IJITCE Jan

INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY & CREATIVE ENGINEERING (ISSN:2045-8711)
VOL.1 NO.1 JANUARY 2011
IJITCE PUBLICATION
INTERNATIONAL JOURNAL OF
INNOVATIVE TECHNOLOGY & CREATIVE
ENGINEERING
Vol.1 No.1
Editorial Team
January 2011
Contents
1. Measurement of Potassium Levels in the Soil using Embedded System based
Soil Analyzer ……….[1]
2. Crowd Safety: A Real Time System For Counting People ……….[6]
3. A Novel design of Electronic Voting System Using Fingerprint…..[12]
4. GPS Tracking Simulation By Path Replaying…………………………[20]
5. Object Oriented Design of E-learning System for Distance Education…[27]
6. On The Relative Character Graph of A Finite Group………………..[31]
7. Detection of Diabetic Retinopathy Using Radial Basis Function….[40]
8. Peristaltic flow of a Williamson fluid in an asymmetric channel through porous

medium………………………………………………………[48]
Measurement of Potassium Levels in the

Soil using Embedded System based Soil
Analyzer
Ms.D. Asha Devi#1, Prof.K.Malakondaiah#2,Mr.M.Suresh Babu#3
#1
Associate Professor, Department Of ECE,Intellectual Engineering College,Anantapur.
#2
Professor, Department Of Instrumentation,Sri Krishna Devaraya University, Anantapur,A.P.
#3
Principal,Intell Institute Of Science,Anantapur.
Abstract: Potassium is important to the plants for metabolic The measured data is represented in the form of web
changes during flowering, and the production of floral clusters. pages, using HTML and the processing of data using
It also promotes general plant-vigor, disease-resistance, and Dynamic C functions. The Rabbit is created as
study growth. Hence, in this paper the concentration of
potassium ion in aqueous soil samples of Anantapur area is
analyzed. Combination of ISE and reference electrode is used web server and is connected to PC by using Ethernet
to determine the ion activities of potassium ion in aqueous soil cross cable (RJ45). Then the Rabbit processor and the
samples. This method is very simple and fast when compared PC are in LAN (Local Area Network). Here data is
with the other methods. The soil samples tested with the accessed through Rabbit processor, webpage can be
embedded system based soil analyzer. viewed on PC, and hence, we can access physical data
through LAN.
Keywords: ESBSA (Embedded System Based Soil
Analyzer); ISEs (Ion Selective Electrodes);
Introduction: The main objective of this research is to

develop “EMBEDDED BASED SOIL ANALYZER” which
is used to analyze the Potassium levels present in the
soil. As Agriculture is one of the most important
occupations in INDIA, it is very much essential to know
the nutrients present in the soil for a suitable crop.
However, in every district only one or two organizations
are there for the testing of soil. To increase this facility,
adding today’s technology towards agricultural fields, a
cost effective soil-testing instrument is developed. In this
paper the measurement of Potassium levels present in
the soil is demonstrated with the help of ESBSA.
Figure.1 consists of two sensors (i.e.

temperature, and potassium) which are connected to Fig.1. Measurement Of Potassium using
ADC MCP3208. The main function of analog to digital Embedded system based Soil Analyzer
converter is to convert analog values into digital values.
The output of ADC is given to Rabbit Processor (Rabbit
3000). The main reason for using Rabbit Processor is This experiment is to develop and implement an
for its excellent feature of inbuilt TCP/IP. embedded system based soil analyzer, which is an
internet based measurement system for analysis of
1
different parameters. In this paper the estimation of After removing any stones or fresh organic material
Potassium present in the soil sample is explained. The (roots, twigs, leaves, worms, insects etc), and breaking
ESBSA is connected to local computer via Ethernet port up any large lumps, soil samples must be air dried by
using RJ45 Ethernet cable. Then internet is provided to laying out in a thin layer on metal or plastic trays in a
the local PC, Apache Web Server Software is used to current of air at no more than 30°C until dry. Then they
configure the PC and made the local PC as Web Server. must be crushed in a pestle and mortar to pass through
Using this software proxy pass and proxy reverse a 2mm sieve. About 200g of material should be
operations are provided to have online communication. sufficient for duplicate analysis and storage.
Now the local system is ready to communicate the
information online. Therefore, with the ESBSA,
measured and calibrated data can be accessed globally. Weigh accurately about 50g of dry soil
sample and add exactly 100ml of de-ionized water and
shake vigorously for 30 seconds to ensure good
dispersion, then leave to stand for 15 mins. After this
time, shake again for 5 seconds, and allow to stand
again. Repeat this procedure three times before finally
allowing to settle. When the solution is clear, take
exactly 50 mls (by decanting or pipetting) and mix with 1
ml of buffer solution in a plastic beaker.
Procedure for determining the concentration of

+
POTASSIUM (K ) in SOILS
Fig:2 Local PC interfaced with ESBSA Water-soluble Potassium is extracted
from dried soil samples by dispersion in de-ionized
water and analyzed by direct potentiometer.
Potassium in Soil Sample measurement technique:
Prepare two standard concentrations C1 and C2

of the sample potassium.First, dip the
Potassium and reference electrodes in the more
dilute standard solution of concentration C1 and
note down the reading E1 mV. Then rinse both
electrodes with double distilled water blot dry
with a soft tissue paper. Now dip the electrodes
in more concentrated standard of concentration
C2, and note the stable reading E2 mV.
Now immerse the electrodes in the sample and
note down the potential reading Ex mV.
Now sample concentration Cx can be calculated
from the equation.
Cx = C2 / anti-log (∆/S).
Where ∆ = E2 – E1, S = Slope of the electrode.
The Potassium Ion-Selective Electrode has a solid-state FIG 3: Potassium Sensor electrode and reference
-3
PVC polymer matrix membrane which is designed for electrode are immersed in 10 M buffer solution
the detection of potassium ions ( K+ ) in aqueous
solutions and is suitable for both field and laboratory
applications. The Potassium Ion is a monovalent action. Calibration
One mole of Before soil sample measurement, the electrodes
( K+ ) has a mass of 39.098grams; 1000 ppm is 0.026 must be calibrated by measuring a series of known
M. standard solutions, made by serial dilution of the
1000ppm standard solution. For a full calibration,
Sample Preparation prepare 100ml of solutions containing 1000, 100, 10, 1,
and 0.1ppm K. If the approximate range of
2
concentrations of the samples is known, and this is 2. The RCM-3700 has 40 pins out of which we use 7
within the specified linear range of the ISE, then it is only pins from port-B and 6 pins from port-A for our
necessary to make two solutions which span this range: application.
e.g. if samples are known to lie between, say, 30 and
130ppm then we could use standards of 10 and 200ppm 3. The 4 pins of port-B are configured as input and
or even 20 and 150ppm.2 ml of buffer solution must be
output pins, 6 pins of port-A are configured for LCD
added to each 100ml standard and mixed thoroughly to
compensate for different activity coefficients between Display.
samples and standards.
4. In which PB-5 is used for clock, PB-7 for D-out, PB-4
Sample Measurement for D-in, PB-2 for chip select, PA-4, PA-5, PA-6, PA-7
Briefly, it is important to note that the are used for data transfer, PA-0 for R/W, PA-1 for
electrodes must be washed and dried between each Enabling the LCD.
sample, to avoid cross contamination, and sufficient time
must be allowed (2 or 3 minutes), before taking a 5. 40th pin is connected to VCC (+5v) and pins 19 and
reading after immersion, to permit the electrode signal to 20 are connected to ground.
reach a stable value. For the highest precision, frequent
recalibration is recommended (see operating
instructions).
The results will be displayed as ppm and mol/l in the
solution. Since buffer solution has been added equally to
standards and samples, these figures will not need
adjusting for this addition. However, the concentration in
the solution (in ppm = micrograms per ml) must be
multiplied by 100 and divided by the sample weight to
give the concentration in the soil (in micrograms per
gram).
Temperature Compensation:
The ISE (POTASSIUM) is a temperature
dependent one. Hence, in this research a temperature
compensated Potassium measurement system is
developed. LM35 series are precision integrated-circuit
temperature sensors, whose output voltage is linearly
proportional to the Celsius (Centigrade) temperature.
The LM35 thus has an advantage over linear
temperature sensors calibrated in ° Kelvin, as the user is
not required to subtract a large constant voltage from its Fig.4: Schematic Diagram of Embedded Soil Analyzer
output to obtain convenient Centigrade scaling. The
LM35 does not require any external calibration or
trimming to provide typical accuracies of ±1⁄4°C at room 6. LCD INTERFACE:
temperature and ±3⁄4°C over a full −55 to +150°C
temperature range. Low cost is assured by trimming and 1, 3, 5, 16th pins are connected to ground;
calibration at the wafer level. The LM35’s low output
impedance, linear output, and precise inherent 2,15th pins are connected to 5volts;
calibration make interfacing to readout or control
4th pin (RS) is connected to PA-0, 6th pin (E) is
circuitry especially easy. It can be used with single th
connected to PA-1, and 11, 12, 13, 14 pins are
power supplies, or with plus and minus supplies.
connected to PA-4, 5, 6, 7 pins for data transferring.
Hardware Description:
7. In MCP3208, ch-0,1,2,4,5 & 7(not used) Temperature
1. Here, we use rabbit-3000 processor along with RCM- sensor-ch-3, Potassium sensor – ch-5
3700 development kit.
3
8. As written in the software, when input changes the

corresponding output is changed and output display is
refreshed.
9. Here when enable button is active, it is indicated in

