Hybrid Method For Image Watermarking Using 2 Level LWT-Walsh Transform-SVD in YCbCr Color Space

International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 11 216 – 221
_______________________________________________________________________________________________
216
IJRITCC | November 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
Hybrid Method For Image Watermarking Using 2 Level LWT-Walsh Transform-
SVD in YCbCr Color Space
Rajeev Dhanda
Assistant Professor, ECE Department
PIET, Samalkha
Panipat, India
rajeevdhanda208@gmail.com
Dr. K. K Paliwal
Director ECE Department
PIET, Samalkha
Panipat, India
captkkpaliwal@gmail.com
Abstract: Due to tremendous development in technology in recent time and availability of abundant tool, it is very easy for an unauthorized
person to imitate crucial information which is present on internet. Therefore to shield valuable information present on internet there are various
advanced techniques for example watermarking technique, cryptography technique, steganography and many more. With pace of time analog
techniques replaced by digital techniques due to various advantages and in current scenario every country moving towards digitalization. Digital
watermarking is a technique through which digital information is embedded into an image and secret digital data can be extracted at receiver side
with authentication otherwise impossible to fetch. Spatial domain and frequency are the two techniques through which secret digital information
can be embedded. In this paper two level lifting wavelet transform (LWT), Walsh Hadamard transform and singular value decomposition (SVD)
technique has been proposed in YCbCr color space. First of all cover image and watermark image converted into YCbCr color space from RGB
color space after that one of channel is selected for embedded purpose. Now perform first level LWT on the Y channel of cover and watermark
image so that image split into four groups. Now apply second level LWT on any one of four bands. Further Walsh hadamard transform
technique applied with singular value decomposition (SVD) technique to get enhanced output. In base paper DWT-DFT-SVD used but in this
paper DWT-DFT replaced by LWT-WHT due to various advantages. One disadvantage of DWT is that the use of larger DWT basis functions
or wavelet filters produces blurring and also ringing noise near edges in images. This disadvantage of DWT is overcome in LWT. Other
advantages of LWT are that it significantly reduces the computation time and speed up the computation process. This method provides better
results in terms of enhanced PSNR values and is able to withstand a variety of image processing attacks and besides this processing time also
reduced.
Keywords- DWT, SVD, LWT, Hadmard Code, Watermarking, PSNR
__________________________________________________*****_________________________________________________
I. INTRODUCTION
Due to huge research in recent time various latest and
advanced technologies came into existence. One should keep
in mind if technologies not used in right direction then
definitely have adverse effect on entire world. In these days
there is huge demand of internet and it is growing
exponentially day by day. Now a day’s popularity of internet
brings its own set of disadvantages in the form of piracy, theft,
ownership issues [1]. To ensure security various techniques
came into existence which can help to protect our confidential
data available on internet for example these techniques include
steganography, watermarking and encryption and many more.
In recent time watermarking technique is tremendously used.
There are different types of watermarking like transform
domain and spatial domain watermarking. Digital image is
made up of pixels and in spatial domain watermarking
technique directly used on image pixels [2]. On the other hand
in transform domain watermarking technique, coefficients are
changed and it can be achieved using different methods for
example discrete cosine transform [6] which is robust to JPEG
compression but susceptible to geometric distortions. DCT
basically used for image compression. Discrete Fourier
transform method used in [7], [8] is rotation invariant and
translation resistant and due to these it gives better resistance
in case of geometric attacks. LWT disintegrate an image into
different four wavelets or sub band LL, LH, HL and HH.
These sub bands actually represent authenticity of original
image. LL sub band provide the entire information of image
into low frequency domain and HH, HL and LH gives
diagonal, vertical and horizontal features of image. Besides
this here L stand for low pass filter and H stand for high pass
filter and as per requirement of our application suitable sub–
band is selected and after that watermark is embedded [9],
[10]. Singular value decomposition is a matrix method to get
smaller set of values which consists optimized signal content
[13], [5]. Singular values of the watermark which are achieved
after SVD are embedded with singular values which obtained
from SVD transformed host to give least truncation error. This
also enhances robustness of the watermarked image as it is
withstand to various attacks. There are various techniques used
for watermarking for example discrete wavelet transform
singular valued decomposition based watermarking technique
used in [13] and this concept further extended to YUV color
space [7]. DWT–DCT–SVD based watermarking technique is
used in [6]. Besides this a combination of DWT–DFT–SVD is
used in Ref. [3] which merges benefits of DFT, DWT and
SVD methods.

