Feature Based watermarking algorithm for Image Authentication using D4 Wavelet Transform

Signal & Image Processing : An International Journal (SIPIJ) Vol.2, No.3, September 2011
DOI : 10.5121/sipij.2011.2309 89
Feature Based watermarking algorithm for Image
Authentication using D4 Wavelet Transform
Kartik Sau1
Amitabha Chanda2
Milan Pal3
1
Department of Computer Science & Engineering, Budge Budge Institute of technology,
BBIT Nischintapur. Kolkata 7000137, WBUT, India.
e-mail: kartik_sau2001@yahoo.co.in
2
Department of Computer Science & Engineering, Guest faculty, UCSTA, University of
Calcutta, India.
e-mail: amitabha39@yahoo.co.in
3
Department of Computer science & Engineering, Institute of engineering and
Management, Salt Lake, WBUT, Kolkata -700091, India.
e-mail: milanpal2005@gmail.com
ABSTRACT:
In this paper we propose a new watermarking schema i.e. the combination of color space and wavelet
transform. Watermarking is a technique that authenticates a digital picture by hiding the secret information
into the image. Now, a lot of algorithms and methods have been developed for greyscale images but the
particularities of color spaces have to be studied. On the other hand, the wavelet transform allows different
possibilities of integrating a mark because of the uses of different parameters: the scale of decomposition,
size, shape and localisation of the mark, and the used color space, etc. The RGB (Red, Green, and Blue)
values of each pixel of the host color image as well as the color key image are converted to HSV (Hue,
Saturation, and Value) values. Then the orthogonal D4 Wavelet transform is applied in each plate of host
image and key image. Now insert the key component into appropriate blocks of host image’s different
plates. The experimental results show the effectiveness of our algorithm.
KEYWORDS:
Digital watermarks, D4wavelet transform, color space, human visual system.
1. INTRODUCTION
Nowadays our life is becoming more and more permeated with Internet or widely-used
communication systems and the technology allows an easy access to an almost unlimited amount
of information. On the other hand with the increasing use of internet and effortless copying,
tempering and distribution of digital data, copyright protection for multimedia data has an
important issue. Digital watermarking emerged as a tool for protecting the multimedia data from
copyright infringement. Before distributing the data, the owner embeds an invisible signature, the

90
watermark, into the host source (audio, text, image, or video) using a secret key. After embedding
the watermark, there should be no perceptual degradation. These watermarks should not be
removable by unauthorized person and should be robust against intentional and
unintentional attacks.
In the literature, several techniques have been developed for watermarking. In [1], three coding
methods for hiding electronic marking in document were proposed. In [2]–[5], the watermarks are
applied on the spatial domain. The major disadvantage of spatial domain watermarking is that a
common picture cropping operation may eliminate the watermark. Other than spatial domain
watermarking, frequency domain approaches have also been proposed. In [6], a copyright code
and its random sequence of locations for embedding are produced, and then superimposed on the
image based on a JPEG model. In [7], the spread spectrum communication technique is also
used in multimedia watermarking.
In this paper, we propose a new technique for embedding digital watermarks with
visually recognizable patterns into the images. Since, in daily life, one claim a document,
a creative work, and so on, by signing one’s signature, stamping a personal seal or an
organization’s logo, such kinds of visually recognizable patterns are more intuitive for
representing one’s identity than a sequence of random numbers is.
First of all, we transform the RGB color space of host image as well as secrete key into
corresponding HSV color space. The components are not changed, because changes in the H
components could change the color balance between HSV components. The orthogonal wavelet
transform [8] is applied on the each plate of the image since it gives the better results than Fourier
transform. Then we distinguee the background and foreground of the host image and we
embedded the secrete key into the appropriate block of background host image. This is selected
by appropriate edge detector.
ORGANIZATION OF PAPER:
The paper is organized as follows: In section 2 we briefly discuss the Wavelet transform. In
section 3 the general watermarking algorithm is described. In section 4 we propose an algorithm
for watermarking and its process flow. Some experimental results are shown in section 5 along
with few remarks. In section 6 some conclusions are drawn.
2. WAVELETS TRANSFORM
Wavelets transform is an efficient tool to represent an image. It allows multi resolution analysis
of an image. The aim of this transform is to extract relevant information from an image [9].
Wavelet transform has received considerable attention in the field of image processing due to its
ability in adapting to human visual characteristics mentioned in [10]. It is most powerful and
widely used tool in the field of image processing. It divides the signal into number of segments;
each corresponds to a different frequency band. The continuous wavelet transform of a one
dimension signal x(t) is given by