Ethernet cable by blinking of light.
10. If we want to end the process, enable is disabled or

connection is removed.
11.Appache Web Server software is used to configure

the local PC as a server.
12.LCD display is used for standalone results.
Results:
S.no. Soil Measurement Present Fig: 5: On line measurement of potassium sample

samples study
@ Anantapur
M/ppm Conclusion:
Soil office
By Using “EMBEDD SYSTEM BASED SOIL
1 6364 20.98 21.33 ANALYSER”, we measure Potassium levels present in
the soil at respective temperature and humidity using
2 6365 19.42 19.45 ION SELECTIVE ELECTRODES.
Potassium is absorbed by plants in larger
3 6366 45.36 45.34 amounts than any other mineral element except nitrogen
and, in some cases, calcium. It helps in the building of
protein, photosynthesis, fruit quality and reduction of
4 6367 23.38 23.14
diseases. Potassium is supplied to plants by soil
minerals, organic materials, and fertilizer.
5 6368 24.48 24.56
From the study of potassium ion content in
soils, collected from literature values and soil office
readings, soils which are having less than 150
Kg/Hectare K2O is poor in potassium deposits. When it
Table .1: Potassium levels measured in soil samples is between 150-250 Kg/Hectare K2O of soils are having
medium levels of potassium. Finally, the soils, with
greater than 250 Kg/Hectare K2O deposits, considered
to have sufficient potassium deposits.
This system is connected to RABBIT
Processor, which has special feature of built in TCP/IP,
by which we can upload the data to the Internet. This
device is more economical, reliable, and portable. Using
this instrument, farmers can access the data through
Net from remote area. By this, farmer can evaluate the
soil nutrients present in the soil and can analyze the
amount of nutrients present in the soil and lack of
percentage of nutrients to be added for a specific crop
by using predefined data provided. A farmer can have
the suggestions from soil analysts or agricultural
scientists through Internet, so that he can improve the
crop yields in an efficient manner.
4
12. References
[1]www.agric.gov.ab.ca/app21/rtw/index.jsp
[2]Advanced sensor technologies, Inc.Toll free: 1-888-wow-ASTI

(964-2784)
[3]“Computer networks” By Andrew S.Tanenbaum, prentice hall

India private limited, third edition 2001.
[5] Wireless LAN Medium Access Control (MAC) and Physical
Layer(PHY) Specification, IEEE Std. 802.11, 1997
[6] J.J.Lingane, Anal.chem./39,881 (1967)
[7] J.J.Lingane, Anal.chem./40,935 (1968)
[8] E.Pungor and K.Toth, Anal.chem.Acta, 47,291 (1969)
[9] J.W.Ross and M.S Frant, Anal.chem, 49,967(1969)
[10] G.A.Rechnitz and N.C.Kenny, Anal.Lett. 2,395(1969)
[11] G.kakabadse, Ion-selective Rev., 3,127(1982).
[12] R.A.Durst, “Ion Sensitive Electrodes”, [13] [13]

Natl.Bur.Stand.Spec.Pub1, 314, Washington D.C., (1969).
[14] Iodide ion selective electrode manual of pH products company

Pvt. Ltd., Hyderabad.
[15] A. Hameed, M. A. Goddal, Z. H. Yamini, A. H. Yahya,

“Significance of pH Measurements in photo catalytic splitting of
water using 355nm UV laser”, Journal of Molecule catalysis A:
chemical, 227,241(2005).
[16] Australian Journal of Basic and Applied Sciences, 4(5): 922-931,

2010;ISSN 1991-8178;© 2010, INSInet Publication
[17] K. Hiiro, A. Kawahara and T. Tanaka, Anal. Chim. Acta,110, 321

(1979).
[18]K. Hiiro, A. Kawahara and T. Tanaka, Nippon Kagaku Kaishi,1980,

1447.
[19]K. Hiiro, A. Kawahara and T. Tanaka, Bunseki Kagaku, 31, E33

(1982).
[20]K. Hiiro, T. Tanaka, A. Kawahara and S. Wakida, Anal. Sci., 2,145

(1986).
[21]K. Hiiro, S. Wakida, T. Tanaka, A. Kawahara and M. Yamane,

Fresenius' Z. Anal. Chem., 326, 362 (1987).
[22] Denshi Kogyo Geppo, 27, 38 (1985).
5
CROWD SAFETY: A REAL TIME SYSTEM

