Hiding Information of user into an image by creating a regions of the host or watermarking image.

1. Region Based Undetectable Multiple Image
Watermarking
SHALU SINGH
Department of Computer Science and Engineering
Central University of Rajasthan
IEEE International Conference on Computational Techniques in Information and
Communication Technologies, Delhi, INDIA
11-13 March 2016

2. Contents
Introduction
Need of Watermarking
Problem Statement
Proposed Work
Conclusion
References

3. Introduction
Watermark:
Watermark is a designed impression(a pattern, logo or an
image) on the piece of paper or in digital media to
authenticate a document and discourage counterfeiting[7].
Digital Watermarking:
It is a process whereby, arbitrary information; a patterned bits
are encoded into digital media, in such a way as to be
imperceptible to observers[6][7][9].
Digital Image Watermarking:
It is perform on an image.it is process to embedded invisible
signature inside an Image, to show authenticity and proof of
ownership[7].

4. Continue
Multiple Watermarking:
Different type of techniques uses different watermarks in
digital media object to accomplish different goals.[2][4]
Watermark to convey Ownership Information,to Convey
Caption, to verify that Object Content.
Multiple Watermarking can be classified into four type
according to their insertion method into Host Image.[10][11]
Single Composite Watermarking
Segmented Watermarking
Successive Watermarking
Interactive Embedding

5. Need of Watermarking[2][7][9]
Proof of ownership
Ownership protection
Redistribution of digital contents
Illegal Copying of digital contents
Tamper Detection

6. Application of Image Watermarking[7][9]
Copyright Protection and Authentication
Fingerprinting and Digital Signatures
Copy Control and Device Control
Broadcast Monitoring
Data authentication
Further Application
In the field of Medical
Meta data of Information
Covert Channel Communication

7. Characteristics of Watermarking[6][9]
Robustness: Robustness is the resistance of watermark against
attacks and distortion
Imperceptibility: Imperceptibility of watermarked image shows
that inserted watermark in it, is not detectable or visible
through naked eyes.
Capacity: Capacity is a length of watermark upto certain limit
into the host image.
Security: Algorthim of embedding and extracting watermark
must be non guessable or decodable. Figure

8. Problem Statement
Detectability of watermark in a domains are major concern.

9. Proposed Method
We are proposing a multiple watermarking technique to insert
more than one watermark into the host image with the
features benefit of spatial and frequency domain. Who’s
overcomes each other’s drawback.

10. Proposed Work: Continue
The proposed method contains
Watermark Embedding Process
Watermark Extracting Process

11. Proposed Work: Continue
Figure : Embedding and Extractinng Process Technique

12. Propsed Work: Cotinue
Figure : Parts of Watermark Embedding and Extraction

13. Propsed Work: Cotinue
Watermark Embedding Algorithm: Insertion of watermarks are
divided into three parts called Part-A, Part-B and Part-C.
Step 1:A host image of size M x M divided into two regions
called Region-A and Region-B. Division of regions must be in
such a way that the size of Region-B is 2(n+1) x 2(n+1).Where,
2<=n<=M. Figure
Step 2: Watermark Insertion in Region-A: Watermark-Ws is
inserting in region-A using a LSB technique, at 6th bit plane.
Step 3: Watermark Insertion in Region-B:
Step 3.1: Part A: Apply one level of DWT on the
Region-B. Figure
Step 3.2: Part B:
i. Select LL sub-band whose size is 2n
x 2n
. Apply DFT on it.
Here we got a modified coefficients of LL whose size is same
as LL sub-band.Coefficients of LL sub-bands divided into four
parts of each size is 2n
2 − 1 x 2n
2 − 1 Figure

14. Proposed Work: Continue
ii. Watermark Insertion for P-1:
a. Select a high and a low DFT coefficients of P-1.
b. Select a block size ‘b’ and a pseudo random number
(PRN-1) of length ‘l’, where b*l = size of P-1.
c. Creation of Watermark W-1: Each frequency band is
replaced by one component of PRN-1 in every block. A
replacement location is selected randomly.
d. Select two scaling factor, ll and hh for low frequencies and
high frequencies respectively. Where ll >= hh.
e. Add W-1 in P-1 by additive scaling method.
iii. Copy P-1 into P-4 because of DFT symmetrical feature.
iv. Watermark Insertion for P-2:
a. Select same frequencies bands for P-2, as selected in P-1.
b. Select a block size ‘b’ and a different pseudo random
number (PRN-2) of length ‘l’, where b*l = size of P-1.
c. Watermark Insertion for P-2.
d. Creation of Watermark W-2: Watermark W-2 is created
same as step 3.2.ii.c with PRN-2.
e. Use same scaling factor to insert W-2 in P-2 as in step
3.2.ii.e.

