The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.

1, Maich 2O1O

10.5121/ijngn.2010.2104 37
Image Compression and Watermarking scheme
using Scalar Quantization
Kilari Veera Swamy
1
B.Chandra Mohan
2
Y.V.Bhaskar Reddy
3
S.Srinivas Kumar
4
1
Professor and HOD,ECE Dept., QISCET, Ongole, India
[email protected]
2
Professor and HOD,ECE Dept., BEC, Bapatla, India
[email protected]
3
Associate Professor, ECE Dept., QISCET, Ongole, India
[email protected]
4
Professor, ECE Dept., JNTU, Kakinada, India
[email protected]
Abstract: This paper presents a new compression technique and image watermarking algorithm based on
Contourlet Transform (CT). For image compression, an energy based quantization is used. Scalar quantization
is explored for image watermarking. Double filter bank structure is used in CT. The Laplacian Pyramid (LP) is
used to capture the point discontinuities, and then followed by a Directional Filter Bank (DFB) to link point
discontinuities. The coefficients of down sampled low pass version of LP decomposed image are re-ordered in a
pre-determined manner and prediction algorithm is used to reduce entropy (bits/pixel). In addition, the
coefficients of CT are quantized based on the energy in the particular band. The superiority of proposed
algorithm to JPEG is observed in terms of reduced blocking artifacts. The results are also compared with
wavelet transform (WT). Superiority of CT to WT is observed when the image contains more contours. The
watermark image is embedded in the low pass image of contourlet decomposition. The watermark can be
extracted with minimum error. In terms of PSNR, the visual quality of the watermarked image is exceptional.
The proposed algorithm is robust to many image attacks and suitable for copyright protection applications.
Keywords- Contourlet Transform, Directional Filter Bank, Laplacian Pyramid, Topological re-ordering,
Quantization.
1.INTRODUCTION
There are many applications requiring image compression, such as multimedia, internet,
satellite imaging, remote sensing, and preservation of art work, etc. Decades of research in this area
has produced a number of image compression algorithms. Most of the effort expended over the past
decades on image compression has been directed towards the application and analysis of orthogonal
transforms. The orthogonal transform exhibits a number of properties that make it useful. First, it
generally conforms to a parseval constraint in that the energy present in the image is the same as that
displayed in the images transform.
Second, the transform coefficients bear no resemblance to the image and many of the
coefficients are small and, if discarded, will provide for image compression with nominal degradation
of the reconstructed image. Third, certain sub scenes of interest, such as some targets or particular
textures, have easily recognized signatures in the transform domain [1]. The Discrete Cosine
Transform has attracted widespread interest as a method of information coding.
The ISO/CCITT Joint Photographic Experts Group (JPEG) has selected the DCT for its
baseline coding technique [2]. The fidelity loss in JPEG coding occurs entirely in quantization and
much of the compression is gained by run length coding of coefficients which quantize to zero.
Methods for determining the quantization table are usually based on rate-distortion theory. These
methods do achieve better performance than the JPEG default quantization table [3,4,5]. However,
the quantization tables are image-dependent and the complexity of the encoder is rather high. WT has
attracted wide spread interest as a method of information coding [6,7]. JPEG-2000 has selected the
The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
38
WT for its baseline coding technique [8]. Lossless image compression algorithms using prediction are
available in the literature [9,10].
Multimedia (image, video clip, text file, etc.,) data transmission over untrusted public
networks is gaining importance in the recent past. For secure transaction over the internet, multimedia
object has to be protected. Digital watermarking [22] is a viable solution for this problem. Digital
image watermarking is a technique wherein, the watermark image is embedded in the cover image
using some known watermark embedding algorithms. At the receiving side, by using known
extraction algorithm, the watermark image can be obtained.
There are two fundamental requirements for any watermarking scheme. Firstly, the visual
quality of the watermarked image should be good. Secondly, the watermark image should survive to
many image attacks, which may be intentional or unintentional. That means, the extracted watermark
should be recognizable even after modifying the watermarked image. Based on the availability of the
cover image and watermark image at the receiver, the watermarking schemes are classified as blind
(oblivious) and non blind (non oblivious).
Watermarking algorithms can also be classified based on the domain used for watermark
embedding. Accordingly, there are two popular techniques, namely spatial domain watermarking and
transform domain watermarking techniques. Spatial domain watermarking techniques are useful for
higher data embedding applications. Transform domain watermarking techniques are suitable in
applications where, robustness is of prime concern. There are several transform domain
watermarking schemes available in the literature. Discrete Cosine Transform (DCT) [23], Discrete
Fourier Transform (DFT) [24], Discrete Wavelet Transform (DWT) [25], Discrete Hadamard
Transform (DHT) [26], Contourlet Transform (CT) [27], and Singular Value Decomposition(SVD)
[28] are some useful transforms for image processing applications.
In this work, CT [11] is considered for image compression and watermarking. For image
compression, low pass version of LP decomposed image is re-ordered and prediction algorithm is
applied. Further, all coefficients of CT are quantized using scalar quantization based on the energy in
that particular band. CT is applied to the entire image. Hence, blocking artifacts gets reduced in CT
than JPEG compressed image. If an image contains more contours, then CT outperforms WT. The
watermark image is embedded in the low pass version of LP decomposition using scalar quantization
technique.
2.JPEG OVERVIEW
Implementation of JPEG is given below
- The image is partitioned into 8x8 blocks of pixels
- Working from left to right , top to bottm the DCT is applied to each block
- Each block is compressed using quantization
- Coding is applied for zig zag scanned coefficients
3.CONTOURLET TRANSFORM
CT uses double filter bank structures for obtaining sparse expansion of typical images having
smooth contours [12]. In this double filter bank structure, Laplacian Pyramid (LP) is used to
capture the point discontinuities, and Directional Filter Bank (DFB) is used to link these point
discontinuities into linear structures. First stage is LP decomposition and second stage is
DFB decomposition and is shown in Figure 1. One level LP decomposition is shown in
Figure 2. Let ) , ( j i f (0<i , j<N) represent the original image and its down sampled low pass
filtered version by ) , ( j i f
lo
. The prediction error is given by