FOR COUNTING PEOPLE
Prabakaran V #1, Dr.Arthanariee A.M *2, Sivakumar M #3
#1
Assistant Professor, Department of Computer Science, Adharsh Vidhyalaya Arts and Science College for
Women, Erode, Tamilnadu, India
#2
Director, Bharathidasan School of Computer Applications, Erode, Tamilnadu, India
#3
Doctoral Research Scholar, Anna University, Coimbatore, Tamilnadu, India
Abstract— Man-made disasters are the outcome means and routes of entrance and exists,
of lack of awareness, lack of sensitivity towards the communication, jamming and queuing [3].
safety measures to be taken to prevent unforeseen It is clear that a stampede death never occurs
accidents. Large crowds always invite accidents if due to the presence of 1-10 people at a place. It
preventive measures are not taken with proper
planning. When the number of people waits in queue
always happens when a huge crowd gathers at
before a shopping mall on a special sales day or one place for some or other reason. Whenever we
gathering at religious functions or at a sports have experienced any stampede death to our
gallery, it is very much evident that large crowding society the crowd of people has been held
is part of normal life. This exposes a new problem responsible. When one person in a crowd starts
space of crowd management. Excessive crowding running due to any kind of rumour the entire crowd
and poor crowd management can cost loss of follows him or her without having any time for the
precious life very easily. Hence the need for second thought to be processed and this causes
developing an automated crowd management the deaths on mass level. With the tremendous
system will be highly appreciated by the society.
Crowd management involves keeping track of the
growth of population and the facts given that
crowd, the space available and balancing between having a bad system can easily jeopardize the
the crowd and space. In this paper, we propose a precious life, “Why do crowds need to be
prototype for counting the people as a part of managed?” is becoming a penny worth question
developing better crowd monitoring system. The and “What is the best way to manage a crowd?” is
system counts people and displays the result in a becoming a million dollar question. The best
user friendly interface. The system has been tested reasons for the “Why?” Includes; Big gatherings of
in different places and found working fine for people raise the odds of a dangerous occurrence
counting people. happening. Secondly, individuals within a crowd
always take for granted that others have the
responsibility. Thirdly, big crowds or gatherings of
Keywords: Crowd Monitoring, LDR & people make changes in action slower and more
Programmable Interface Controller complicated. Fourthly, big crowds or gatherings of
people make communications slower and more
I. INTRODUCTION complicated. And most importantly, big crowds of
Crowd monitoring is the process of monitoring people raise the possible number of victims [2].
and controlling large groups of people for their Even though crowd is normal in all over the
safety and security. It also provides ways to world, India often faces many problems with crowd
efficiently utilize space and reduce cost involved in particularly in religious functions. Religious is the
maintenance [1]. It includes different phases like soul of India and its culture. There are number of
planning, organizing, guiding and evaluating holy places where the devotees go as pilgrims
results of corrective actions. Crowd safety and throughout the year. Whenever there are such
security in public areas are primarily the mammoth gatherings, there should be proper
organizer’s or operator’s responsibility. A health arrangements to control the people.
and safety management system is required to Traditionally, crowd management is performed
monitor and control potential crowding risks in by employing extensive closed circuit television
public areas. The four interacting elements that system. This involves extensive cost, time and
contribute to a better and efficient system are; human effort in setting up and establishing the
time, space, information, and energy. The system. As routine monitoring is tedious, the
behaviour of the crowd is another attribute that observers are likely to lose concentration. The
plays a vital role in efficient crowd monitoring need and necessity of the automatic surveillance
system. Other attributes include facility, size,
6
for routine crowd monitoring and controlling is Examples of these types are turnstiles and mat-
again proven [6] type foot switches. The above sensors are suitable
In order to save the life of the public, a model for counting a few people and are not adequate for
has been developed to count the entry of the crowd monitoring. On the other hand, there are
people entering into an area. The system has non-contact and non-obstruct sensors. From these
been developed using both the hardware and types, is CCTV visual camera. In CCTV based
software implementations. This paper has been research efforts, individuals are first detected by
organized in the following ways. Section II human models or detectors and then tracked in
describes some of the incidents occurred due to order to count the number of pedestrians. Masoud
overcrowding and Section III discuses previous et al. [2] used difference images to detect moving
work done. The Section IV describes prototype of blobs, and pedestrians are detected by rectangular
the system and Section V discuses experimental patches based on their dynamic behavior. Liu et al.
results. Conclusion and future work is given in [3] proposed a method based on human
Section VI. appearance models to detect pedestrians. In this
method, human models are created using a set of
II. LIST OF DANGERS ENCOUNTERED low level image features, and the number of
Numerous incidents have been recorded in people is counted using these human models.
which uncontrolled crowding has resulted in Also, Sidla et al. [4] proposed a model-based
injuries and, in some instances, death. There are method to detect the number of humans in scenes.
inherent dangers associated with every large In this method, humans were successfully
public gathering. Every year there are reports of detected using active shape models (ASM) even if
overcrowding and crushing incidents from around their bodies were partially occluded. However,
the world. It is very particular to say that more these methods cannot be applied to multicamera
number of incidents is occurring in India during systems or embedded systems because they
Temple festivals. Hence there is a growing need in usually require a great amount of computational
India to develop a very useful automated crowd power.
monitoring system. To put the problem into In this paper, we propose a simple and effective
perspective, the following list highlights just some method to count the number of people by using
of the events that have ended in tragedy. Light Dependent Sensors (LDR). The sensors are
fitted in two sides of the doors, so that when light
January 15, 1999 - 51 Hindus killed and 100 beam is cut, automatically the counting process
injured in a stampede after part of a shrine starts by the program stored in programmable
collapsed. Over 1.5 million present in a ceremony Interface Controller.
in Kerala, India. IV. PROTOTYPE DESCRIPTION
The prototype of the crowd monitoring system
August 24, 2005 – 56 die, hundreds of people
includes the following main components in the
injured in a stampede at Vaishnavi Devi temple,
hardware part. The components are i) Power Unit
India
ii) Step-down Transformer iii) Light Dependent
Sensor (LDR) iv) Programmable Interface
September 30, 2008: 147 people were killed
Controller (PIC) v) LCD Display. The Software part
during the Chamunda Devi stampede at the
includes development of a system which accepts
Chamunda Devi temple in Jodhpur, India
output from the hardware unit as input and
March 4, 2010: At least 71 killed and over 200 displays the result. The working of prototype is as
injured at Ram Janki Temple in Kunda, India follows: The 230V power supply is connected to
the Transformer. The Step-Down Transformer
The above listing shows some incidents takes the voltage and converts it into 5V. The 5V
occurred in India, where there is a growing need supply is connected to Light Dependent Sensors
for better crowd monitoring system. (LDR). The sensor has two hands where 5 mA
III. RELATED W ORK light is emitted between two hands of the sensor.
In general, the detection methods are broadly The two hands are fitted in the left and right sides
classified into two categories. The first one is of the door. Fig. 1 a) shows an entry gate where
obstructive and the second is non-obstructive [6]. hands of the sensor are fixed. Fig. 1 b) shows
The first type to detect the number of people people movement through the gate. The picture
requires personal contact, which obstruct the path. shows the entrance of Sri Sangameswarar Temple
7
at Bhavani where a flow rate of people is very high

throughout the year. Fig. 2. a) Passage of people through wooden barriers (Front
view)
Fig. 1. a) Entry gate
Fig. 2. b) Passage of people through wooden barriers (Back

View)
When a person enters through the door where

sensor hands are fixed, the 5mA light passed is
cut and the information is passed to the LDR and
then to Programmable Interface Controller (PIC).
An assembly language program has been written
to process the data received from LDR. The
assembly language program stored in the PIC is
working and increments the count by 1. When the
person enters, the light beam is cut, then counter
is incremented by 1 and, when a person exits from
the area via the door, the counter is decremented
by 1. This information is displayed in the Liquid
Crystal Display (LCD) and also in software user
Fig. 1. b) People flow through the entry gate
interface. Fig. 3 provides the prototype of crowd
monitoring system.
For effective counting, the pathway is designed
as a single entry by using wooden barriers. The
following Fig. 2 a) and 2 b) show the scenario.
8
and serial programming and re-programming with

flash memory capability. There are five ports in this
microcontroller, namely PORT A, PORT B, PORT
C, PORT D and PORT E. Each pin in the ports can
be used as either input or output pins.
V. EXPERIMENTAL OUTPUTS
In virtue of Visual Basic, the proposed method
has been successfully implemented as an actual
system for counting people. Comparing to previous
methods, the method has higher accuracy in
counting the people. The interface designed for
displaying output is user friendly. The output
shows the number of persons inside a particular
area. It also includes date and time. The data
collected from this system can be used for various
analysis purposes viz. hourly, daily, monthly and
Figure 3. Prototype of crowd monitoring system yearly report for understanding the crowd behavior
in one particular place. The following tables show
The components of the prototype are discussed the result of people counting process held at Sri
one by one in the following sections. Sangameswarar Temple, Bhavani. We conducted
the experiment by allowing the individual to pass
A. Power Supply Unit through a queue where the system is installed.
The following table depicts the output of our
Since all electronic circuits work only with low dc system.
voltage it needs a power supply unit to provide the
appropriate voltage supply. This unit consists of a
transformer, rectifier, filter and regulator. AC Percentage
S.
voltage typically 230v is connected to the Actual System
Date Time
transformer that steps the AC voltage down to the Strength Count of
No
level to the desired AC voltage. A diode rectifier correctness
then provides a full wave rectified voltage that is
initially filtered by a simple capacitive filter to 1 1/11/10 09:09:10 80 80 100%
produce a DC voltage. In our prototype, the power
supply unit provides pure DC 5V
2 1/11/10 09:16:20 120 120 100%
B. Step down Transformer
Step down transformers are designed to reduce 3 1/11/10 09:22:50 160 160 100%
electrical voltage. This kind of transformer "steps
down" the voltage applied to it. In our prototype, 4 1/11/10 09:29:40 183 183 100%
the step down transformer converts 230 AC to 15-
0-15 AC. 5 1/11/10 09:35:50 200 200 100%
C. Light Dependent Resistors
LDRs or Light Dependent Resistors are very 6 1/11/10 09:40:59 230 230 100%
useful especially in light/dark sensor circuits.
Normally the resistance of an LDR is very high, 7 1/11/10 09:45:10 270 270 100%
sometimes as high as 1000 000 ohms, but when
they are illuminated with light, resistance drops TABLE I
dramatically. PEOPLE COUNTING RESULT OF CROWD MONITORING
D. Programmable Interface Controller SYSTEM WITHOUT OCCLUSION
From the above table, it is clear that the system
The PIC microcontroller device used in this accurately counts the number of people entering
prototype is PIC16F877A. The PIC16F877 is 40 into an area. The following table depicts the result
pin IC. For this prototype development, we used where occlusions are in the crowd while passing
PICs because it is of low cost, wide availability, through the entry gate
extensive collection of application notes,
availability of low cost or free development tools,
9
Percentag Advanced Video and Signal Based

