International Journal of Pure and Applied Mathematics

Volume 116 No. 16 2017, 471-477

A STUDY ON THE GEOMETRIC CORRECTION USING SATELLITE IMAGES

1
E.Venkatesan, 2S.Selvaragini
1,2
Asst.Professor, Department of MCA
BIST, BIHER ,Bharath University, Chennai-73
1
venkatelumalai12@yahoo.co.in, 2slecturer1989@gmail.com

Abstract: The geometric correction processing remote step is Rectification. This means alignment of image to
sensing images. In the paper available for different kind a map so that the image is plan metric, just like the
of satellite images, application and variety of geometric map. Also known as geo - referencing. The next level is
correction techniques focus. This survey focuses the Geocoding. A special case of rectification that includes
source of geometric correction and compares different scaling to a uniform standard pixel GIS. The use of
kind of models and geometric process. This survey standard pixel sizes and coordinates permits convenient
analysis many error correction methods and processing layering of images from different sensors and maps into
function, algorithms, models and processing steps. The a GIS. The final level is Ortho-rectification. This level
paper reviews the satellite image based on geometric focuses correction of the image, pixel by pixel for
correction technique processing and geometric topographic distortion. The preprocessing of satellite
correction best model analysis. The geometric images prior to image classification and change
correction apply filed ground control point detection is essential. Preprocessing commonly
measurement and map update models, medical image comprises a series of sequential operations, including
measurement technique, neurology surgery systems, atmospheric correction or normalization, image
environment and disaster's monitoring and land registration, geometric correction, and masking (e.g.,
management, evaluation of forest regeneration and for clouds, water, irrelevant features). The
vegetation recovery. preprocessing of satellite images data often will include
radiometric correction and geometric correction. The
Keywords: Satellite Imagery, Ground Control Points, radiometric corrections are made to the raw digital
and Image registration. image data to correct for the brightness values of the
object on the ground, that have been distorted because
1. Introduction of sensor calibration or sensor malfunction problems.
The distortion of images is caused by the scattering of
The satellite images are originally used in military and
reflected electromagnetic light energy due to a
environmental filed. But nowadays it is used more and
constantly changing atmosphere. This is one source of
more in the field of map production, agriculture, forest,
sensor calibration error. The geometric corrections are
planning national land, and establishment of city plan.
made to correct the inaccuracy between the location
This satellite images are used today world in many
coordinates of the picture elements in the image data,
applications, monitoring in satellite images is day to
and the actual location coordinates on the ground.
day processing because of weather report climate
Several types of geometric corrections include system,
changing, natural diastral, and the Security of nation.
precision, and terrain corrections [2].
The satellite images raw data are not directly used in
the geographical information system, (GIS) some error
correction technique is included in a radiometric
2. Geometric correction technique
correction and Geometric correction. The radiometric
corrections are made to the raw digital image data to The geometric correction technique is one of the image
correct the brightness values. In systematic the error correction technique and images quality process.
geometric correction present remote sense images to The geometric correction technique is in different kiddy
previous image to compare and measure ground of models, algorithms and new models used. Today
controls points (GCP), in non-systematic present image world remote sense data is very important. In day to
to old map to compare and measure ground control day life a remote sense data is used to human. [3].
points, other one is global position system (GPS) is RamaKrishnan R et al. carried out a
used image to GPS is measure ground control point.[3]. comparative analysis of various methods for
This section discusses about different levels of preprocessing of satellite imagery in their research
geometric correction of remotely sensed imagery. The work [4]. A comparative study of the various
first level is Registration. In this stage the alignment of mathematical techniques has been discussed in this
one image to another image of the same area. The next paper for the betterment of preprocessing of satellite