Volume: 5 Issue: 11 216 – 221
_______________________________________________________________________________________________
217
_______________________________________________________________________________________
Figure 1 Decomposition of the image into four components [2]
II. FUNDAMENTAL CONCEPT
A. Problem Defination
First of all very first task of research is to find out problem and
to fetch need to study various research papers deeply and
analyzing the result. After gone through so many research
papers, main problem came into existence is the low value of
PSNR (peak signal to noise ratio). If PSNR value will be high
then it represents high quality of the watermarked image and
vice versa. On the other hand if there is high value of
correlation which signify the immunity of the watermarked
image to different types of attacks. Therefore to achieve
objectives, a propose technique which will be illustrated
below:
1. Usage of YCbCr color space instead of RGB color
space due to its high decorrelation value as compared
to RGB color space which ensures better correlation.
2. Combination of two level LWT, WHT, SVD is used
to watermark an image which actually combines the
advantages of all the transforms.
B. YCbCr Color Space
YCbCr color space actually represents brightness and signal
color difference on the other hand RGB depicts color as red,
green and blue components. In YCbCr, Y represents
luminance, Cb is actually B-Y means difference of blue
component and luminance component and in the same way Cr
is the difference of red component and luminance component
(R − Y). YCbCr given by following equation
Y = (0.257 × R) + (0.504 × G) + (0.098 × B) + 16
Cb = (0.439 × R) − (0.368 × G) − (0.071 × B) + 128
Cr = (0.148 × R) − (0.291 × G) + (0.439 × B) + 128
C. Singular Value Decomposition
Singular value decomposition is factorization three matrices
and these matrices can be complex or real. Consider an m×n
matrix M which could be complex or real and its Singular
value decomposition represented as follows,
M = USVT
Here U is an unitary matrix of order m×m which could be
complex or real, S is a rectangular diagonal matrix of order an
m×n with non–negative real numbers on the diagonal, V is an
also unitary matrix of order n×n which could be complex or
real. The diagonal entries of S are the singular values of M in
the decreasing order. Main advantage of SVD is that very few
values of host image changed that’s why little disturbance
occurred in an image and these small alteration in singular
values can be avoidable.
III. METHODOLOGY
A. Lifting Wavelet Transform
With pace of time so many techniques came into existence for
watermarking. Wavelet transform is a time domain localized
analysis method. Wavelet transform disintegrate the image
into four various spatial domains. Discrete wavelet transform
basically used to compress an image. Actually it is better
technology than discrete cosine transform.
Figure 2 Block Diagram of Proposed Architecture
These sub band are known as HH, HL, LL and LH. Out of
these one sub band is low frequency band and rest of three are
high frequency band. Figure 1 depicts the one level discrete
wavelet transform decomposition process. Lifting wavelet
transform can be also applied in same manner as DWT and
this is known as first level transform. In second levele
transform out of these four sub band, again one sub band is
selected as per requirement of application and then again
LWTapplied and further four sub band produce that why it is
known as second level LWT. One disadvantage of DWT is
that the use of larger DWT basis functions. Besides this
wavelet filters generate ringing and blurring noise near edges
in images. DWT disadvantage overcome by lifting wavelet
transform technique. Besides this comutational time of LWT is
fatser than DWT. Therefore LWT significantly reduces the
computation time and speed up the computation process.

Volume: 5 Issue: 11 216 – 221
_______________________________________________________________________________________________
218
_______________________________________________________________________________________
B. Walsh Hadmard ransform
Fourier transforms is basically used to convert time domain
into frequency domain and Hadamard transform is an example
of this class. Hadamard matrix is tremendous to calculate
various complicated problem into simple way. It executes an
orthogonal, symmetric, linear operation on 2m
complex
numbers or real number but it is known that Hadamard matrix
is purely real. The Hadamard transform can be regarded as
being built out of size-2 discrete Fourier transforms (DFTs). It
disintegrates an input vector into a superposition of Walsh
functions. Hadamard transform matrix is a square matrix and
which having only two element -1 and 1 besides this it is
orthogonal matrix. This transform is also known as WHT
(Walsh Hadamard transform). 𝐻1 is the smallest Hadamard
matrix, and it is defined as [15]
Higher size of matrix can be calculated by simply using this
elementary matrix.
In general any higher order matrix is computed by generalized
formula
IV. EXPERIMENTAL RESULT
SOFTWARE: MATLAB R2015A: It is powerful software
that provides an environment for numerical computation as
well as graphical display of outputs. In Matlab the data input is
in the ASCII format as well as binary format. It is high-
performance language for technical computing integrates
computation, visualization, and programming in a simple way
where problems and solutions are expressed in familiar
mathematical notation.
Figure 3 Experimental Dataset
Figure 4 Flow Chart of Embedding System
Figure 5 Complete output overview of Algorithm
Figure 6 Extracted cover and watermark image

