Digital Watermarking

Course Instructor: Dr. Durgesh Singh

Assistant Professor
Dept. of Computer Science and Engineering,
PDPM IIITDM, Jabalpur
Evaluation Scheme
• Quiz I : 10 Marks
• Mid term : 20 Marks
• Quiz II : 10 Marks
• Project : 20 Marks
• End term: 40 Marks (Including 5 Marks Attendance )
• Attendance Marks
If Attendance >=95% (5 Marks)
Else if Attendance >=90% (3 Marks)
Else if Attendance > =85% (2 Marks)
Else if Attendance > =80% (1 Marks)
Else if Attendance > =75% (0 Marks)
Else “Not allowed in End Sem Exam and Quiz II”
Main objectives of this course
Authentication and restoration

(a) original image (b) forged image

Fig: Example of original image, and forged image

Copyright Protection
 Copy Move Forgery

Part of a digital image is copied and then pasted to another portion of the
same image.
Original Image Forge Image

Figure (a) original image Figure (b) Forge image

In image (b) Lady with bag in red circle is copied from the same image and pasted. Above is the example of
copy move forgery.
 Image Splicing
Image Splicing is a process of making a composite picture by cutting some
object from images.
and adding it to the other image.

Figure 2(a) Figure 2(b) Figure 2(c)

Figure 2: Image 2 (a),2 (b) is the original image while 2(c) is the spliced image.

In image 2(c) Tiger is copied from image (a) and pasted in the image 2(b). Above is the example of Image
splicing.
Definitions of Digital Watermarking
Watermarking is the practice of imperceptibly altering a work
(image, video, document etc.) to embed a message about that
work.
Original
work
Detected
message
Watermark Watermarked Watermark
embedder work detector
(looks like
Message original)
(regarding
work)
Definitions of watermarking: Other definitions
• Imperceptibility is not always considered essential (allows for visible
watermarking).
• Sometimes more narrowly defined as owner identification
(watermarks must indicate identity of owner).
“Watermarking is an art/ technique for embedding secret
message/information(watermark) into digital signal (Cover)which can
be later extracted or detected to verify the authenticity or identity of
its owner. ”
Basic Watermarking Model
Main Components of a Watermarking System
1. Cover/Host Work
• Cover or host works are those digital signals which must be protected
by the watermarking approach.
• A cover work may be either an image or audio or video or document.
• In image watermarking, Cover may be binary image, gray scale or
color image
• The complexity and performance of the image watermarking
approach depends on the type of image as well as the dimensions of
the cover image.
Main Components of a Watermarking System
2. Watermark/Message/Secret Information
• The watermark/message is the information that must be hidden in
the cover image.
• A watermark is used to protect the cover image and it may be
related/unrelated to the cover image in which it is going to be hidden.
• A watermark may be derived from the cover image itself. The
watermarking method in which watermark is derived from cover itself
known as self-embedding watermarking method/scheme.
• It may also be a random image or random pattern of bits.
Main Components of a Watermarking System
3. Embedding System: The embedding system is the one which
embeds the watermark into the cover image.
4. Watermarked Image: Watermarked images are nothing but the
cover images embedded with the watermark.
5. Transmission Channel: This is the medium by which the
watermarked image travels up to the receiver
6. Extraction or Retrieving System: This is the entity where the
watermark extraction algorithm is executed to extract the watermark
from the watermarked image.
Basic Image Processing
What is Digital Image ?
• An image may be defined as a two-dimensional function, f(x, y),
where x and y are spatial (plane) coordinates, and the
amplitude of f at any pair of coordinates (x,y) is called the
intensity or gray level of the image at that point.
• When x, y, and the intensity values of f are all finite, discrete
quantities, we call the image a digital image.
RGB vs Gray Scale Image
• RGB Image: It is three dimension(channel) image. It contains the Red
color , Green color and Blue color image in separate matrix.
• Gray Scale Image: It is one dimension (channel) image. It is
derived from RGB image.
• RGB Image to Gray Scale conversion formula
𝑌 = 0.299𝑅 + 0.587𝐺 + 0.114𝐵
RGB vs Gray Scale Image
RGB image representation
• An RGB image can be viewed as three different images(a red scale
image, a green scale image and a blue scale image) stacked on top of
each other.
• When fed into the red, green and blue inputs of a color monitor, it
produces a color image on the screen.
Colour planes of RGB image in MATLAB