) , ( j i Lo = ) , ( j i f - ) , ( ' j i f
lo
(1)
The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
39


Figure 1: Contourlet Filter bank
Here,
) , ( ' j i f
lo
represents the prediction of
) , ( j i f
. Encoding step is performed on
) , ( j i Lo
as it is
largely decorrelated and requires less number of bits than
) , ( j i f
. In Equation 1,
) , ( j i Lo
represents a
band pass image. Further encoding can be carried by applying Equation 1 on
) , ( j i f
lo
iteratively to
get
) , (
1
j i f
l
,
) , (
2
j i f
l
,. . . . . . .
) , (
ln
j i f
, where n represents the pyramidal level number. In Figure 2,
H and G represent analysis and synthesis filters respectively. M represents sampling matrix. One level
LP reconstruction is shown in Figure 3. The band pass image obtained in LP decomposition is further
processed by a DFB. A DFB is designed to capture the high frequency content like smooth contours
and directional edges. This DFB is implemented by using a k-level binary tree decomposition that
leads to 2
k
sub bands. In CT, DFB is implemented using quincunx filter bank with fan filters [13]. The
second building block of the DFB is a shearing operator, which amounts to just reordering of image
samples.
CT gives two important properties [14].
- Directionality. The representation should contain much more directions.
- Anisotropy. To capture smooth contours in images, the representation contains basis elements
using a variety of elongated shapes.

These two properties are useful for image compression, image watermarking, and Content Based
Image Retrieval.