15. Proposed Work: Continue
v.Copy P-2 into P-4 because of DFT symmetrical feature.
Step 3.2: Part C:
i. Combine all Watermarked parts P-1, P-2, P-3, P-4 and 2n
2
th
row and col. And apply reverse DFT on Combined image.
ii.Apply reverse DWT on watermarked LL and its respective
sub-bands. This gives us a watermarked Region-B.
Step 4: Combine Watermarked Region-A and Watermarked
Region-B to get a Watermarked Image.

16. Proposed Work: Continue
Watermark Embedding Algorithm:
To extract a Watermarks a reverse process of embedding is
applied.
Step 1:Divide Watermarked Image into Regions A and B.
Step 2:Extract Watermark Ws from Region-A, by selecting
LSB’s values from 6th bit plane.
Step 3:Apply part-A and part-B on Region-B, and extract
Watermark W1 and W2 from it.
For the detection of watermark from the watermarked image
reference –[5]– is followed.

17. Conclusion
We propose a multiple watermarking technique with taking
benefits of spatial and frequency domain.
Our motto is to make our watermarks secure and undetectable
with the help of inserting watermarks into different regions.
This method can be widely used in the field of medical to
insert patients sensitive diseases information. And in covert
channel for communication purpose where the insertion of
watermark is undetectable.

18. References
[1].Radu Ovidiu Preda,“Self-recovery of unauthentic images using
a new digital watermarking approach in the wavelet domain” IEEE
10th International Conference on Communications (COMM), 2014.
[2].Ouazzane Hana, mahersia hela, Hamrouni kamel,“A robust
multiple watermarking scheme based on the DWT”, IEEE 10th
International Multi-Conference on Systems, Signals & Devices
(SSD), 2013.
[3].Yang Qianli,CaiYanhong,“A Digital Image Watermarking
Algorithm Based on Discrete Wavelet Transformand Discrete
Cosine Transform”, in the Proceedings of IEEE International
Symposium on Information Technology in Medicine and Education,
2012.
[4].Sudhir Keshari, Shri Gopal Modani,“Dual watermarking based
on Multiple parameter fractional Fourier transform and LSB
technique”, in the Proceedings of IEEE International Conference

19. References
[5].Ganic, Emir, Scott D. Dexter, and Ahmet M.Eskicioglu,
“Embedding Multiple Watermarks in the DFT domain using
low-and high Frequency Bands”, in the Proceeding of International
Society for Optics and Photonics, Electronic Imaging, 200
Technologies for the Asia-Pacific Region, 2003.
[6].Mohammed,Ghassan N.,Azman Yasin and Akram M.Zeki,
“Digital Image Watermarking,Analysis of Current Methods”, in the
Proceedings of IEEE International Conference on Advanced
Computer Science Applications and Technologies, 2012.
[7].Frank Y.Shih,“Digital Watermarking and Steganography:
Fundamentals and Techniques”,CRC Press 2012.
[8].B.Sridhar and C. Arun,“On secure multiple image watermarking
techniques using DWT”,in the Proceedings of IEEE 3rd
International Conference on Computing Communication &
Networking Technologies, 2012

20. References
[9].Vidyasagar M.Potdar,Song Han,Elizabeth Chang,“A Survey of
Digital Image Watermarking Techniques”,in the Proceedings of
IEEE 3rd International Conference on Industrial
Informatics(INDIN), 2005
[10].Sheppard,Nicholas Paul,Reihaneh Safavi-Naini and Philip
Ogunbona,“On multiple watermarking”,in the Proceedings of
ACM, workshop on Multimedia and security:new challenges,2001
[11].Mintzer,Fred and Gordon W.Braudaway,“If one watermark is
good,are more better?”,in the Proceedings of IEEE International
Conference on Acoustics,Speech and Signal Processing, 1999
[12].P. Tao and A.M. Eskicioglu, “ A Method for Image Recovery
in the DFT Domain.”

21. .....................................ThankYou...............................

22. Back to page

23. Back to page

24. Back to page

25. Back to page