade
the image quality.

7. Reliability: To ensure that the projec
t application returns the watermark each time.


With the development of the JPEG2000 standard digital image watermarking schemes that are based on
Discrete Wavelet Transform (DWT) are becoming more and more interesting. The ma
jority of watermarking
schemes consider Discrete Cosine Transform (DCT) as the method of choice. An overview can be found in .
Though, there is evidence that DWT could enhance robustness of the watermark against intentional and
unintentional attacks, work
based on DWT for watermarking is comparatively rare but exists.. The reason mainly
being that the former JPEG standard relied on DCT. With the advent of JPEG 2000 schemes based on DWT are
gaining interest. Though, watermark robustness varies with the under
lying transform algorithms, provisions must
be taken to harden a watermark against attacks. Ways are, e.g. multiple embedding and the application of error
correction codes are in use to regain the embedded watermark. Robustness of watermarking schemes bein
g far
from satisfactory this paper investigates application of DWT in conjunction with ECC.

The transform of a signal is just another form of representing the signal. It does not change the information
content present in the signal. The Wavelet Transform
provides a time
frequency representation of the signal. It
was developed to overcome the short coming of the Short Time Fourier Transform (STFT), which can also be
used to analyze non
stationary signals. While STFT gives a constant resolution at all freque
ncies, the Wavelet
Transform uses multi
resolution technique by which different frequencies are analyzed with different resolutions.

A wave is an oscillating function of time or space and is periodic. In contrast, wavelets are localized waves. They
their energy concentrated in time or space and are suited to analysis of transient signals. While Fourier
Transform and STFT use waves to analyze signals, the Wavelet Transform uses wavelets of finite energy.


Watermarking has been used as a

tool for embedding image for copyright protection. It is also used for
transmission of the data securely and also provided assistance to management of information. In our project we
have considered an RGB image that is a bit map of characters. We have suc
cessfully embedded the image
information using the MATLAB.


The work presented in this project is not user friendly i.e. one can provide a graphical user interface for the work
that has been presented. Also the effective retrieving o
f the image when passed through a lossy channel can be
considered for further work. The compression of the watermarked image can also be considered in this process