Figure 2: One level LP decomposition

M
Low pass
image
) , ( j i f
lo


Band pass
image
) , ( j i Lo
Original
Image
) , ( j i f

M
G
) , (
'
j i f
lo

.
.
.
.
.
.
.
LP
stage

(2,2)
Image
DFB
stage LP
stage
.
.
.
.
.
.
.

H
The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
40


Figure 3: One level LP reconstruction

I. PROPOSED IMAGE COMPRESSION METHOD
In this method, image is decomposed using CT to obtain different bands. Down sampled low pass
version of LP decomposed image is compressed in lossless procedure by steps, viz., topological re-
ordering of coefficients, scanning, prediction and calculation of residues. Finally, all coefficients of
CT are quantized.
3.1Topological re-ordering
The rearrangement of coefficients is based purely upon coefficient position, rather than a function of
coefficient value, hence it is designated as topological re-ordering [15]. The advantage of re-ordering
is better accessibility of the successive coefficients for the estimation of current coefficient value.
Various re-ordering possibilities are given in an earlier work [16]. Re-ordering scheme used in this
work is given in Figure 4.







Figure 4: Square re-ordering scheme.

3.2Scanning, prediction and residues
Transformed image coefficients can be processed by different scanning techniques [17]. The
raster scan consists of processing the matrix by scanning the image matrix from left to right within
each row and top to bottom in a matrix. Ideally, the value guessed for the current coefficient P
should depend on its neighbours. Notation of neighbouring coefficients of P is shown in Figure
5.Prediction [18] is based on one or combination of the neighbouring coefficients. West coefficient
(W) is used for estimating the current coefficient value P. This is a simple technique when the
coefficient values are uniform.
_

M
Low pass
image
) , ( j i f
lo

Band pass
image
) , ( j i Lo

G
H
M
+
+
Original Image
-
The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
41



Figure 5: Notation of neighbouring coefficients of P.

Residual is the difference between actual and estimated coefficients. Pseudo code for the algorithm is
given below to find the residual matrix for low pass version of LP decomposed image.
Y= Leasterror (X)
Input(X): Low pass version of LP decomposition
Output(Y):Residual(current coefficient predicted value)
m: rows, n: columns
{
Let Y=
R=Re-order(X)
Scan R using scan path and at each coefficient P do
{
for the first coefficient in image predicted value (Q)=0
Y=P-Q
else predicted value (Q) = W
Y=P-W
}
Return Y
}

The reverse of the above procedure is considered to reconstruct the original re-ordered matrix. In both
the cases, if P is the first coefficient to be scanned in the beginning of the matrix then the error at P
is defined to be of same value. The reverse process of topological re-ordering is used to reconstruct
the original matrix.
3.3Quantization
Energy of a sub band ) , ( j i , N j i , 0 is computed by
=

i j
j i
2
) , ( (2)
Based on the energy value, each band is divided by a suitable number. The band with higher energy
values is divided with less value and vice versa. All bands of CT decomposed image are quantized
using scalar quantization. Quantized coefficients are rounded to get better compression. At the
reconstruction, each band is multiplied with corresponding number.

W

P


E


NW

N

NE
The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
42
4.PROPOSED IMAGE WATERMARKING ALGORITHM

The image watermark embedding algorithm is as follows.

The cover image ) , ( j i f is decomposed into low pass image and directional subbands by
using CT decomposition.
The low pass image ) , ( j i f
lo
coefficients are quantized using the following rule
) ), , ( mod( Q j i f z
lo
=
If 0 ) , ( = j i w & 2 / Q z <
No modification in ) , ( j i f
lo


If 0 ) , ( = j i w & 2 / Q z >=
) , ( j i f
lo
= 2 / ) , ( Q j i f
lo


If 1 ) , ( = j i w & 2 / Q z >=
No modification in ) , ( j i f
lo


If 1 ) , ( = j i w & 2 / Q z <
) , ( j i f
lo
= 2 / ) , ( Q j i f
lo
+

Here, Qis the quantization coefficient and may be selected based the experimentation on cover
image. This is usually a trial and error process. After modifying ) , ( j i f
lo
, inverse CT is applied with
the modified ) , ( j i f
lo
and the watermarked image )' , ( j i f is obtained.