Surveillance, pp. 306–311 (2005)
e
S. [4] Sidla, O., Lypetskyy, Y., Brandle, N., Seer, S.’
Actual System Pedestrian Detection and Tracking for
Date Time
Strength Count of Counting Applications in Crowded Situations’
No
correctnes Advanced Video and Signal Based
Surveillance, pp. 70–75 (2006)
s
[5] Casburn, L., Srinivasan, M., Metoyer, R. A.
and Quinn, M. J. (2005) ‘A Data-Driven
1 1/11/10 10:01:02 18 16 89% Model of Pedestrian Movement’. Proceedings
of the Third International Conference on
2 1/11/10 10:02:40 24 22 92% Pedestrian and Evacuation Dynamics
(September 2005)
3 1/11/10 10:03:50 30 27 90% [6] M.I.Kamel, M.Fkry, A.Mashat, N.M.Biqami,
H.Barhamtoshy, and I.Beedewy, ‘Monitoring,
Surveillance and control of the crowds in the
4 1/11/10 10:06:40 35 32 91% holy sites using SCADA System’, The
Seventeenth National Computers Conference,
5 1/11/10 10:09:50 42 36 86% Kingdom of Saudi Arabia, April 2004
[7] Fruin, J. J., ‘The Causes and Prevention of
6 1/11/10 10:12:59 51 46 90% Crowd Disasters’. First International
Conference on Engineering for Crowd Safety,
93% London, England, March 1993. Document on
7 1/11/10 10:20:10 59 55
the web,
http://www.crowddynamics.com/Main/Fruin%
TABLE II PEOPLE COUNTING RESULT OF CROWD
MONITORING SYSTEM WITH OCCLUSION
20%20causes.htm
[8] Still, K., Crowd Dynamics, PhD. Thesis,
The above table shows the result of our Mathematics Department University of
experiment where the people moving through the Warwick, 2000
entry gate are partially or fully occluded. [9] Tang W., Wan T., Patel, S., ‘Real-time crowd
movement on large scale terrains’, Theory and
VI. CONCLUSION Practice of Computer Graphics, p.p. 146-153,
2003.
In this paper, we suggested a simple and user
friendly prototype for counting the number of
people entering into a particular area. The system Prabakaran V., received his MCA from
has been developed by keeping in mind that University of Madras and M.Phil from
successful counting of people entering into an Bharathiar University. As an enthusiastic
area as part of the development of full pledged researcher, he presented papers in many
crowd monitoring system. In future work, to resolve National conferences and an International
the occlusion issue, we try to incorporate a camera Conference. His area of interest includes
in our system and it can also be used to store the Object oriented programming, Data mining
images of the public for future references. and Image processing
REFERENCES
[1] M.Rossi and A.Bozzoli, ‘Tracking and
Counting Moving People’, Proc. Second IEEE
International Conference on Image Processing,
pp. 212-216, 1994
[2] Masoud, O., Papanikolopoulos, N.P., ‘A Novel
Method for Tracking and Counting Pedestrians
in Real-time Using a Singe Camera’, Vehicular
Technology 50, 1267–1278 (2001)
[3] Liu, X., Tu, P.H., Rittscher, J., Perera, A.,
Krahnstoever, N.: ‘Detecting and Counting Dr. Arthanariee A.M., holds a Ph.D degree
People in Surveillance Applications’. in Mathematics from Madras University as
well as Masters Degree in Computer
10
Science from BITS, Pilani. He holds a

patent issued by the Govt. of India for the
invention in the field of Computer Science.
He has directed teams of Ph.D
researchers and industry experts for
developing patentable products. He
teaches strategy, project management,
creative problem solving, innovation and
integrated new product development for
last 37 years.
Sivakumar M has 10+ years of experience

in the software industry including Oracle
Corporation. He received his Bachelor
degree in Physics and Masters in
Computer Applications from the Bharathiar
University, India. He holds a patent for the
invention in embedded technology. He is
technically certified by various professional
bodies like ITIL, IBM Rational, Clear case
Administrator, OCP - Oracle Certified
Professional 10G and ISTQB
11
A Novel design of Electronic Voting System

Using Fingerprint
D. Ashok Kumar#1, T. Ummal Sariba Begum#2
#1 #2
Department of Computer Science, V .S.S. UGC Research Fellow, V.S.S. Government Arts
Government Arts College, College,
Pulankurichi – 630 405, Sivagangai, Tamil Nadu, Pulankurichi – 630 405, Sivagangai, Tamil Nadu,
India India
Abstract— The heart of democracy is voting. The heart of For many years, paper-based ballot is used as a way
voting is trust that each vote is recorded and tallied with to vote during campus election day. This matter put
accuracy and impartiality. The accuracy and impartiality are an inefficient way of voting process as students have
tallied in high rate with biometric system. Among these to queue up to register their name before they can
biometric signs, fingerprint has been researched the
longest period of time, and shows the most promising
vote. Furthermore, the traditional way of voting will
future in real-world applications. Because of their take a long process and time. So, the novel electronic
uniqueness and consistency over time, fingerprints have voting using minutiae will become the best solution
been used for identification over time. However, because of for the matters; besides provide easier way of voting.
the complex distortions among the different impression of Compared to existing voting system the Electronic
the same finger in real life, fingerprint recognition is still a voting has several advantages like: Electronic voting
challenging problem. Hence in this study, the authors are system is capable of saving considerable printing
interested in designing and analysing the Electronic Voting stationery and transport of large volumes of electoral
System based on the fingerprint minutiae which is the core material. It is easy to transport, store, and maintain. It
in current modern approach for fingerprint analysis. The
new design is analysed by conducting pilot election among
completely rules out the chance of invalid votes. Its
a class of students for selecting their representative. use results in reduction of polling time, resulting in
Various analysis predicted shows that the proposed fewer problems in electoral preparations, law and
electronic voting system resolves many issues of the order, candidate’s expenditure, etc. and easy and
current system with the help of biometric technology. accurate counting without mischief at the counting
Keywords— Biometric, Fingerprint, Minutiae, centre. It is also eco friendly [8].
Electronic Voting. Biometrics is the automated recognition of
individuals based on their behavioural and biological
I. INTRODUCTION characteristics. Biometric recognition means by
measuring an individual's suitable behavioural and
biological characteristics in a recognition inquiry and
Elections allow the populace to choose their comparing these data with the biometric reference
representatives and express their preferences for data which had been stored during a learning
how they will be governed. Naturally, the integrity of procedure, the identity of a specific user is
the election process is fundamental to the integrity of determined. A fingerprint is an impression of the
democracy itself. The election system must be friction ridges, from the surface of a fingertip.
sufficiently robust to withstand a variety of fraudulent Fingerprints have been used for personal
behaviours and must be sufficiently transparent and identification for many decades, more recently
comprehensible that voters and candidates can becoming recognition is nowadays one of the most
accept the results of an election. In context of important and popular biometric technologies mainly
Western democracies' current crisis, electronic voting because of the inherent ease in acquisition the
has become a very popular topic of discussion in numerous sources (ten fingers) available for
academic and technical circles. collection, and the established use and collections by
Voting is a method for a group such as a meeting or law enforcement agencies. Automatic fingerprint
an electorate to make a decision or express an identification is one of the most reliable biometric
opinion—often following discussions, debates, or technologies. This is because of the well known
election campaigns. It is often found in democracies fingerprint distinctiveness, persistence, ease of
and republics. Electronic voting (also known as e- acquisition and high matching accuracy rates.
voting) is a term encompassing several different Fingerprints are unique to each individual and they
types of voting, embracing both electronic means of do not change over time. Even identical twins do not
casting a vote and electronic means of counting carry identical fingerprints. Scientific research in
votes.
12
areas such as biology, embryology, anatomy and of students to chose their representative and Section
histology has supported these findings [28]. V concludes and states the future work plans.
Because biometric identifiers cannot be easily
misplaced, forged, or shared, they are considered II. ISSUES OF PRESENT VOTING SYSTEM
more reliable for person recognition than traditional There has been several studies on using
token or knowledge based methods. The objectives computer technologies to improve elections [5, 38,
of biometric recognition are user convenience (e.g., 21, 22, and 29].
money withdrawal without ATM card or PIN), better These studies caution against the risks of moving
security (e.g., difficult to forge access), and higher too quickly to adopt electronic voting system,
efficiency (e.g., lower overhead for computer because of the software engineering challenges,
password maintenance). The tremendous success of insider threats, network vulnerabilities, and the
fingerprint based recognition technology in law challenges of auditing.
enforcement applications, decreasing cost of Researchers in the electronic voting field have
fingerprint sensing devices, increasing availability of already reached a consensus pack of following core
inexpensive computing power, and growing identity properties that an electronic voting system should
fraud/theft have all ushered in an era of fingerprint- have [30]:
based person recognition applications in commercial, Accuracy: (1) it is not possible for a vote to be
civilian, and financial domains. So the Electronic altered, (2) it is not possible for a validated vote to be
voting system has to be improved based on the eliminated from the final tally, and (3) it is not
current technologies viz., biometric system. possible for an invalid vote to be counted in the final
There are some previous works which uses tally.
fingerprint for the purpose of voter identification or Democracy: (1) it permits only eligible voters to vote
authentication. As the fingerprint of every individual is and, (2) it ensures that eligible voters vote only once.
unique, it helps in maximizing the accuracy. A Privacy: (1) neither authorities nor anyone else can
database is created containing the fingerprint of all link any ballot to the voter who cast it and (2) no voter
the voters in the constituency. Illegal votes and can prove that he voted in a particular way.
repetition of votes is checked for in this system. Verifiability: anyone can independently verify that all
Hence if this system is employed the elections would votes have been counted correctly.
be fair and free from rigging. Collusion Resistance: no electoral entity (any server
Fingerprint recognition or fingerprint authentication participating in the election) or group of entities,
refers to the automated method of verifying a match running the election can work in a conspiracy to
between two human fingerprints. Fingerprints are introduce votes or to prevent voters from voting. If all
one of many forms of biometrics used to identify an entities conspire this property isn’t achieved. So, this
individual and verify their identity. Extensive research characteristic should be measured in terms of the
has been done on fingerprints in humans. Two of the total number of entities that must conspire to
fundamentally important conclusions that have risen guarantee a successful interference in the election.
from research are: (i) a person's fingerprint will not Availability: (1) the system works properly as long as
naturally change structure after about one year after the poll stands and (2) any voter can have access to
birth and (ii) the fingerprints of individuals are unique. it from the beginning to the end of the poll.
Even the fingerprints in twins are not the same. In Resume Ability: the system allows any voter who had
practice two humans with the same fingerprint have interrupted his/her voting process to resume it or
never been found [7]. In this study, for the fingerprint restart it while the poll stands
authentication the minutiae based matching is The existing elections were done in traditional way,
considered for higher recognition accuracy. Also, the using ballot, ink and tallying the votes afterward. But
matching accuracy of fingerprint based authentication this system prevents the election from being
systems has been shown to be very high. Fingerprint accurate. Problems encounter the usual elections are
– based authentication systems continue to dominate as follows:
the biometrics market by accounting for almost 52% • It requires human participation, in tallying the
of authentication systems based on biometric traits votes that makes the elections time consuming
[2]. and prone to human error.
This paper is organized as follows: The section II
describes the issues of the present voting system, • The voter find the event boring resulting to a
section III discusses the fundamentals of finger print
small number of voters.
authentication system Section III describes the
proposed novel application for Electronic Voting
• Deceitful election mechanism.
Systems, Section IV describes the Experimental
Results of a pilot election conducted among a class
13
• Constant spending funds for the elections staff