91
dt
t
x
∫ 





=
−
α
α
ψ
a
b
-
t
*
)
(
a
1
b)
a,
(
W f (1)
The continuous wavelet transform is a function of two variables a and b. Here a is scaling
parameter and b is a shift parameter. Ψ(x) is the mother wavelet or the basis function and it is
used as a prototype for generating all the basis functions. The transformation parameter or the
shift parameter b gives the time information in the wavelet transformation. It indicates the
location of the window as it is shifted through the signal. The scale parameter a gives the
frequency information in the wavelet transforms. A low scale corresponds to wavelets of smaller
width, which gives the detail information in the signal. A high scale corresponds to the wavelets
of larger width which gives the global view of the signal. The IWT is given by
a
c
da
db
t
b
a b
a 2
)
(
)
,
(
1
X(t) ,
f
W ψ
α
α
ψ
∫ ∫
−
=
,
∞
<
Ψ
= ∫
∞
∞
−
Ψ
w
dw
w
C 2
|
)
(
|
(2)
The wavelet transform of a signal using the CWT is obtained by changing the scale of the
analysis of the window, shifting the window in time, multiplying the signal & integrating the
result over all time. According to the orthogonal wavelet transform, the S values are decomposed
by the following equation
y)
x,
(
y)
x,
(
=
y)
F(x,
1
-
n
0
j 0
1
-
n
0
j
j
A ψ
φ jk
n
k
jk
jn D
∑ ∑
∑ = =
=
+ (3)
Here φ is the scale function and Ψ is the wavelet transform. The former component of the
decomposition is the approximate components and the latter one is the detail components. Aj are
the approximate and Djk are the detail components. As the transformation is an orthogonal
transformation, each Aj is in the range [0,255].
The image is reconstructed by using the inverse Wavelet transformation as indicated by the
equation:
)
,
(
)
,
(
)
,
(
'
1
0 0
1
0
' y
x
y
x
y
x
F
jk
n
j
n
k
jk
jn
n
j
j D
A ψ
φ ∑ ∑
∑
−
= =
−
=
+
= (4)
3. GENERAL METHOD OF WATERMARKING
Abstractly, a watermarking schema is six-tuple (X, W, K, G, E, D) where [11]
1. X represents the set of original or host images.
2. W is the set of all watermarks w such that k)
G(x,
w
K,
k
X,
x =
∈
∈
∃ .
3. K is the set of watermark key.
4. G be the algorithm that generates W using K and
X: k)
G(x,
w
W,
K
X
:
G =
→
× .

92
5. E denotes the encoding algorithm that embeds a watermark w in an image x with some
strength α :
)
,
,
(
,
: α
w
x
E
x
X
W
X
E w =
→
ℜ
×
×
6. D is the decoding algorithm that detects whether a watermarking w is present in an image
x:
otherwise
0
in x
exists
w
if
1
w)
D(x,
{0,1},
W
X
:
D
=
=
→
×
4. PROPOSED METHOD AND ITS PROCESS FLOW
The proposed watermarking algorithm is described as follows:
Step1: transform the RGB color space to HSV color space of host image as well as key image.
Step2: select the proper blocks for inserting key component using suitable edge detection key
technique in such way that H block of host maps into S block of key image, V block of host
image maps into H block of key image and S block of host image into V block of key image.
Step4: Let T be the transformation of H,S and V component of host and key images, compute
T(x) to get transform coefficient h0 h1, h2and h3 such that
4
0
2
3
1+
=
h , 4
1
2
3
3+
=
h ,
4
2
2
3
3 −
=
h and
4
3
2
3
1−
=
h
Each step of the wavelet transform applies the scaling function to the data input. If the original
data set has N values, the scaling function will be applied in the wavelet transform step to
calculate N/2 smoothed values. In the ordered wavelet transform the smoothed values are stored
in the lower half of the N element input vector. The wavelet function coefficient values are: g0 =
h3, g1 = -h2, g2 = h1 and g3 = -h0. Each step of the wavelet transform, applies the wavelet
function to the input data. If the original data set has N values, the wavelet function will be
applied to calculate N/2 differences (reflecting change in the data). In the ordered wavelet
transform the wavelet values are stored in the upper half of N element input vector. The scaling
and wavelet functions are calculated by taking the inner product of the coefficients and four data
values. The equations are shown below. The Daubechies D4 scaling function is
]
3
2
[
]
2
2
[
]
1
2
[
]
2
[
]
[ 3
2
1
0
3
2
3
2
2
2
1
2
1
2
0
+
+
+
+
+
+
=
+
+
+
= +
+
+
i
s
h
i
s
h
i
s
h
i
s
h
i
a
s
h
s
h
s
h
s
h
a i
i
i
i
i
(5)
Daubechies D4 wavelet function is
]
3
2
[
]
2
2
[
]
1
2
[
]
2
[
]
[ 3
2
1
0
3
2
3
2
2
2
1
2
1
2
0
+
+
+
+
+
+
=
+
+
+
= +
+
+
i
s
g
i
s
g
i
s
g
i
s
g
i
c
s
g
s
g
s
g
s
g
c i
i
i
i
i
(6)