471
International Journal of Pure and Applied Mathematics Special Issue

images. They discusses about images obtained by requires corrections for comparisons between them. In
remote sensing systems are not quite often sufficient this paper we show how highly deformed radar images
for high precision applications due to various can be geometrically corrected and compared to map
distortions. The distortions can be divided into three data coming from digital terrain models and also with
layers Geometric, Radiometric and Atmospheric data coming from SPOT satellite. Radar images that we
distortions/Errors. use are from the sensor airborne radar Varan, which is
Goncalves.H.et al. This research analysis the used for data acquisition campaign in the South East of
geometric correction process is a crucial step in remote France. Applications include structural geology, land
sensing applications. This process is frequently cover or study of coastline. We propose a solution to
manually performed, which is a laborious task in many rectify radar image in the geometry of a numerical
situations as automatic image registration methods are terrain model. The method adopted here, is to produce a
still far from being broadly applied. One of the reasons synthesis radar image by encoding all flight parameters
that justify the absence of a broad application of of aircraft or satellite from a digital terrain model; radar
automatic image registration methods is the lack of image can then be compared to the synthetic image
measures for an objective and automated analysis of the because points of landmarks can be clearly identified.
image registration process quality. The root mean Finally, we obtain a correspondence between the points
square (RMS) of the residuals is the only quantitative of real radar image distorted, and those in the land or
evaluation which is generally used in this process, with map. From this method, it is possible to simulate radar
the final validation of the geometric correction process image of aircraft or satellite missions as ERS1 or ERS2
being a qualitative analysis. Therefore, in both and propose some geometrical corrections. However,
ldquohumanrdquo and automatic image registration Satellite images acquired on areas of complex
processes, an objective evaluation of its quality is landscapes are of course much more distorted than
required. In this letter, we propose several measures for those acquired from airborne radar. Finally, we can
an objective evaluation of the geometric correction consider other comparisons: we then apply the
process, as a complement to the traditional RMS of the superposition of radar data and SPOT images of
residuals and visual inspection. Two scenarios of terrains with or without landforms. [7].
control point distribution and the most common Jianghao Wang et al. this research focus about
residual distributions were considered. With the acquirement of ground control points (GCPs) is a basic
proposed measures, we intend to cover the most and important step in the geometric correction of
common qualitative analysis aspects. This has remotely sensed imagery. In particular, the spatial
particular importance under the scope of automatic distribution of GCPs may affect the accuracy and
image registration methods, where an automatic quality of image correction. In this paper, both a
evaluation of the results is also required [1]. simulation experiment and actual-image analyses are
A.L. Choo et al. This research about geometric carried out to investigate the effect of the sampling
correction is an important step in the Synthetic design for selecting GCPs on the geometric correction
Aperture Radar (SAR) image processing. A SAR image of remotely sensed imagery. Sampling designs
has to be geocoded before it is usable for further compared are simple random sampling, spatial
processing or information extraction. There are three coverage sampling, and universal kriging model based
major processes involved in a typical geometric sampling, and the experiments indicating that the
correction: error modeling, ground control point sampling design of GCPs strongly affects the accuracy
selection and image transformation. In this research of the geometric correction. The universal kriging
analysis on four geometric correction methods namely model based sampling design considers the spatial auto
global polynomial model, 2D direct linear covariance of regression residuals and yields the most
transformation model, hybrid model and genetic accurate correction. This method is highly
algorithm with global polynomial model will be recommended as new GCPs sampling design method
presented. By choosing appropriate ground control for geometric correction of remotely sensed imagery.
points, these geometric correction methods are applied [5].
onto an actual UAV SAR data for performance Xiaojun Shan et al. This research is automatic
evaluation using root mean square error technique. geometric precision correction system (GPCS) based
[19]. on the automatic registration between HJ-1 images and
Philippe Durand et al. We propose in this paper Landsat Thematic Mapper images. First, the coarse
an original method to correct the geometric distortions image matching method based on geometric-restricted
of a radar image. The comparison of satellite data scale-invariant feature transform (SIFT) is used to
reveals a specific problem. Data can be more or less determine the coarse global transformation between the
noisy, but especially the geometry of their acquisition HJ-1 image and the reference image. Second, inspired