every year.
So, the proposed electronic voting system has to

be addressed these problems.
III FUNDAMENTALS OF FINGERPRINT

AUTHENTICATION SYSTEM
The types of information that can be collected from

a fingerprints friction ridge impression can be
categorized as Level 1, Level 2, or Level 3 features Fig 2 Characteristic Attributes of a Minutiae
as shown in fig 1. Level 2 features or minutiae refer
to the various ways that the ridges can be
discontinuous. These are essentially Galton In a recently published World Biometric Market
characteristics, namely ridge endings and ridge Outlook (2005-2008), analysts predict that the
bifurcations. A ridge ending is defined as the ridge average annual growth rate of the global biometric
point where a ridge ends abruptly. A bifurcation is market is more than 28%, by 2007 [11]. The
defined as the ridge point where a ridge bifurcates technologies that would be included in this are
into two ridges. Minutiae are the most prominent fingerprint technology by 60%, facial & iris by 13%,
features, generally stable and robust to fingerprint keystroke by 0.5% and digital signature scans by
impression conditions. The distribution of minutiae in 2.5% Basically there are two types of fingerprint
a fingerprint is considered unique and most of the Recognition System:
automated matchers use this property to uniquely (1) AFAS ( Automatic Fingerprint
identify fingerprints. Uniqueness of fingerprint based Authentication System)
on minutiae points has been quantified by Galton [7]. (2) AFIS ( Automatic Fingerprint
Statistical analysis has shown that Level 2 features Identification / Verification System )
have sufficient discriminating power to establish the
individuality of fingerprints [34]. 1) AFAS (Automatic Fingerprint Authentication
System) Components of AFIS are: [40] [10][42]
1. Physical Fingerprint required as input.
2. Input is processed by using various image
processing tools and databases and Classification of
Fingerprints. The basic fundamental steps of these
systems (see Fig (3) are image acquisition, pre-
processing segmentation, enhancement etc), feature
extraction, matching along with classification through
databases. Authentication or verification systems
authenticate the person's identity by comparing the
own biometric template(s) stored in database (One-
to-One comparison). An identification system
recognize an individual by searching the entire
templates in database for match (One-to-Many
Fig 1 Fingerprint features at Level 1, Level 2, and Level 3 [36, 23] Comparison) [31] [17].
The Fig 2 shows the clear view of minutiae. A
minutia is characterized by its location and
orientation.
Fig 3 Typical Structure for Fingerprint System [16]
14
2) AFIS (Automatic Fingerprint Identification/ significant percentage of fingerprint images is of poor

Verification System) quality. In fact, a single fingerprint image may contain
A fingerprint recognition system operates either in regions of good, medium, and poor quality. Thus an
verification mode or in identification mode. The enhancement algorithm which can improve the
various stages in a fingerprint verification system are quality of ridge structure is necessary. A survey on
shown in Fig 4. different enhancement techniques can be found in
[2]. The paper [9] describes the popular
enhancement algorithm by Sharat et al. [33], which
used contextual filtering in Fourier domain. In paper
[24], the enhancement technique like histogram,
Fourier and Gabor are compared and best technique
gabor is found.
There has been lot of interesting work done in
enhancing fingerprints. Sherlock [35] proposed
enhancing the features in a fingerprint image by
directional Fourier filtering. This frequency domain
filtering is computationally less expensive than the
spatial convolution of the image with filters. The
filtered image is usually binarized or thinned for
Fig 4 Architecture of Fingerprint Verification System
feature extraction. But there has been an effort to
The first stage is the data acquisition stage in extract features from grey scale images, Maio and
which a fingerprint image is obtained from an Maltoni [16] proposed an algorithm to extract features
individual by using a sensor. The next stage is the from gray scale images. The feature extraction
pre-processing stage in which the input fingerprint is algorithm has usually been employed on thinned
processed with some standard image processing images. Jain [11] and Ratha [27] developed
algorithms for noise removal and smoothening. The algorithms for thinned images, their approach has
pre-processed fingerprint image is then enhanced involved local neighbourhood based processing on
using specifically designed enhancement algorithms the images.
which exploit the periodic and directional nature of Many authors have identified the need to perform
the ridges. The enhanced image is then used to post processing on fingerprint images to remove the
extract salient features in the feature extraction false minutiae, Ratha et al., where the minutiae are
stage. Finally, the extracted features are used for validated based upon heuristics like distance. Since
matching in the matching stage. the fingerprint based system rely on matching
Data Acquisition: Traditionally, in law enforcement between the query fingerprint and the database
applications fingerprints were acquired off-line by fingerprint, classification of the database results in
transferring the inked impression on a paper. the query only searching in a particular class. Many
Nowadays, the automated fingerprint verification attempts [16] [20] have been made to classify the
systems use live-scan digital images of fingerprints fingerprints based upon core as well as delta points;
acquired from a fingerprint sensor. These sensors these have been point based approach. The
are based on optical, capacitance, ultrasonic, thermal matching forms the heart of any fingerprint; the query
and other imaging technologies. The techniques fingerprint of even a client is usually a transformed
followed in these sensors are discussed in [2]. version of the database fingerprint. This involved
Image Pre-processing: The preprocessing steps try registration of the images before obtaining the match.
to compensate for the variation in lighting, contrast There have been several prior approaches that
and other inconsistencies which are introduced by addressed this. Ranade and Rosenfield [26]
the sensor during the acquisition process. The paper proposed an iterative approach for obtaining point
[1] discusses the preprocessing steps generally correspondences.
used, which are Gaussian Blur, Sliding-window The fingerprint enhancement techniques proposed
Contrast Adjustment, Histogram based Intensity by Chen et al. [27], is based on the convolution of the
Level and etc., image with Gabor filters which has the local ridge
Fingerprint Image Enhancement: The Performance orientation and ridge frequency. The algorithm
of fingerprint feature extraction and matching includes normalization, ridge orientation estimation,
algorithms relies heavily on the quality of the input ridge frequency estimation and filtering.
fingerprint images. Due to various factors such as The paper [24] evaluates the performance of three
skin conditions (e.g., we, dry, cuts, scars and types of image enhancement techniques and their
bruises). Non-uniform finger pressure, noise impact in minutiae detection. In this work we have
introduced by sensor and inherently poor-quality taken the account of hough transformation and
fingers (e.g., manual workers, elderly people), a
15
analyzed the results with previous referred to the core point and finger code obtained. The paper
transformations. [32] proposed a fingerprint matching which is more
robust at shift and rotation of the fingerprints while it
Feature Extraction: In this section I describe is of high accuracy. A Survey on Ridge feature based
various levels of feature in fingerprint. The levels of matching techniques is proposed in paper [1].
features which is to be extracted are Minutiae, Pores, Minutiae based Matching
Ridge Contour Extraction. Let T and Q be the feature vectors, representing
Minutiae Extraction The next step after enhancement minutiae points, form the template and query
of the image is the extraction of minutiae. The fingerprint, respectively. Each element of these
enhanced image is binarized first in this step. The feature vectors is a minutiae point, which may be
skeleton of the image is then formed. The minutiae described by different attributes such as location,
points are then extracted by the following method. orientation, type, quality of the neighbourhood region,
The binary image is thinned as a result of which a etc. The most common representation of a minutiae
ridge is only one pixel wide. The minutiae points are is the triplet x, y, θ where x, y is the minutiae location
thus those which have a pixel value of one (ridge and θ is the minutiae angle. Let the number of
ending) as their neighbour or more than two ones minutiae in T and Q be m and n, respectively.
(ridge bifurcations) in their neighbourhood. This ends T=m1, m2,…, mm, mi = xi, yi, θi, i =1…m
the process of extraction of minutiae points.
Let (x, y) denote a pixel on a thinned ridge, and Q=m’1, m’2,…,m’ni m’j = x’j, y’j, θ ‘j, j=1…n
N0, N1,…, N7 denote its eight neighbours. A pixel (x, (9)
y) is a
7
Ridge ending if ( ∑ Ni ) = 1 A minutiae mi in T and mj’ in Q are considered
i =0 matching, if following conditions are satisfied:
7
if (
Ridge bifurcation
∑ Ni ) > 2
i =0
Pores extraction: Pores are extremely fine details 2 2