93
Each iteration in the wavelet transform step calculates a scaling function value and a wavelet
function value. The index i is incremented by two with each iteration, and new scaling and
wavelet function values are calculated. This pattern is discussed on the section 3.
Step3: If the blocks satisfy the threshold value, then insert the corresponding key component into
selected blocks randomly.
Step6: output xw= )
..........
,
( '
2
'
1
'
1
n
h
h
h
T −
Step7: Now modified H component, modified S and modified V components are converted to
RGB color space of host image.
Step8: show the watermarked image.
Block diagram of encoding technique is as follows:
Figure: 1(Process flow of encoding technique)
The key extraction from watermarked image algorithm is described as follows:
Step1: convert the RGB color space of watermarked image to HSV color space of watermarked
image.
Step2: Compute )
( w
x
T to get the transform coefficients
∧
∧
∧
n
2
1 v
.
..........
v
,
v

94
Step3: Extract the watermarked w’
from the transform coefficient
∧
∧
∧
n
2
1 v
.
..........
v
,
v .
Step4: Compute C (w, w’
)>t where t be the detection threshold.
∑
=
⊕
−
=
mn
i
i
i w
w
w
w
C
1
'
'
1
)
,
( Such that
C (w,w’
)=1 if w exist in x
0 elsewhere.
Step6: now extracted H, S & V components are converted to RGB color space on each block.
Step7: show the key.
Block diagram of decoding technique is as follows:
Figure:2(Process flow of decoding technique)

95
6. EXPERIMENTAL RESULT
We present a novel algorithm for color image watermarking, which consists of orthogonal
wavelet transform of HSV color space. We performed fidelity tests to analyze the
unobtrusiveness of the watermarks after watermark embedding, whether perceptual distortion
occurred to the host images or not, i.e., why we give the few examples, which shows the
effectiveness of our proposed method.
We also tested the robustness against standard noise attacks to the watermarked images which is
obtained by PSNRs [12] between host image and watermarked image under standard noise
degradations, which is shown in Table 1 and table 2. In this paper all experiments are done in
Matlab 7.1 Version.
The source image set X is given as follows.
Image: 1 Image: 2 Image: 3

96
The secret key set K is given bellow:
The watermark image set W is given by
Image:1 Image:2 Image:3

97
Image:7 image:8 image:9
The extracted secret key set K from watermark image set W is given as follows:

98
Sl no. Image
name
Key name No of key in
watermarked image &
it’s possitions
PSNR
1 Image:1 ieee logo 9(1,2,3,4,5,8,11,15,16) 51.3038
2 Image:2 wipro logo 15(1,2,3,4,5,7,8,9,10,1
1,12,13,14,15,16)
57.1478
3 Image:3 cts logo 5(3,4,11,12,16) 57.8626
4 Image:4 wipro logo 6(1,3,5,9,13,16) 61.5340
5 Image:5 infosys
logo
11(1,2,3,4,5,8,9,10,12,
13,16)
53.8647
6 Image:6 webel logo 14(1,2,3,4,5,6,7,8,9,11
,12,13,14,16)
58.7990
7 Image:7 cts logo 6(1,3,5,8,9,16) 51.5364
8 Image:8 infosys 6(1,2,4,5,8,16) 59.6433
9 Image:9 webel logo 5(1,2,3,4,16) 54.4222
Table:1(various position of key in watermarked image)
Sl
no.
Image
name
PSNR
after Salt
& pepper
attack
PSNR after
Gaussian
attack
PSNR
after
compress
ion
1 Image:1 51.1744 51.2100 51.3038
2 Image:2 56.9560 57.1236 57.1478
3 Image:3 57.5971 57.8382 57.8626
4 Image:4 60.9982 61.2032 61.5340
5 Image:5 53.7788 53.7611 53.8647
6 Image:6 58.5117 58.7126 58.7990
7 Image:7 51.4622 51.3570 51.5364
8 Image:8 59.2868 59.3065 59.6433
9 Image:9 54.2168 54.2727 54.4222
Table:2(PSNR calculation after various attacks)