472
International Journal of Pure and Applied Mathematics Special Issue

by the hierarchical method of non-rigid registration for relief are two important problems in geometric
medical images, a hierarchical image matching correction, especially in the case of imaging with large
approach is proposed based on the combination of SIFT view angles. General polynomial correction model is
feature points and template matching. This approach only effective to flat area and ineffective to correct
decomposes a matching problem of a whole image into projection errors caused by both curvature of the earth
numerous matching problems of image blocks and can and relief. According to the generated characteristic of
overcome the impact of local distortions in HJ-1 the projection errors, this research analysis proposes a
images. Hierarchical random sample consensus self - adjective polynomial model which adds an
(RANSAC) based on digital elevation model (H- adjective factor in the large view angle direction to
RANSAC) is used to remove incorrect control points. make effective correction on distorted images with both
Third, an HJ-1 image is rectified using a triangulated kinds of projection errors in the absence of precise
irregular network. Finally, the automatic evaluation attitude parameters. This paper first analyzes the
method based on automatic image matching between principle of projection errors caused by curvature of the
the corrected HJ-1 image and the reference image is earth and relief, and then proposes the improved
adopted to evaluate the geometric precision. On the one polynomial model. Experiments show that the proposed
hand, experiments on eight HJ-1 images demonstrate model has greatly improved accuracy for correcting the
the efficiency and accuracy of the different steps of projection errors in the large angle direction compared
GPCS. On the other hand, experiments on 1000 HJ-1 with general polynomial models. [18].
images also demonstrated the robustness, accuracy, and Jaehong Oh et al. this research disuses about
suitability for batch processing [13]. As the need for efficient methods to accurately update
This research focus K.S.S. Sekhar et al. the and refine geospatial satellite image databases is
purpose of this paper is two-fold. First, the use of the increasing, we have proposed the use of 3-dimensional
rational polynomial coefficients (RPCs) model is digital maps for the fully-automated RPCs bias
analysis for geocoding of Medium Resolution Scan compensation of high resolution satellite imagery. The
(MRS) ground range (GR) images from the RISAT-1 basic idea is that the map features are scaled and
SAR mission. As the GR images are obtained after aligned to the image features, except for the local shift,
many preprocessing image corrections for topographic through the RPCs-based image projection, and then the
effect, range cell migration, etc., the number of ground shifts are automatically determined over the entire
control points (GCPs) required for ortho-rectification to image space by template-based edge matching of the
meet desired geometric quality needs to be established. heterogeneous data set. This enables modeling of RPCs
This assumes importance due to difficulty in visual bias compensation parameters for accurate geo-
identification of the GCPs in 18 m-resolution MRS referencing. The map features are selected based on
SAR images. Second, three possible methods of bias- four suggested rules. Experiments were carried out for
compensated RPC models are studied for geocoding. three Kompsat-2 images and stereo IKONOS images
These cases are (A) modified RPC with shift bias, (B) with 1:5000 scale Korean national topographic maps.
regenerated RPC with shift bias, and (C) regenerated Image matching performance is discussed with
RPC with affine transform model. Experiments are justification of the parameter selection, and the geo-
carried out with a set of eight scenes acquired over referencing accuracy is analyzed. The experimental
planar regions especially to avoid the impact of SAR- results showed the automated approach can achieve
specific geometric effects such as foreshortening and one-pixel level of geo-referencing accuracy, enabling
layover. Geometric accuracy of the ortho images economical hybrid map creation as well as large scale
obtained from these cases is verified at GCPs used for map updates. [15].
processing as control points and at new GCPs used as Ali E. Said et al. this research analysis High
check points. It is observed that the modified RPC with resolution satellite images (HRSI) such as Quick Bird
the shift bias case required more GCPs to meet the supply a lot of opportunities in mapping, GIS and many
desired geo positioning accuracy. Even though both the other applications. HRSI have to be geometrically and
regenerated RPC models have shown near similar precisely processed with GCPs to get the accuracy
performance, the regenerated RPC with shift bias needed for map updating. In this study, North-West
compensation is found to reach the required geo part of LIBYA (in Tripoli Region) test field which has
positioning accuracy with least number of the GCPs, a flat topography is chosen as a test area. The test study
suggesting it as a strong candidate for realizing area was covered with Ortho rectified Quick Bird
operational high precision RISAT-1 geocoded products image 60 cm spatial resolution, large scale topography
for multi-temporal data analysis. [6]. map produced from stereo aerial photo have been used
This research analysis about Ye Zhang et al. as a reference data, ground control points (GCPs) and
projection errors caused by curvature of the earth and check points (CPs) was extracted. Geometric correction