The image watermarking extraction algorithm is as follows.

The watermarked image )' , ( j i f is decomposed into low pass image and directional
subbands by using CT decomposition. The watermark image is extracted by using the
following rule
) , )' , ( mod( ' Q j i f z
lo
=
If 2 / Q z < , 0 ) , ( = j i w
If 2 / Q z >= , 1 ) , ( = j i w
5.EXPERIMENTAL RESULTS
Experiments are performed on six grey images to verify the proposed method. These six
images are represented by 8 bits/pixel and size is 512 x 512. Images used for experiments are shown
in Figure 6.


Figure 6. Images(Lena, Mandrill, Zelda, Peppers,Airplane, Barbara).

The entropy (E) [19] is defined as
The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
43
) ( log ) (
2
e p e p E
s e

= (3)
where, s is the set of processed coefficients and p(e) is the probability of processed coefficients . By
using entropy, number of bits required for compressed image is calculated. Compression ratio is
defined as

2
1
n
n
C
R
= (4)
let
1
n ,
2
n denote the number of bits in the original and compressed images.
An often used global objective quality measure is the mean square error (MSE) defined as


( ) ( ) [ ]
2
1 1
'
, ,
) )( (
1

= =
=
n
i
m
j
j i f j i f
m n
MSE
(5)
where, nxm is the number of total pixels. ( ) j i f , and ( )
'
, j i f are the pixel values in the original and
reconstructed image. The peak to peak signal to noise ratio (PSNR in dB) [20] is calculated as


MSE
PSNR
2
10
255
log 10 =

(6)
usable grey level values range from 0 to 255. The proposed method and JPEG methods are lossy
compression techniques. JPEG process results in both blurring and blocking artifacts. Blocking effect
occurs due to the discontinuity of block boundaries. The blockiness is estimated as the average
differences across block boundaries. This feature is used to constitute a quality assessment model.
Accordingly, the score (S) of the image [21] is calculated. The score typically has a value between 0
and 10 (10 represents the best quality, 0 the worst).The score is calculated using the formulae

3 2 1
Z A B S + = (7)

where , , 1 , 2 , 3 are the model parameters that must be estimated with the subjective test data.
B is related to the blockiness, A is the activity measure (relative blur) and Z is the zero-crossing (ZC)
rate.
Experiments are performed with JPEG 9-7 filters because they have been shown to provide
the best results for images. 9-7 filter coefficients are used both for pyramidal filter and directional
filter. LP decomposition n=(0,1,2) is considered for experiments. At each successive level, a number
of directional sub bands are changed. Finally, LP decomposition generates 64x64 lowpass version of
the original image and three 64x64, two 128x256 , four 256x256 directional sub bands are generated.
Energy is calculated for all bands. For Lena image, energy value of each band is 1111300, 1178,
3854, 1047, 297, 131, 20, 34, 12, 17 respectively. Hence, higher energy band is divided by less value
number and lower energy value is divided by higher value number. For simplicity, lowpass band is
divided with value 2, directional sub bands are divided with 4 and 8 based on the energy value. In
both methods (JPEG, proposed) coding part is not considered. Entropy is calculated for rounded
quantized coefficients. By using entropy number of bits required for compressed image is calculated.
By using Equation 4, Compression ratio is calculated. Results are tabulated in Table 1.