• Consider an RGB image array ‘I’ then,

I(:, :, 1) represents the Red colour plane of the RGB image
I(:, :, 2) represents the Green colour plane of the RGB image
I(:, :, 3) represents the Blue colour plane of the RGB image
Gray Scale Image
• The gray level of each pixel in a digital image is stored as one or more
bytes in a computer.
• For an 8-bit image, 0 is encoded as 00000000 and 255 is encoded as
11111111. Any number between 0 to 255 is encoded as one byte.
• The bit in the far-left side is referred as the most significant bit (MSB):
❖because a change in that bit would significantly change the value
encoded by the byte.
• The bit in the far right is referred as the least significant bit (LSB):
❖because a change in this bit does not change the encoded gray value
much.
MSBs vs LSBs in Gray Scale Image
b7 b6 b5 b4 b3 b2 b1 b0
1 0 0 0 1 0 1 1
Why Multimedia Security?
• Multimedia communication plays an important role in multiple areas
in today’s society including politics, economics, industries, militaries,
entertainment, etc.
• Easier to transfer multimedia documents over the Internet
• Due to advancement of editing software and insecure Internet, illegal
operations such as duplication, modification, forgery have become
easy and difficult to prevent.
• It’s also important to determine where and how much a multimedia
file differs from its original.
• Therefore, security has become one of the most significant problems
for distributing new information technology.
• As the possible solutions, Cryptography, and digital watermarking
come to our help.
Cryptography
• Cryptography (i.e. encryption technique) is the science of secret
writing.

➢ A cipher is a secret method of writing, where by

plaintext is transformed into a ciphertext.
➢The process of transforming plaintext into ciphertext
is called encryption.
➢The reverse process of transforming ciphertext into
plaintext is called decryption.
➢Encryption and decryption are controlled by
cryptographic keys.
Cryptography …
Cryptographic Systems

• Symbols
✓Plaintext: M (for message) or P (for plaintext)
✓Ciphertext: C
✓Encryption function: E
✓Decryption function: D
• Formulations
✓E(M)=C, the encryption function operates on plaintext to produce
ciphertext
✓D(C)=M, the decryption function operates on ciphertext to produce
plaintext
Cryptosystem requirements

• Efficient enciphering/deciphering
• Systems must be easy to use
• The security of the system depends only on the keys, not the secrecy
of E or D
• The key (usually denoted by K ) may be any one of a large number of
values
• The range of possible values of the key is called the keyspace.
Why is Watermarking becoming the key
method for protecting Multimedia?
• The major drawback of cryptography is that once the data is
decrypted by an intruder/attacker, there is no way to protect the data
and track its illegal distribution.
➢In cryptography where the content is encrypted prior to delivery
and a decryption key is offered only to those who have purchased
legitimate copies of the content.
➢ Cryptography can protect content from manipulation only in
encrypted form but once decrypted, the content has no further
protection from illegal duplication.
Why is Watermarking becoming the key
method for protecting Multimedia?
• In other hands, digital watermarking is a technique that employs to
guard digital content from illegal copying and manipulation even after
it is in decrypted form.
• Therefore, digital watermarking has drawn much attention of
research community to resolve these problems due to its advantages.
Note:
1. Cryptography can protect content only in encrypted form
2. Cryptography can not help after decryption
History
• The Italians where the 1st to use watermarks in the manufacture of
paper in the 1270's.

• A watermark was used in banknote production by the Bank of

England in 1697.
Other Related methods: Cryptography

• Cryptography is a method of protecting information and

communications through the use of Encryption, so that only those for
whom the information is intended can Decrypt it in original
meaningful information using key.
.
Other Related methods: Steganography

• A term derived from the Greek words “steganos” and “graphia” (The
two words mean “covered” and “writing”, respectively).
• Writing hidden messages in such a way that no one, apart from the
sender and intended recipient, suspects the existence of the
message.
– The art of concealed communication.
– The very existence of a message is kept secret.
Watermarking v.s. Steganography

• Watermarking is the subset of Steganography.

• Watermark messages contain information related the cover work
• In steganographic systems, the very existence of the message is kept
secret.
❖In simple words we can say that in watermarking, the cover is a
primary object, and the secret message (watermark) is secondary
object which is actually used to protect the cover. Whereas in
steganography, the secret message is the primary object, and the
cover is secondary which is used only to cover/protect the secret
message.
Visual Cryptography
• Visual Cryptography (VC) is a type of secret sharing scheme proposed
by Naor and Shamir in 1994.
• Visual cryptography is a cryptographic technique which allows visual
information (pictures, text, etc.) to be encrypted in such a way that
the decrypted information appears as a visual image.
• Visual cryptography provides a powerful technique by which one
secret image can be divided into two or more shares.
• When a predefined set of these shares are superimposed exactly, the
original secret image can be discovered without any computation
otherwise nothing will be revealed.
 1. The original image or text is divided into two or more
shares, each of which appears as a random and
How Visual meaningless pattern.
Cryptography works? 2. The shares are then distributed to the intended
recipients.
3. When a predefined set of these shares are
superimposed exactly, the original secret image can be
discovered without any computation otherwise nothing
will be revealed.
k out of n (k,n) Visual Cryptography
• After making transparencies of n generated shares, they are
distributed among n participants, one to each.
• Any single share does not reveal any visual content of the secret
image.
• But when k or more assigned transparencies of the participants are
properly superimposed together, visual content of the secret image
will be exposed.
• It means that at the receiver end, the secret decoding process
requires no computation at all, i.e., the beauty of the visual
cryptography.
Example:
(k,n)visual
cryptography
Classifications of Watermarking