473
International Journal of Pure and Applied Mathematics Special Issue

of the image will be done based on mathematic model routinely used in diffusion and functional MR imaging
(2D-polynomial) approach. From the results it can be due to its rapid acquisition research time.[28] However,
say that the high resolution satellite images like Quick the long readout period makes it prone to susceptibility
Bird images are good source for updating large scale arte facts which results in geometric and intensity
maps (1:5000). [12]. distortions of the acquired image. The use of these
Hui Deng et al. this discusses about in recent distorted images for neuron navigation hampers the
years, data collected from remote sensing satellite and effectiveness of image-guided surgery systems as
aero photography has been showing a geometric critical white matter tracts and functionally eloquent
sequence increase. A method of Scale Invariant Feature brain areas cannot be accurately localized. In this
Transform (SIFT) algorithm could be employed for the paper, we present a novel method for correction of
automatic geometric fine correction.[23] This method distortions arising from susceptibility arte facts in EPI
could avoid the impact of the rotation and zooming of images.[29] The proposed method combines field map
template matching during the image matching process, and image registration based correction techniques in a
and it can also save the labor during the image unified framework. A phase unwrapping algorithm is
processing operation.[27] Based on the SIFT algorithm, presented that can efficiently compute the B0 magnetic
this research analysis a two-step method,[24] which field inhomogeneity map as well as the uncertainty
firstly conducts coarse match on feature points, and associated with the estimated solution through the use
then further conducts fine correction on the coarsely of dynamic graph cuts. This information is fed to a
matched feature points by using the least squares subsequent image registration step to further refine the
technique.[25] The result indicates that, this method is results in areas with high uncertainty.[30] This work
an effective automatic matching method for remote has been integrated into the surgical workflow at the
sensing images. [8]. National Hospital for Neurology and Neurosurgery and
P.V. Radhadevi et al. this research analysis its effectiveness in correcting for geometric distortions
describes the development of an algorithm for due to susceptibility arte facts is demonstrated on EPI
geometric correction of Terrain Mapping Camera images acquired with an interventional MRI scanner
(TMC) imagery of Chandrayaan-1 (CH-1). The during neurosurgery. [17]. Wei Zhang et al. this discus
correction is based on a rigorous sensor model. The about Small satellite constellation of environment and
algorithm ;[24]incorporates the camera geometry disaster's monitoring and predicting (shorted for HJ-1)
model, satellite data and lunar control points in a is not a mapping satellite, and its parameters of attitude
rigorous bundle adjustment updating the satellite model and orbit cannot satisfy the requirement of geometric
parameters. The model is tested for different strips over correction using strict imaging model. On the other
the polar and equatorial regions and the results are hand, due to the 12000 CCD detectors and large
presented. The plan metric control is identified from the overlay of multispectral payload named CCD carried
100 m/pixel USGS Clementine base map mosaic and by HJ-1 satellite, the error caused by CCD distortion
vertical control is derived from Lunar Orbiter Laser cannot be ignored. Aiming at these problems of HJ-1,
Altimeter (LOLA) data. RMS error of the order 200– this paper proposes a strict orbit model algorithm based
300 m in latitude, longitude and height with respect to on Ground Control Point (GCP) and collinear condition
the references could be achieved using the sensor equations. Through the robust estimation of parameters,
model with few distributed controls over the strip. this algorithm can effectively set up imaging geometric
[22]The model is used for the orientation of long strips model of CCD, and satisfy the requirement of high
with little or no degradation in orientation quality precision geometric correction.[16].
attainable with a short scene. The results are This research work done by George H et al. the
representative of the stability of the platform and analysis Careful evaluation of forest regeneration and
potential of CH-1 for accurate lunar mapping. The vegetation recovery after a fire event provides vital
algorithm for geometric correction described in this information useful in land management.[35]The use of
paper is a part of Lunar Mapping System (LMS), which remotely sensed data is considered to be especially
will handle full-pass data for operational generation of suitable for monitoring ecosystem dynamics after
Digital Elevation Models (DEM) and Ortho products fire.[36] The aim of this work was to map post-fire
from TMC images of CH-1. [11]. forest regeneration and vegetation recovery on the
Mediterranean island of Thasos by using a combination
3. Geometric correction application of very high spatial (VHS) resolution (Quick Bird) and
hyper spectral (EO-1 Hyperion) imagery and by
The geometric correction is different kind of employing object-based image analysis. More
application in the section discuses. [14]. Pankaj Daga et specifically, the work focused on first the separation
al. this discuses about Echo Planar Imaging (EPI) is and mapping of three major post-fire classes (forest