The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
44
Table1. Results of different images





Results in Table 1, indicating that blocking artifacts are more in JPEG than the proposed method at
the same compression ratio. PSNR values in dB are 37.0625 and 38.0159 for Lena and Zelda images
for the proposed method. These values for JPEG are 36.9238 and 37.9238 dB. Almost at the same
PSNR and compression ratio, the proposed method gives better images in terms of reduced blocking
artifacts. By using WT the Score values for Mandrill and Barbara are 9.22 and 9.07 respectively. For
other image-score values are same as CT. Mandrill and Barbara contains more contours. Hence, CT is
more effective in capturing smooth contours and geometric structures in images than wavelet
transform.
Experiments further extended to other image like Boat, Goldhill, and Nature. JPEG
compression gives 5.0400, 4.8765 and 4.9357 as score values respectively. Where as the
proposed method gives better score (10.0). Due to block processing, JPEG method gives
more blocking artifacts than CT. CT and WT are applied for whole image. Hence, blocking
artifacts are not available. Reconstructed images using three methods at 17:1 CR is shown in
Figure 7. JPEG reconstructed image does not give good visual quality of image. CT and WT
give same quality images. Score values for JPEG, WT and CT are 4, 10 and 10 respectively.


JPEG WT CT
Figure 7. Reconstructed Lena images ( Compression ratio 17:1).

Experiments are conducted on LENA to test the performance of the proposed watermarking
algorithm. LENA of size 512x512 is used for experimentation and is shown in Figure 8(a). The
watermark image is a 32x32 binary logo as shown in Figure 8(b).


8(a) 8(b)
Figure 8(a) Cover Image Lena (b) Watermark Image
Image Compression
Ratio
JPEG
(Score)
Proposed
(Score)
Lena 7:1 8.63 10.0
Mandrill 9:1 5.89 10.0
Zelda 10:1 8.04 10.0
Peppers 10:1 7.68 10.0
Airplane 9:1 8.10 10.0
Barbara 9:1 6.80 10.0
The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
45
Several attacks are used to assess the performance of the proposed watermarking algorithm. Two
important performance metrics considered for the algorithm are PSNR (Peak Signal to Noise Ratio)
and Normalized Cross correlation Coefficient (NCC) [29]. PSNR of the watermarked image is 51.12
dB indicates that the quality of the watermarked image is exceptional. Almost error free extraction is
possible with the proposed algorithm. Extracted watermarks are listed in Table 2. NCC and PSNR are
indicated in Table 2. Rotation is a lossy operation. The watermarked image is rotated by 10
0
and 50
0

to the right and then rotated back to their original position using bilinear interpolation. For low pass
filtering attack, a 3x3 mask consisting of an intensity of 1/9 is used. When median filtering is applied
to watermarked image, each output pixel in the attacked image contains the median value in the 3-by-
3 neighbourhood around the corresponding pixel in the input image. Resizing operation first reduces
or increases the size of the image and then generates the original image by using an interpolation
technique. This operation is a lossy operation and hence the watermarked image also looses some
watermark information. In this experiment, first the watermarked image is reduced from 512x512 size
to 256x256. By using bicubic interpolation, its dimensions are increased to 512x512.

The watermarked image is compressed using lossy JPEG (Joint Photographic Experts Group)
compression. The index of the JPEG compression ranges from 0 (best compression) to 100 (best
quality). The watermarked image is attacked by salt & pepper noise with a noise densities of 0.001.
All the extracted watermarks are clearly visible indicating the proposed methods resilience to noise
attack.
Table2. Extracted Watermarks
Rotation
(5 degrees)
PSNR=
20.3456