which are lost after the enhancement stage. sd ( m’j , mi)= ((x’j-xi) +(y’j-yi) ≤ ro
Kryszczuk et al. [15] and [3] have proposed
skeletonization based approach for pore extraction. dd (m’j , mi) = min ( |θ’j – θi|, 360 - |θ’j – θi|) ≤ θo
Jain et al. [13] have proposed a pore extraction (10)
technique directly from gray scale image. . A recent
study [9] by the International Biometric Group has
proposed a new approach for pore extraction which
utilizes orientation information of pores along with the Here, r0 and θ0 are the parameters of the tolerance
location information. window which is required to compensate for errors in
Ridge Contour Extraction: Ridge contours can be feature extraction and distortions caused due to skin
extracted by using classical edge detection plasticity.
algorithms. Jain et al [13] have proposed an
algorithm to extract the ridge contours which used a The number of “matching” minutiae points can be
simple filter to detect ridge contours. maximized, if a proper alignment (registration
parameters) between query and template fingerprints
Fingerprint Matching: A variety of automatic
can be found. Correctly aligning two fingerprints
fingerprint matching algorithms have been proposed
in the pattern recognition literature. A useful literature requires finding a complex geometrical
survey on fingerprint recognition can be found in [2]. transformation function (map ()), that maps the two
One family uses correlation based matching [4]. [6] minutiae set (Q and T) the desirable characteristics of
and [19]. Correlation matching is less tolerant to map () functions are: it should be tolerant distortion; it
rotational and translational variances of the should recover rotation, translation and scale
fingerprint and of extra noise in the image. Another parameters correctly.
family uses Minutiae-based matching [37], [10], [14].
Minutiae matching are certainly the most well known For the fingerprint enhancement technique, we
and widely used method for fingerprint matching. In compare the four types of fingerprint enhancement
general minutiae matching are considered by most to technique viz., Histogram, Fourier filter, Hough
have, higher recognition accuracy. The last family Transform and Gabor filter and find the best
uses Ridge feature based matching [12]. Jain et al enhancement based on the following measures
[12] proposed a local texture analysis where the
fingerprint area of interest is tessellated with respect
16
III. A NOVEL ELECTRONIC VOTING SYSTEM allowed to vote. Otherwise he is rejected and give the
The main core of this study is to design an beep sound. The person who is authenticated may
electronic voting system based on fingerprint vote for their beloved one by giving his fingerprint to
minutiae is discussed in this section by two phases: i) the fingerprint scanner of corresponding nominee.
Enrolment Process and ii) Voting Process. This is the innovation we made so that no person is
allowed to press voting button as it is one of the
i) Enrolment Process
drawbacks of the present voting machine. After the
The Fig 6 shows the enrolment process clearly. completion of voting, one can know the status of the
The Process involved in using fingerprint scanner for nominees by clicking the count button.
election is very simple. First, the chosen finger for
example, the thumb is captured and extracted. The
fingerprint template is then enrolled and store in a
local repository, a database. This primary process is IV. EXPERIMENTAL RESULTS
done during the registration process. After that, the In this work, we have conducted the Pilot Election
chosen finger can be live scan. The fingerprint using a Personal Computer with four fingerprint
template is then processed and extracted. It will scanners for selecting class representative. For that,
subsequently match the scanned fingerprint against we have created the database which consists of the
the stored template. Upon verification, they will have fingerprint of the Computer Science department
the access to vote for their desired candidates. students with the number of 80 (45 males and 35
Mismatched fingerprint certainly would indicate denial females). The database is created based on the
form the access. digital personal scanner. This primary process is
done during the registration process. After that, the
chosen finger can be live scan. The fingerprint
Fingerprint Scanner template is then processed and extracted. It will
subsequently match the scanned fingerprint against
the stored template. Upon verification, they will have
Capture the the access to vote for their desired candidates.
Fingerprint image Mismatched fingerprint certainly would indicate denial
from the access.
During the voting, the voter first places his thumb
on the touch sensitive region. If the fingerprint
FP Enhancement matches he is allowed to vote. In case the print is not
stored before, a single beep is given, so the person
cannot vote OR if the same person votes again, the
Minutiae Extraction system should give a double beep, so that the
security can be alerted. The system is programmed
to recognize a fingerprint twice, but to give a beep for
more than once.
There are three nominees for the selection of
Fingerprint representative and each student is asked to vote for
the candidates they wish by checking their identity
Database through fingerprint and allowing them to vote by
giving thumb impression against the fingerprint
Fig 6 Enrollment Process scanner of candidate.
The Table 6 shows the pilot election results.
ii) A Novel Design for E-Voting Process TABLE 6: PILOT ELECTION RESULT
In the Fig 7, the first process is capture the input S. Name of the Count of the Votes
No Candidate Polled
image, the captured image is then enhanced by using
the best enhancement technique Gabor. The next 1 M. Jeyaraj 20
step after enhancement is the extraction of minutiae. 2 S. Ashik 15
After extracting minutiae, it is compared with the 3 P. Kokila 10
template which is stored in the database based on
4 P. Krishna 35
minutiae based matching as proposed in the previous
chapter. If the matching result is true, the person is Total 80
17
Fingerprint Scanner
From the results of Table 6 it is declared that Mr.

Krishna has been elected as Representative of the
Class of Students. Capture the Fingerprint image
V. CONCLUSIONS AND FUTURE DIRECTION