474
International Journal of Pure and Applied Mathematics Special Issue

regeneration,[31] other vegetation recovery, unburned Various Methods for Preprocessing of Satellite
vegetation) existing within the fire perimeter, and Imagery", International Journal of Engineering Science
second the differentiation and mapping of the two main & Technology, Vol. 5(2), pp. 431-437, 2013.
forest regeneration classes, namely, Pinusbrutia [5] Xiaojun Shan, Ping Tang,Changmiao Hu., ”An
regeneration and Pinusnigra regeneration. The data automatic geometric precision correction system based
used in this study consisted of satellite images and field on hierarchical Registration for HJ-1 A/B CCD
observations of homogeneous regenerated and images”, A International Journal of Remote Sensing
vegetated areas.[32] The methodology followed two ,pp.7154-7178 Vol 35(20), pp.7154-7178, 2014.
main steps three level image segmentation, and, a [6] Chunyuan Wang and Ye Zhang,PigangLiu,Qi
classification of the segmented images.[33] The process Xu., “A self adjustive geometric correction method for
resulted in the separation of classes related to the seriously oblique Aero image Geoscience and Remote
aforementioned objectives.[34] The overall accuracy Sensing Symposium”, IEEE, pp.1433-143624-29,
assessment revealed very promising results 2011.
(approximately 83.7% overall accuracy, with a Kappa [7] Jianghao Wang Yong Ge ,Gerard B.M.
Index of Agreement of 0.79). The achieved accuracy Heuvelinkc and Chenghu Zhou,Dick Brus” Effect of
was 8% higher when compared to the results reported the sampling design of Ground control points on the
in a previous work in which only the EO-1 Hyperion geometric correction of remotely sensed imagery”
image was employed in order to map the same classes. International Journal of Applied Earth Observation and
Some classification confusions involving the classes Geoinformation Vol, 18, pp. 91–100,2012.
of P.brutia regeneration and P.nigra regeneration were [8] Radhadevi.P.V, Solanki.S.S, Nagasubramanian,V
observed. This could be attributed to the absence of ,SudheerReddy, T,KrishnaSumanth,J,Saibaba and
large and dense homogeneous areas of regenerated pine Geeta Varma “An algorithm for geometric correction of
trees in the analysis area. full passes TMC imagery of Chandrayaan-1” Planetary
and Space Science Elsevier, Vol 79-80, pp.45-51, 2013,
4. Conclusion [9] Yifu Chen , Zhong Xie., “The Trajectory Model
for ZY-3” The scientific World Journal PP 9. Vol 2014.