0.4515
Resizing
34.7539

0.9624
LPF
32.2469


0.6035
JPEG
compression
48.0304

0.7945
Median
Filtering
36.1475


0.4329
Salt &
Pepper
noise with
density
0.001

0.5298
6.CONCLUSIONS
In this paper, a CT based image compression and watermarking scheme is presented. LP decomposed
low pass version of the image is compressed using prediction algorithm with re-ordering. Further,
these coefficients are quantized using scalar quantization. All directional sub bands are quantized
based on the energy of a particular sub band. Compression is improved by considering lossless
compression in LL band. Due to multiresolution decomposition of CT, performance of compression is
improved in terms of reduced blocking artifacts than JPEG compression. Results are also compared
with wavelet transform. CT based image compression is more effective in capturing smooth contours
(due to anisotropy property) than WT based compression. A CT based robust watermarking algorithm
is also proposed in this work. By using scalar quantization, the watermark image is embedded in the
low pass sub band of the CT decomposition. The extracted watermark is visible even after some
common image attacks. Hence, it can be used for image copy right protection.
Acknowledgment
The first author would like to thank the management of QIS College of Engineering and Technology,
Ongole, A.P, India, for their support to carry this work.

The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
46
References
[1] Z.Fan and R.D.Queiroz. Maximum likelihood estimation of JPEG quantization table in the identification of
Bitmap Compression. IEEE, 948-951,2000.
[2] V.Ratnakar and M.Livny. RD-OPT: An efficient algorithm for optimizing DCT quantization tables. Data
Compression Conference, IEEE, 332 - 341. March, 1995
[3] D.M.Monro and B.G.Sherlock. Optimum DCT quantization. Data Compression Conference, IEEE, Dec,
1993.
[4] D.M.Bethel,D.M.Monro and B.G.Sherlock. Optimal quantization of the discrete cosine transform for
image compression. , Conference Publication No443, IEE ,69-72, 1997.
[5] L.W.Chang,CY Wang and SM Lee Designing JPEG quantization tables based on Human Visual System.
Signal Processing: Image Communication, Volume 16, Issue 5,501-506, Jan 2001.
[6] Sonja G,Mislav G and Branka Z. Performance analysis of image compression using wavelets. IEEE
transactions on Industrial Electronics. 48 (3), pp.682-695,2001.
[7] Averbuch A and Zheludev VA. A new family of spiline-based biorthogonal wavelet transforms and their
application to image compression. IEEE transactions on image processing, 13(7), pp 993-1007, 2004.
[8] A.Skodras, Christopoulos and T Ebrahimi. The JPEG 2000 still image compression standard. IEEE Signal
processing magazine, pp 36-58, 2001.
[9] Weinberger, Seroussi and Saprio. The LOCO-I lossless image compression algorithm:principles and
standardization into JPEG-LS. IEEE transactions on Image Processing, 9 (8),pp 1309-1324,2000.
[10] G.Manzini. An analysis of the burrows-wheeler transform. Journal of the ACM,48 (3),pp 407-430,2001.
[11] Minh N.Do. Directional multiresolution image representation. Ph.D dissertation, Swiss Federal Institute of
Technology,Lausanne, Switzerland.
[12] Minh N.Do and M. Vettterli. Pyramidal directional filter banks and curvelets. IEEE international
conference on image processing. Greece, Oct 2001.
[13] M.Vetterli. Multidimensional suband coding:some theory and algorithms. Signal Processing, 6(2), pp 97-
112,1984.
[14] Minh N.Do and Vetterli M.The contourlet transform:An efficient multiresolution image representation.
IEEE transactions on Image Processing, Vol.14, No.12, December, pp. 2091-2106,2005 .
[15] Civarella and Moffat. Lossless image compression using pixel reordering. Proceedings of twenty seventh
Australian Computer Science conference,pp 125-132,2004.
[16] K.Veeraswamy, S.Srinivaskumar and B.N.Chatterji. Lossless image compression using topological pixel
re-ordering. IET international conference, India, pp 218-222,2006.
[17] Memon and Shende. An analysis of some scanning techniques for lossless image coding. IEEE
Transactions on Image Processing,9 (11),pp 1837-1848,2000.
[18] Memon and Wu X. Recent developments in context based predictive techniques for lossless image
compression. The computer Journal,40,pp 127-136,1997.
[19] K.Sayood. Introduction of data compression. Acdemic press, 2
nd
edition,2000.
[20] D.Salomon. Data Compression. Springer,2
nd
edition, 2000.
[21] Wang Z, Sheikh and Bovik. No reference perceptual quality assessment of JPEG compressed images.
IEEE international conference,pp 477-480,2002.
[22] I.J.Cox, M.L.Miller, and J.A.Bloom, Digital Watermarking, Morgan Kaufmann Publishers, 2002.
[23] RR.Chandramouli, G.Benjamin, and R.Collin A multiple description framework for oblivious
watermarking, in Proceedings of Security, Watermarking and Multimedia, Vol. 4314, No. 4314, 2001,
pp.585-593.
[24] P.Premaratne, A novel watermark embedding and detection scheme for images in DFT domain,
Proceedings of IEE 7
th
International Conference on Image Processing & Applications, Vol.2, 1999,pp.780-
783
[25] .D.Kundur and D.Hatzinakos, Digital watermarking using multiresolution wavelet decomposition,
Proceedings of IEEE International conference on Acoustics, Speech and Signal Processing, Seattle,
Washington, Vol.5, 1998,pp.2969-2972.
[26] T.S.Ho.Anthony, J.Shen, S.H.Tan, and A.C.Kot, Digital image-in-image watermarking for copyright
protection of satellite images using the fast hadamard transform, IEEE International Geoscience and
Remote Sensing Symposium, Vol.6, 25-28 June 2002, pp.3311-3313.
[27] M.Jayalakshmi, S.N.Merchant, and U.B.Desai, Digital watermarking in contourlet domain, 18
th