For over a century, fingerprints have been one of
the most highly used methods for human recognition; FP Enhancement
automated biometric systems have only been
available in recent years. This work is successfully
implemented and evaluated four different models and
PC based electronic voting system under Matlab 7.5. Minutiae Extraction
The arrived results were significant and more
comparable. It proves the fact that the fingerprint
image enhancement step will certainly improve the
verification performance of the fingerprint based
recognition system. Referring
Not Allow to False Match
The best enhancement technique Gabor is used to Vote and Give Fingerprint
Beep Sound -ing
enhance the fingerprints for electronic voting and the Database
report of the pilot study for students’ election shown
the better accuracy. By the use of this PC based
True (Allow to Vote)
voting system, the student’s representative is elected
in a proper way with high security. Because
fingerprints have a generally broad acceptance with
the general public , law enforcement and the forensic FPS1 Nominee-1
science community, they will continue to be used with FPS-
many governments’ legacy systems and will be Fingerp
FPS2 Nominee-2
utilized in new systems for evolving applications that rint
require a reliable biometric. Scanner
In this work, we counted the spurious minutiae and FPS3 Nominee-3
did not address impact of image enhancement
algorithm with spurious minutiae removal algorithms Nominee-4
FPS4
and also we are designed only a PC based electronic
voting system. In future, we will design a device with
Biometric Technology which can be used as if Indian
Electronic Voting Machine.
Election Result
Fig.7 A Novel Design for E-Voting Process
ACKNOWLEDGMENT
This work is a part of a Research Project and
authors are thankful to UGC for funding the Project
(File No. F-38-258/2009 (SR) Dt: 19.12.2009).The
authors would like to thank the anonymous reviewers
for their thorough reviews, and constructive
suggestions which significantly enhance the
presentation of the paper.
REFERENCES
Abishek Rawat,, A Hierarchical Fingerprint Matching System,
Indian Institute of Technology, Kanpur, July 2009
18
Anil K. Jain and David maltoni. , Handbook of Fingerprint Rajnikannan M., Ashok Kumar D., Muthuraj, Estimating the Impact
Recognition, Springer-verlag New York, Inc., Secaucus, NJ, of Fingerprint Image Enhancement Algorithms for Better
USA, 2003 Minutia Detection, International Journal of Computer
Ashbaugh D. R., Quantitative-Qualitative Friction Ridge Analysis: Application, No.1 Article 7, 2010
An Introduction to basic and advanced Ridgeology. CRC Raju Sonavane, Dr. B.S. Sawant., Noisy Fingerprint Image
Press, 1999 Enhancement Technique for Image Analysis: A Structure
Bahuguna R., Fingerprint Verification using Hologram matched Similarity Measure Approach, SNS International Journal of
filterings, Biometric Consortium Eighth Meeting, San Jose, Computer Science and Network Security, Vol. 7, No. 9,
CA, 1996 September 2007
California Internet Voting Task Force. A Report on the Feasibility of Ranade A. and Rosenfeld A, Point pattern matching by relaxation.
Internet Voting, Jan.2000. Pattern Recognition, 12(2):269–275, 1993.
http://www.ss.ca.gov/executive/ivote/ Ratha, N., Chen, S., and Jain, A. Adaptive flow orientation based
Coetzee L. and Botha E.C., Fingerprint Recognition in low quality feature extraction in fingerprint images. Pattern Recognition
images, Pattern Recognition , Vol.26, No.10, pages. 1441- 28, 11, 1657–1672., 1995
1460, 1993 Ridges and Furrows – history and science of fingerprint
Francis Galton. Fingerprints, Macmillan, London, 1892 identification technology and legal issues.
Frank Vahid and Tony Givargis, Embedded System: Design A http://ridgeand.furrows.homestead.com/fingerprint.html
unified Hardware/Software Introduction, John Wiley & Sons, Rubin A.D. Security considerations for remote electronic voting.
Inc, 2002. Communications of the ACM, 45(12):39-44,Dec.2002.
International Biometric Group. Analysis of Level 3 features at High Rui Joaquim, André Zúquete, Paulo Ferreira Revs- A Robust
Resolutions. http://level3tk.sourcforge.net/, 2008 Electronic Voting Voting System Instituto Superior Técnico
Jain A. K., Hong L. , Bolle R., Online Fingerprint Verification, IEEE Salil Prabhakar, "Fingerprint classification and matching using
Trans Patt Anal Mach Intell, Vol 19, No. 4., Pages 302-314, filterbank", Ph. D. Thesis, 2001.
1997 Shaharam Mohammadi, Ali Frajzadeh A . Matching Algorithm of
Jain A., Hong, L., Pankanti, S., and Bolle, R. An identity Minutiae for Real Time Fingerprint Identification System ,
authentication system using fingerprints. In Proceedings of World Academy of Science, Engineering and Technology 60
the IEEE ,vol. 85, pp. 1365–1388.Sep 1997 ,2009
Jain A.K, Prabhakar S., Hong L. , Pankanti S., Filter bank based Sharat Chikkerur, Alexander N. Cartwright and Venu Govindaraju.,
fingerprint matching, IEEE Trans. Image Processing, 9(5): Fingerprint Enhancment using STFT analysis., Pattern
Pages 846–859, 2000. Recognition., 40(1):198-211, 2007
Jain A.K., Chen Y. and Demirkus M., Pores and Ridges: High Sharath Pankanti, Salil Prabhakar and Anil K. Jain., On the
Resolution Fingerprint Matching using Level 3 features, Individuality of Fingerprints, IEEE Trans. Pattern Anal. Mach.
PMAI, 29(1):15-27, January 2007 Intell, 24(8):1010-1025, 2002
Jie Y., Yifang Y., Renjie Z. and Qifa S., Fingerprint minutiae Sherlock, D. B. G., Monro, D. M., and Millard, K. Fingerprint
matching algorithm for real time system, Pattern enhancement by directional Fourier filtering. In IEEE Proc.
Recognition , pp. 143–146 , 2006 Vis. Image Signal Processing , vol 141, pp. 87–94., 1994
Kryszczuk K., Drygajlo A. and Morier P., Extraction of Level 2 and The Thin Blue Line.
Level 3 features for Fragmentary Fingerprints. In proc. http://policensw.com/info/fingerprints/finger06.html, October
Second COST Action 275 Workshop, pages 83-303, 1996. 2006
Maio D. and Maltoni D. , Direct gray scale minutia detection in Tico M., Kuosmanen P., Fingerprint Matching using an Orientation
fingerprints. Transactions on PAMI, 19(1), 1997. based minutia descriptor, IEEE Trans. On Patt. Analy and
Maltoni D., Dmaio, Jain A.K., Prabhakar S. , "Hand book of Mach Intell, Vol. 25, No. 8, Pages 1009-1014, 2003
Fingerprint Recognition", Springer, 2003. Voting: What Is; What Could Be, July 2001.
Mariam BT. Samawi , Web Based Campus Election using Thumb http://www.vote.caltech.edu/Reports/
Recognition, Mara University of Technology Faculty of WUZHILI, "Fingerprint recognition," Student project, Hong Kong
Information Technology and Quantitative Science, May 2006 Baptist University, April 2002.
Marsh R.A., Petty G. S., Optical Fingerprint Correlator, US Patent Xia X. and O’Gorman, L. Innovations in fingerprint capture
5050220, 1991 devices. Journal of Pattern Recognition, Pergamon Press,
Meltemp Ballan & F. Ayhan Sakarya & Brian L.Evans, "A Vol. 36, No. 2, pp. 361-370, 2002
Fingerprint Classification Technique Using Directional Xuejun Tan Bhanu, Yingqiang Lin B. , Fingerprint classification
Images". based on learned features, Center for Res. In Intelligent
Mercuri R.. Electronic Vote Tabulation Checks and balances. PhD Syst., Univ. of California, Riverside, CA, USA; Aug 2005.
thesis, University of Pennsylvania, Philadelphia, PA, Zhou Wang ,Alan Conrad Bovik’sJain,A., “Image Quality
Oct.2000 Assessment :from error visibility to structure imilarity IEEE
National Science Foundation. Report on the National Workshop on transaction On image processing” Vol.13 No4, April 2004
Internet Voting: Issues and Research Agenda, Mar.2001.
Nieuwendijk H.Y.D, Fingerprints.
http://www.xs4all.nl/~dacty/minu.htm, October 2006
19
GPS Tracking Simulation by Path Replaying