This paper analysis many type of geometric correction [10] Goncalves.H Goncalves, J.A.Corte and Real, L.”
technique focus. The satellite images using different Measures for an Objective Evaluation of the Geometric
geometric correction of models, algorithms and Correction Process Quality Geoscience and Remote
applications. This survey analysis many research papers Sensing Letters”IEEE,Vol.6(2) pp.292 – 296, 2002.
in geometric correction apply medical images and [11] Bagheri, H Sadeghian, S. ” Geometric
forest monitoring, models image-guided surgery rectification of high resolution satellite images using
systems. This survey analysis which one model is best, mathematical intelligent & classical Modeling” IEEE,
the new research model trajectory model is compare PP.1- 6,2014..
with the other models, polynomial model (LPM), [12] Hui Deng,,Limin Wang,,Jia Liu, Dandan
piecewise polynomial model (PPM), and Lagrange Li, Zhongxin Chen, Qingbo Zhou” Study on
cubic polynomial model (LCPM) any other models Application of Invariant Feature Transform Algorithm
compare trajectory model is best and valued model. on Automated Geometric Correction of Remote
Sensing Images” Vol.393, pp.352-358,2013.
References [13] Sekhar.K.S.S, Senthil Kumar and Dadhwal.V.K,
“Geocoding RISAT-1 MRS images using bias-
[1] A.L. Choo, Y.K. Chan, V.C. Koo, T.S. Lim” compensated RPC models” International Journal of
Study on Geometric Correction Algorithms for SAR Remote SensingVol.35(20) 2014.
Images “International journal of microwave and optical [14] pankajdaga , tejaspendse, marcmodat, markwhite,
technology, Vol.9(1), 2014. lauramancini, gavin,p.Wnston,andrew,w.mcevoy, john
[2] Hiroshi Higuchi, Masayuki Saito and Tsutomu thorntn, tarek yousry, ivana drobnjak, john
Iwahashi, Sumio Usui “Network based high accuracy s.duncan, sebastien ourselin” Susceptibility arte fact
real time GPS Positioning for GCP correction of high correction using dynamic graph cuts Application to
resolution satellite imagery”, IEEE, 2004. neurosurgery” Elsevier, Vol 18(7),pp.1132-1142, 2014.
[3] Santhosh Baboo.S, and Renuka Devi.M” [15] Ali E. Said, Hadi, M.Omar Shandoul, YekHlef,
Geometric Correction in Recent High Resolution Z.Yekhlef “Up Dating Large Scale Maps using
Satellite Imagery” International journal of computer Resolution Satellite Image” 4th International
applications, Vol. 14(1), pp. 0975 – 8887 2011. Conference on Environmental Science and
[4] RamaKrishnan, R., Manthira Moorthi.S, Development-ICESD , pp.435-440 2013 Elsevier.
Prabu.S, and P. Swarnalatha. "Comparative Analysis of