▪ Based on the various features, watermarking approaches can be

classified as follows:
➢Attached Media
➢Visibility or Perceptibility
➢Resist Attack
➢Watermark Embedding Method
➢Requirements for Watermark Extraction or Detection
Basis on Attached Media

Figure: Types of watermark basis on attached media

Basis on Attached Media
• Image watermarking: Image watermarking represents to add
watermark in still image is used to embed watermark in a still image.
• Video watermarking: Video watermarking refers to embedding digital
watermark in the video stream to control video applications.
• Audio watermarking: Audio watermarking refers to embedding
digital watermark in the audio stream to control audio applications.
• Text watermarking: Text watermarking refers to embedding
watermark in DOC, PDF and other text file to prevent changes of text
• Graphic watermarking: Graphic watermarking the watermark is
embedded into two-dimensional or three-dimensional computer-
generated graphics to prevent changes.
Basis on Visibility or Perceptibility

Figure: Types of watermarking basis on visibility

Basis on Visibility or Perceptibility

• Visible watermarking: In case of visible (or perceptible)

watermarking, the watermark is embedded in a host image in such a
way that the watermark is noticeable to a human observer.
• In other words, those watermarking approaches in which the
embedded watermark is visible on the watermarked image are called
visible watermarking approaches.
• There are many areas in which we need such types of visible
watermarks so that just by seeing the watermarked image one can
identify the watermark.
Visible Watermarking

Fig. :Example of visible watermark.

Basis on Visibility or Perceptibility

• Invisible watermarking: In the case of invisible (or imperceptible)

watermarking, the embedded data is not detectable, but can possibly
be extracted by some software or a computer program.
• Unlike the visible watermarking approaches, invisible watermarking
approaches are those in which the embedded watermark does not
appear on the watermarked image.
• We can only extract the watermark using an extraction mechanism.
• These types of watermarks are used where we do not want to
compromise the imperceptibility of the watermarked image.
Invisible Watermarking

Fig. :Example of Invisible watermark.

Basis on Watermark Embedding Method

Figure: Types of watermarking basis on

embedding method
Basis on Watermark Embedding Method
• Transform domain Watermarking: Transform domain techniques
perform the watermark by changing the coefficients in the frequency
domain of host image. Discrete Cosine Transform (DCT) and Discrete
Wavelet Transform (DWT) are commonly used frequency domain
techniques.
• Spatial domain watermarking: In the spatial domain techniques,
watermark is directly applied on pixel value of the host/ cover
image. Spatial domain algorithms are easier to implement.
Basis on Resist Attack
• Some watermarking cannot tolerate an attack and some can
tolerate very efficiently.
• Watermarking can also be classified based on their degree of
resistance to attacks in four categories as follows:

Figure: Types of watermarking basis on resist attacks

Robust Watermarking
• A robust watermark should be able to resist intentional or
unintentional manipulations.
• The watermark is embedded into the cover image in such a
way that if the content of the watermarked image is
tampered with any way, the associated robust watermark
will remain same.
• Some important applications for robust watermarking are
copyright protection and ownership verification.
Fragile watermarking:

• A fragile watermark is intended to be destroyed even after the

minor unintentional or intentional manipulation in the
watermarked image.
• Fragility is essential in order to check the integrity of the
given cover image.
• A fragile watermark is embedded into the cover image in
such a way that if content of the watermarked image is
tampered, the associated fragile watermark with that
content is also tampered or destroyed.
Fragile watermarking (Cont….)

• At the receiver end, the unavailability of the watermark

confirms that the associated principal content of the cover
image has been tampered.
Semi-fragile watermarking
• A Semi-fragile watermarking should be able to resist unintentional
manipulations caused by common image processing operations like
JPEG compression and is fragile against intentional, malicious
manipulations.
• Semi-fragile watermarks are those watermarks, that are robust
watermarks for unintentional attacks and fragile for intentional
attacks.
• A threshold is used to separate the amount of alteration of
content.
• If alteration of content is less than the threshold value then, it
will be treated as an unintentional attack, otherwise it is treated
as an intentional attack
Note: The main applications field of fragile and semi-fragile
watermarking are image and video content authentication.
Basis on Requirements for Watermark Extraction or
Detection
• This phase is executed on the receiver side.
• At the receiver end, we extract/detect the watermark.
• Based on the method of detection of the watermark, it is classified
into three type as follows:

Figure: Types of watermarking basis on requirements for detection

Basis on Requirements for Watermark Extraction or
Detection
• Non-blind Watermarking: The non-blind watermarking (i.e. private
watermarking) requires the original cover image and secret key(s) to
identify the watermark.
• Semi-blind watermarking: The Semi-blind watermarking demand the
presence of both the secret key(s) and the watermark bits sequence.
• Blind watermarking: On the other hand, the blind (or public)
watermarking requires only the secret key(s) for extraction.