International conference on Pattern Recognition, Vol.3, 2006, pp.861-864.
[28] B.C.Mohan and S.S.Kumar, A robust image watermarking scheme using singular value decomposition,
Journal of Multimedia, Vol.3, No.1, 2008, pp.7-15.
[29] J.S.Pan, H.C.Huang, and L.C.Jain, Intelligent Watermarking Techniques, World Scientific publishers,
2004, pp.7-8.

The InleinalionaI }ouinaI of Nexl Ceneialion Nelvoik (I}NCN), VoI.2, No.1, Maich 2O1O
47
Authors:
Dr.K.Veeraswamy is currently working as Professor & HOD in ECE
Department, QIS College of Engineering & Technology, Ongole, India.
He received his M.Tech and Ph.D. from JNTU Kakinada, India. He has
twelve years experience of teaching undergraduate students and post
graduate students. His research interests are in the areas of image
compression, image watermarking, and networking protocols.

Dr.B.Chandra Mohan is currently working as Professor & HOD in ECE
Department, Bapatla Engineering College, Bapatla, India. He received his
Ph.D from JNTU,Kakinada, India. He received his M.Tech from Cochin
University of Science & Technolgoy, Cochin, India. He has twenty years
experience of teaching undergraduate students and post graduate students.
His research interests are in the areas of image watermarking, and image
compression.

Y.V.Bhaskar reddy is presently working as an associate professor in ECE
Department, QIS College of Engineering & Technology, Ongole, India.
He received his B.Tech from S.V.University, Tirupathi, India and M.Tech
from Bharath University, Chennai, India. He has eight years of experience
of teaching undergraduate students and post graduate students. His
research interests are in the areas of image processing and frequency
tunable antennas.

Dr. S. Srinivas Kumar is currently Professor in ECE Department and
Director Sponsored Research, JNT University Kakinada, India. He
received his M.Tech. from JNTU, Hyderabad, India. He received his
Ph.D. from E&ECE Department IIT, Kharagpur. He has twenty three
years experience of teaching undergraduate and post-graduate students
and guided number of post-graduate and Ph.D thesis. He has published 20
research papers in National and International journals. Presently he is
guiding five Ph.D students. His research interests are in the areas of digital
image processing, computer vision, and application of artificial neural
networks and fuzzy logic to engineering problems.