99
The experimental results shows that, there is no perceptual distortion in original image and
watermarked image, which means that scheme has satisfied the criteria that an efficient
watermark should be unobtrusive, discreet and easily extracted.
6. CONCLUSION
This paper presented a novel algorithm, which is based on color space and wavelet transform.
The key is inserting in the appropriate plates of the images depending upon the nature of the
plates. Compared with other techniques, our technique has higher capabilities and lower
computational cost under the same robustness. It can be extended to multimedia watermarking.
The performance of the watermarking is evaluated in common image processing attacks such as
additive noise, filtering, histogram equalization, JEPG compression, scaling and rotation.
Experimental result demonstrates that watermark is robust against those attacks.
REFERENCES
[1] J. T. Brassil, S. Low, N. F. Maxemchuk, and L. O’Gorman, “Electronic marking and identification
techniques to discourage document copying,” IEEE J. Select. Areas Commum., vol. 13, pp. 1495–
1504, Oct. 1995.
[2] I. Pitas and T. H. Kaskalis, “Applying signatures on digital images,” in Proc. IEEE Nonlinear Signal
and Image Processing, June 1995, pp. 460–463.
[3] O. Bruyndonckx, J. J. Quisquater, and B. Macq, “Spatial method for copyright labeling of digital
images,” in Proc. IEEE Nonlinear Signal and Image Processing, June 1995, pp. 456–459.
[4] S. Walton, “Image authentication for a slippery new age,” Dr. Dobb‘s J., pp. 18–26, Apr. 1995.
[5] W. Bender, D. Gruhl, and N. Morimoto, “Techniques for data hiding,” Proc. SPIE, vol. 2420, p. 40,
Feb. 1995
[6] E. Koch and J. Zhao, “Toward robust and hidden image copyright labeling,” in Proc. IEEE
Nonlinear Signal and Image Processing, June 1995, pp. 452–455.
[7] I. J. Cox, J. Kilian, T. Leighton, and T. Shammoon, “Secure spread spectrum watermarking for
multimedia,” Tech. Rep. 95-10, NEC Res. Inst., Princeton, NJ, 1995.
[8] M. J. Shensa, “Discrete wavelate transform: wedding the trous Mallat algorithms”, IEEE Trans. On
signal processing Vol. 40, No. 10, pp, 2464-2482.
[9] C.C Yang & J. J. Rodrgues, “Efficient luminance and saturation processing technique for bypass
color coordinate transformation”, Proc. IEEE Conf. on systems, Man and Cybernetics, Bogota,
Columbia, 1995, 56-67
[10] B.A. Tomas and R.N.Strickland, “Color image enhancement using spatially adaptive saturation
feedback”, IEEE Ind., Conf. on Image Processing, Vol. 3, pp, 30-33, 1997
[11] G.Voyatzis and I.Pitas,”the use of watermarks in the protection of multimedia products”, proceedings
of the IEEE, vol. 87,no. 7,July 1999, pp.1197-1207.

100
[12] K.Hameed, A.Mumtaz and S.A.M.Gilani “Digital image watermarking in the wavelate transform
domain”, World Academy of Science and Technology,13 2006
Authors
Kartik Sau: Associate Professor, Department of Computer Science & Engineering, Budge Budge Institute
of technology. He completed his B.Sc. in mathematics from RK Mission Vidyamandira, University of
Calcutta. And M. Sc. in the same subject from Indian Institute of Technology, Kharagpur. M. Tech in
Computer Science from Indian School of Mines, Dhanbad. Currently he is Head of the Department of
BBIT and member of Govt. body of the same Institute; He has presented many papers in International and
National journals and Conference. His area of interest includes Digital Image processing, Artificial
Intelligence, Pattern Recognition, Soft computing, etc. He has more than eight years teaching and research
experience in his area of interest.
Dr. Amitabha Chanda: Visiting Professor; Department of Computer Science, UCSTA; Calcutta
University. He completed his B.E in Chemical Engineering (Jadavpur), M.A. in Pure Mathematics and
Ph.D in Mathematics from university of Calcutta. He was a faculty member of Indian Statistical Institute
(ISI), Kolkata. Now he is also guest faculty member of ISI, Kolkata; Department of Computer Science,
Rajabazar Science College, Kolkata. His area of interest includes Digital Image processing, Pattern
Recognition Fuzzy logic, Genetic Algorithms, Computer Graphics, Control system Turbulence, Fractal,
multifractals, Clifford Algebra, Nonlinear dynamics. He has more than fifty years teaching and research
experience in his area of interest. He has presented more than 100 papers in International and National
journals and Conference. Dr. Chanda is an Associate member of American Mathematical Society.
Milan Pal: He received his B. Tech degree in Computer Science & Engineering from Bankura Unnayani
Institute of Engineering, WBUT, India. He completed his M.Tech in Computer Science & Engineering
from Institute of Engineering & Management, Salt Lake Kolkata, India. His area of interest includes Digital
Image Processing, Pattern Recognition Fuzzy logic, Genetic Algorithms and DBMS.

Feature Based watermarking algorithm for Image Authentication using D4 Wavelet Transform

More Related Content

Similar to Feature Based watermarking algorithm for Image Authentication using D4 Wavelet Transform (20)

Recently uploaded (20)

Feature Based watermarking algorithm for Image Authentication using D4 Wavelet Transform