475
International Journal of Pure and Applied Mathematics Special Issue

[16] George H. Mitri, Ioannis Z. Gitas”Mapping., online voting system over network, Indian Journal of
“post-fire forest regeneration and vegetation recovery Science and Technology, v-6, i-SUPPL.6, pp-4831-
using a combination of very high spatial resolution and 4836, 2013.
hyper spectral satellite imagery” International Journal [29] Thooyamani K.P., Khanaa V., Udayakumar R.,
of Applied Earth Observation and Geo information Efficiently measuring denial of service attacks using
Vol. 20, pp.60–66 2013. appropriate metrics, Middle - East Journal of Scientific
[17] Wei Zhang,Wei Wu,Yan Cui,Wei Wang” HJ-1- Research, v-20, i-12, pp-2464-2470, 2014.
A/B optical satellite image geometric correction” IOP [30] R.Kalaiprasath, R.Elankavi, Dr.R.Udayakumar,
Conference Series Earth and Environmental Science, Cloud Information Accountability (Cia) Framework
2014. Ensuring Accountability Of Data In Cloud And
[18] Jaehong Oh , Changno Lee, “Automated bias- Security In End To End Process In Cloud Terminology,
compensation of rational polynomial coefficients of International Journal Of Civil Engineering And
high resolution satellite imagery based on topographic Technology (Ijciet) Volume 8, Issue 4, Pp. 376–385,
maps” ISPRS Journal of photogrammetry and Remote April 2017.
Sensing, 2014. [31] R.Elankavi, R.Kalaiprasath, Dr.R.Udayakumar,
[19] Philippe Durand , Luan Jaupi, Dariush A fast clustering algorithm for high-dimensional data,
Ghorbanzadeh,” Geometric Correction of Airborne International Journal Of Civil Engineering And
Radar SAR Image on a Digital Technology (Ijciet), Volume 8, Issue 5, Pp. 1220–1227,
[20] Udayakumar R., Kaliyamurthie K.P., Khanaa, May 2017.
Thooyamani K.P., Data mining a boon: Predictive [32] R. Kalaiprasath, R. Elankavi and Dr. R.
system for university topper women in academia, Udayakumar. Cloud. Security and Compliance - A
World Applied Sciences Journal, v-29, i-14, pp-86-90, Semantic Approach in End to End Security,
2014. International Journal Of Mechanical Engineering And
[21] Kaliyamurthie K.P., Parameswari D., Technology (Ijmet), Volume 8, Issue 5, pp-987-994,
Udayakumar R., QOS aware privacy preserving May 2017.
location monitoring in wireless sensor network, Indian [33] Thooyamani K.P., Khanaa V., Udayakumar R.,
Journal of Science and Technology, v-6, i-SUPPL5, pp- Virtual instrumentation based process of agriculture by
4648-4652, 2013. automation, Middle - East Journal of Scientific
[22] Brintha Rajakumari S., Nalini C., An efficient Research, v-20, i-12, pp-2604-2612, 2014.
cost model for data storage with horizontal layout in the [34] Udayakumar R., Thooyamani K.P., Khanaa,
cloud, Indian Journal of Science and Technology, v-7, Random projection based data perturbation using
i-, pp-45-46, 2014. geometric transformation, World Applied Sciences
[23] Brintha Rajakumari S., Nalini C., An efficient Journal, v-29, i-14, pp-19-24, 2014.
data mining dataset preparation using aggregation in [35] Thooyamani K.P., Khanaa V., Udayakumar R.,
relational database, Indian Journal of Science and Virtual instrumentation based process of agriculture by
Technology, v-7, i-, pp-44-46, 2014. automation, Middle - East Journal of Scientific
[24] Khanna V., Mohanta K., Saravanan T., Recovery Research, v-20, i-12, pp-2604-2612, 2014.
of link quality degradation in wireless mesh networks, [36] Udayakumar R., Thooyamani K.P., Khanaa,
Indian Journal of Science and Technology, v-6, i- Deploying site-to-site VPN connectivity: MPLS Vs
SUPPL.6, pp-4837-4843, 2013. IPSec, World Applied Sciences Journal, v-29, i-14, pp-
[25] Khanaa V., Thooyamani K.P., Udayakumar R., A 6-10, 2014.
secure and efficient authentication system for
distributed wireless sensor network, World Applied
Sciences Journal, v-29, i-14, pp-304-308, 2014.
[26] Udayakumar R., Khanaa V., Saravanan T.,
Saritha G., Retinal image analysis using curvelet
transform and multistructure elements morphology by
reconstruction, Middle - East Journal of Scientific
Research, v-16, i-12, pp-1781-1785, 2013.
[27] Khanaa V., Mohanta K., Saravanan. T.,
Performance analysis of FTTH using GEPON in direct
and external modulation, Indian Journal of Science and
Technology, v-6, i-SUPPL.6, pp-4848-4852, 2013.
[28] Kaliyamurthie K.P., Udayakumar R.,
Parameswari D., Mugunthan S.N., Highly secured

476
477
478