Volume: 5 Issue: 11 216 – 221
_______________________________________________________________________________________________
219
_______________________________________________________________________________________
PSNR and RMSE value of watermarked and cover image
calculated by given equations
𝑅𝑀𝑆𝐸 𝑥 =
1
𝑁
||𝑥 − 𝑥^||2 =
1
𝑁
(𝑥 − 𝑥^)²
𝑁
𝑖=1
Where x is cover image, x^ is watermarked image, N is the
size of the cover image
𝑃𝑆𝑁𝑅 𝑥 =
10 𝑋log⁡((255))
𝑅𝑀𝑆𝐸(𝑥)
Where m is the maximum value of the cover image
Figure 7 Flow Chart of Extraction System
Figure 8 Process to get watermark image
TABLE I PSNR COMPARISON BETWEEN REF AND PROPOSED FOR
WATERMARKING.
Tick
Label
Cover
Image
Watermark
Image
Ref
PSNR
Proposed
PSNR
A Airplane House 52.1186 59.80
B Tulips Pepper 52.1670 52.91
C Pepper Airplane 52.1812 65.44
D Lena Cameraman 52.0408 59.04
E Baboon Lifting body 52.1232 55.24
F Bridge Boat 52.2080 51.56
TABLE II TIMECOMPARISON BETWEEN REF AND PROPOSED FOR
EMBEDDING.
Tick
Label
Watermarked
Image
Ref
Embedding
Time
Proposed
Embedding
Time
A Airplane 0.8413 0.5476
B Tulips 0.7334 0.5756
C Pepper 0.6598 0.5278
D Lena 0.5987 0.5948
E Baboon 0.6806 0.5341
F Bridge 0.6450 0.5396
Table II shows the comparison between Ref and proposed
scheme using Embedding Time. It describes the time of
adding two images using proposed algorithm.
TABLE III PSNR COMPARISON BETWEEN REF AND PROPOSED
WHEN DIFFERENT CHANNELS ARE CHOSEN TO EMBED.
Tick
Label
Ref PSNR (in dB) Proposed PSNR
Y Cb Cr Y Cb Cr
A 24.46 10.67 10.84 50.79 58.90 59.48
B 14.57 10.75 9.36 42.95 63.82 63.26
C 13.54 9.79 7.89 64.11 70.59 70.67
D 12.55 12.03 11.31 49.57 68.95 68.58
E 14.67 14.64 8.95 45.29 67.45 67.75
F 22.41 9.48 13.37 41.41 57.21 57.69
TABLE IV RMSE AFTER VARIOUS ATTACKS WHEN Y-CHANNEL
WAS USED FOR WATERMARKING.
Tick
Label
Cover
Image
Watermark
Image
Attacks
Blur Avg

Volume: 5 Issue: 11 216 – 221
_______________________________________________________________________________________________
220
_______________________________________________________________________________________
TABLE V RMSE AFTER VARIOUS ATTACKS WHEN CB-CHANNEL
WAS USED FOR WATERMARKING.
Tick
Label
Cover
Image
Watermark
Image
Attacks
Blur Avg
TABLE VI RMSE AFTER VARIOUS ATTACKS WHEN CR-CHANNEL WAS USED
FOR WATERMARKING.
Tick
Label
Cover
Image
Watermark
Image
Attacks
Blur Avg
Figure 9 PSNR Comparision of LWT WHT SVD and DWT DFT SVD (Base)
Figure 10 Processing Time Comparision of LWT WHT SVD and DWT DFT
SVD (Base)
V. CONCLUSION
In this research we proposed a two level LWT-WHT-SVD
Watermarking method on YCrCb color space. In base paper
DWT-DFT-SVD watermarking method is used and value of
peak signal to noise ratio is very less. Now in our research
prime focus is to enhance the value of PSNR. Now a day’s
privacy of people is prime concern and to shield their data.
Therefore to shield valuable information present on internet
there are various advanced techniques for example
watermarking technique, cryptography technique,
steganography and many more. On the other hand
watermarked images are robust against various numbers of
attacks like blurring, average, crop and Gaussian etc. In this
paper lifting wavelet transform applied in same manner as
DWT in base paper. In research paper two levels LWT
technique is used. When first level LWT applied on an image
then four sub band generated and in second level
transformation out of these four sub band, again one sub band
is selected as per requirement of application and again LWT
applied and further four sub band produced that why it is
known as second level LWT. One afterdeal of DWT is that it
uses larger basis functions. Besides this wavelet filters
generate ringing and blurring noise near edges in images. This
disadvantage of DWT is overcome in LWT. Other advantages
of LWT are that it significantly reduces the computation time
and speed up the computation process. Finally in this research
paper value of PSNR enhanced as compared to base paper and
processing time also reduced. This algorithm could be
extended to watermark the images using color images as the
watermarks. Further this research can be extend on different
color space for example HSV, CMYK to find the color space
with maximum efficiency using different techniques like
FWT, SWT.
ACKNOWLEDGMENT
With a deep sense of gratitude and heartiest honor, I would
like to express my immense thanks to Mr. Rakesh Tayal,
Member BOG, Panipat Institute of Engineering and
Technology, Samalkha for providing me all the facilities,
valuable and sustained guidance, constant encouragement and
careful supervision during the entire span of time which made
the research successful.
REFERENCES
[1] Advith J, Varun K R and Manikantan K, “Novel Digital Image
Watermarking Using DWT-DFT-SVD in YCbCr Color Space”
International Conference on Emerging Trends in Engineering,
Technology and Science (ICETETS), 24-26 Feb 2016,
DOI: 10.1109/ICETETS.2016.7603032
[2] Nikhil Purohit, M. Chennakrishna and K. Manikantan, “Novel Digital
Image Watermarking in SWT+SVD Domain” Proceedings of the
International Conference on Signal, Networks, Computing, and Systems,
in Electrical Engineering 395, DOI 10.1007/978-81-322-3592-7_2
[3] Rahim Ansari, Mrutyunjaya M Devanalamath, K. Manikantan and S.
Ramachandran, “Robust Digital Image Watermarking Algorithm in
DWT-DFT-SVD Domain for Color Images” 2012 International

Volume: 5 Issue: 11 216 – 221
_______________________________________________________________________________________________
221
_______________________________________________________________________________________
Conference on Communication, Information & Computing Technology
(ICCICT), Oct. 19-20, Mumbai, India
[4] Amit Kumar Singh, Mayank Dave and Anand Mohan, “Hybrid
Technique for Robust and Imperceptible Image Watermarking in DWT–
DCT–SVD Domain” The National Academy of Sciences, pp. 351–358
India 2014, 19 July 2014
[5] V. Santhi and Arunkumar Thangavelu, “DWT-SVD Combined Full
Band Robust Watermarking Technique for Color Images in YUV Color
Space” International Journal of Computer Theory and Engineering, Vol.
1, No. 4, October 2009 , 1793-8201
[6] Huang–Chi Chen, Yu–Wen Chang, Rey–Chue Hwang, “A
Watermarking Technique based on the Frequency Domain,”
Journal of Multimedia, Vol. 7, No. 1, 2012.
[7] Huang–Chi Chen, Yu–Wen Chang, Rey–Chue Hwang, “A
Watermarking Technique based on the Frequency Domain,”
Journal of Multimedia, Vol. 7, No. 1, 2012.
[8] Ante Poljicak, Lidija Mandic, Darko Agic, “Discrete Fourier
transform–based watermarking method with an optimal
implementataion radius,” Jounal of electronic imaging, 2011
[9] Awanish Kr Kaushik, “A Novel Approach for Digital Watermarking of
an Image Using DFT,” IJECSE, Vol. 1, No. 1, pp. 35–41, 2012
[10] Pallavi Patel, D. S. Bormane, “DWT Based Invisible Watermarking
Technique for Digital Images,” International Journal of Engineering and
Advanced Technology, Vol. 2, No. 4, 2013
[11] Anuradha, Rudresh Pratap Singh, “DWT Based Watermarking
Algorithm using Haar Wavelet,” International Journal of Electronics and
Computer Science Engineering, Vol. 1, No. 1, 2012.
[12] R. A. Ghazy, M. M. Hadhoud, M. I. Dessouky, N. A. El–Fishawy,F. E.
Abd El–Samie, “Performance evaluation of block based svd image
watermarking,” Progress In Electromagnetic Research, pp. 147–159,
2008.
[13] Kuo–Liang Chung, Wei–Ning Yang, Yong–Huai Huang, Shih–Tung
Wu, Yu–Chiao Hsu, “On SVD based watermarking algorithm,” Applied
Mathematics and Computation, pp. 54–57, 2007.
[14] Chih–Chin Lai, Cheng–Chih Tsai, “Digital Image Watermarking using
Discrete Wavelet Transform and Singular Value Decomposition,” IEEE
Transactions on Instrumentation and Measurement, Vol. 59, No. 11,
2010.
[15] Piyush Pandey, Rakesh Kumar Singh, “Novel Digital Image
Watermarking Using LWT-WHT-SVD in YCbCr Color Space”
International Journal of Innovative Research in Computer and
Communication Engineering, Vol. 5, Issue 6, June 2017

Hybrid Method For Image Watermarking Using 2 Level LWT-Walsh Transform-SVD in YCbCr Color Space

More Related Content

What's hot (19)

Similar to Hybrid Method For Image Watermarking Using 2 Level LWT-Walsh Transform-SVD in YCbCr Color Space (20)

More from rahulmonikasharma (20)

Recently uploaded (20)

Hybrid Method For Image Watermarking Using 2 Level LWT-Walsh Transform-SVD in YCbCr Color Space