G. Rajendran#1, Dr. M. Arthanari#2 , M. Sivakumar#3
#1
Assistant Professor of Computer Science,
Government Arts College (Autonomous), Salem-636007,India.
#2
Director, Bharathidasan School of Computer Applications,
Ellispettai-638116, Tamilnadu, India.
#3
Doctoral Research Scholar, Anna University, Coimbatore, Tamilnadu, India.
Abstract—GPS (Global Positioning System) has a Private and Government owned transport, shipping
variety of applications among which real-time tracking and cargo companies have started using GPS to
finds significance in day-to-day life. GPS tracking is track their vehicles. They maintain a tracking
defined as the measurement of object position and database of their own vehicles and they do not want
orientation in a given coordinate system using GPS to share this database with anybody else because of
data at different points of time. GPS data are collected
from GPS receivers attached to the moving objects and
the fear that these data may be used by their
these data are used for tracking objects in real-time. competitors. But researchers need GPS databases
Researchers who work in GPS tracking need GPS generated by hundreds of GPS receivers. This
databases which contain huge volume of GPS data introduces a new problem domain of non-availability
generated by hundreds of GPS receivers. But the of sample data to develop and test GPS
presently available GPS databases are owned by applications. Hence the need for a GPS tracking
private players and are not available for use by the simulator is vital to generate GPS databases. Such a
researchers. This work is an attempt to generate a simulator will greatly reduce the expenses in
database of GPS data which can be used by the research area and enable us to have more
researchers to develop and test GPS applications. The
approach consists of three successive steps:
researches carried out in the allocated budgets.
Collecting floating car data (FCD) of each path once in
a log file; refining the log file; and replaying multiple This paper presents a GPS tracking simulation
instances of several log files simultaneously after process which produces a database of GPS tracking
replacing some old values with new values to simulate data. GPS simulation has already been carried out by
GPS tracking. Thus a single path tracked previously some of the researchers, but with limitations like
can be used to produce a tracking simulation of a hardware dependency, involvement of certain cost,
number of moving objects by path replaying and each complexity, lack of provision for GPS database
and every execution of the simulation generates a set creation and lack of provision for integration of digital
of new GPS tracking data of several moving objects.
These data are stored in a database and can be used
maps. These limitations have been addressed in this
as sample data for developing and testing GPS work. The remainder of this paper is organised as
applications. follows. Section 2 of this paper describes the tracking
of moving objects using GPS. Previous work in this
Keywords: GPS receiver, GPS simulation, GPS data, area is discussed in Section 3. In Section 4, the
Real time tracking. simulation scenario is introduced. The results of the
simulation along with a comparison of output data
I. INTRODUCTION items with real-time data are dealt in Section 5. The
work is concluded and the possible improvements
The applications of GPS tracking in real time have are discussed in Section 6. The path replaying
found its place in almost all walks of life, for instance, simulator has been designed using Matlab 7.6.
navigation, map making, land surveying, fishing and
trekking. GPS has many technical and economical
II. TRACKING USING GPS
benefits to almost all industries and nowadays many
companies are developing GPS [1] enabled A. A. Global Positioning System
applications and systems. More research is being
carried out in this domain as GPS has the unique
There are many thousands of civil users of GPS
capability of locating any moving object over the
system world-wide. GPS is a Satellite Navigation
earth in terms of latitude, longitude and altitude with
System funded, controlled and operated by the U. S.
high accuracy.
Department of Defense [2, 3]. The GPS system
consists of three segments viz., satellites that
transmit the position information, the ground stations
20
that are used to control the satellites, and finally there Many of the tracking systems combine GPS, GSM
is the receiver that computes its location anywhere in technologies. In less than ten years since the first
the world based on information it gets from the GSM network was commercially launched, it became
satellites [4]. the world's leading and fastest growing mobile
standard, spanning over 200 countries. There is at
The satellite segment consists of a minimum of 21 least one cell tower in every 900m-1000m radius in
satellites and 3 working spares. The GPS satellites the high traffic regions (city limits) and in the low
broadcasts two signals, PPS (Precise Positioning traffic regions (high ways) a single cell tower can
Service) which is available for use by military and cover a radius up to 10 km. The operation of GPS
government and SPS (Standard Positioning Service) tracking is explained in Fig. 1. The GPS receiver
which is available for use by public [5]. The Control captures position data from the satellites, computes
Segment consists of a system of control stations the position of the object, say, a vehicle, and sends
located around the world. The Master Control facility this information to a central base station, using SMS
is located in Colorado. These stations measure (Short Message Service) or GPRS(General Packet
signals from the satellites which are incorporated into Radio Service). If the optional storage module is
orbital models which in turn compute precise orbital installed, location data can even be stored when the
data and satellites clock corrections for each and vehicle is out of range of the cellular operator and
every satellite. The Master Control station uploads retrieved later. GSM technology is used to transmit
ephemeris and clock data to the satellites. The this information which in turn is collected by the
satellites then send subsets of the orbital ephemeris server at the base station. The geographical position
data to GPS receivers over radio signals. The of the object can be displayed at the base station
Receiver Segment consists of GPS receivers which using a suitable application.
are used for navigation, positioning, time
dissemination and other applications. A GPS receiver B. B. NMEA Specification
receives signals from more satellites than are actually
needed for a position fix. The reason for this is that if The NMEA (National Marine Electronics
one satellite becomes unavailable, the receiver Association) has developed a specification that
knows exactly where to find the best possible defines the interface between various pieces of
replacement. Three satellites are required for two marine electronic equipments. The NMEA standard
dimensional positioning and four satellites are permits marine electronics to send information to
required for three dimensional positioning. Two computers and to other marine equipments [6] in
dimensional positioning reports position only in terms predefined formats. GPS receiver communication is
of latitude and longitude whereas three dimensional defined with NMEA specification. Most computer
positioning reports position in terms of altitude as programs that provide real time position information
well. In general, a GPS receiver can provide position recognize data that are in NMEA format which
information with an error of less than 10 meters, and includes the latitude, longitude, velocity and time
velocity information with an error of less than 5 computed by the GPS receiver. In NMEA
meters per second. specification system, data is sent as a line of text,
called a sentence which is totally self contained and
independent from other sentences. The data is
contained within this single line and the data items
are separated by commas. The commas act as
terminators for the sentences and the programs that
read the data should only use the commas to
determine the end of a data item.
The GPS receivers produce GPS data in the form

of standard NMEA sentences. The most important
NMEA sentences include the $GPGGA which
provides the current fix data, the $GPRMC which
provides the minimum GPS sentences information,
and the $GPGSA which provides the Satellite status
data. The $GPRMC sentence is used for the tracking
of moving objects.
Fig. 1 GPS tracking with the help of a GPS receiver fixed in a
moving object.
21
III. PREVIOUS W ORK researchers to develop and test GPS applications.

This generator can be used not only by software
GPS simulators help the researchers to carry out developers, but also by users, who want to learn
their research work with minimal cost and with navigation software before buying GPS receiver. This
accurate data which they need. Though some work system generated random directions, but not based
has been done in this area, more GPS simulators are on the existing routes available on ground.
being developed to suit the custom needs of the
hour. Thus a few number of GPS tracking simulators are
available but with some limitations. Mostly these
A GPS device emulator [6] namely, the GPS simulators involve some cost and hardware
Generator PRO, has been designed for providing dependent. Some simulators suffer from lack of
assistance in developing, testing and debugging support of customized digital maps. Some of the
programs and equipment working with the NMEA- generators require keyboard or mouse control for
0183 protocol. This emulator generates NMEA path creation. In some generators, there is no
messages from different data inputs. It can operate in provision for database creation. The GPS tracking
2 modes: 1)User select map, start point, heading, simulator proposed in this work addresses these
speed; 2)User select NMEA log file. The user can problems.
change speed and heading during simulation. The
generated NMEA sentences can be used by other
IV. MODELLING DYNAMIC ENVIRONMENT OF MOVING
mapping software. But buying this software involves
OBJECTS BY PATH REPLAYING
cost.
A device for generating NMEA sequences for The dynamic environment of moving objects is
testing embedded GPS reception firmware and modelled using the following steps.
hardware is described by Sinivee V [7]. This work
describes a prototype GPS data simulator designed C. Collecting Floating Car GPS data in a log file
and built in Department of Physics of Tallinn D. Log File Pre-processing
University of Technology, Estonia. Device can work E. Replaying multiple instances of several log file.
in standalone mode and also in conjunction with
F. A. Collecting Floating Car GPS data in a log file
control software. Configuration program can be used
to generate test strings without tester hardware as
well. First version of the device was limited to Nowadays, the main research focus in the
generating only one NMEA message and enabled community of Intelligent Transport Systems (ITS) is
simulation of communication errors. Later versions how to acquire real-time and dynamic transportation
were developed to a more universal device with information. This information can be applied in the
control via a GUI running on an ordinary PC. But transportation area like vehicle tracking, navigation,
there is no option provided in this software to create a road guidance and so on. GPS is one such system
database of NMEA sentences. Hardware dependent which is used to provide real time information on
GPS simulators [1, 8] are also available and they moving objects.
operate by generating pseudo GPS signals.
A keyboard or mouse controlled NMEA sentence

generator, Virace GPS Simulator [9] V0.01, can
produce 3 COM port outputs. This simulator supports
NMEA sentences like $GPRMC, $GPGGA, $GPGSA
and $GPGSV. A lot of defined keys for steering and
speed are available in this simulator. It supports three
display and input formats of latitude and longitude.
The disadvantage of using this is that it does not
support GPS track replaying.
A work in this area has recently been done by the

authors [10] to generate a sequence of NMEA
sentences which in turn are used to simulate a GPS
tracking environment. The generated data were Fig. 2 A moving vehicle (Floating car) fixed with a GPS receiver
similar to the data generated in real time by a GPS
receiver and this simulator was used to create a
database of sample GPS data which can be used by
22
The Floating Car (Probe Car) technique is one of 07742.4325,E : Longitude 77 deg 42.4325' E
the key technologies adopted by the ITS to get the 31.6 : Speed over the ground in knots
traffic information in recent years [11]. Its basic 317.52 : Course over the ground
principle is to periodically record the location, 140510 : Date – 14th of May 2010
direction, date, time and speed information of the A : Autonomous mode
traveling vehicle from a moving vehicle with the data *62 : The checksum data, always begins with *
of the GPS as shown in Fig 2. The information can be
processed by the related computing model and
algorithm so that the floating car data can be TABLE I
associated with the city road in real time [12]. This $GPRMC DATA FORMAT
data can also be used as a source of data for
creating research and commercial applications on
vehicle tracking and road guidance systems. Data Item Format Description
Message ID $GPRMC RMC protocol header.
Mostly, the GPS receivers generate $GPGGA, UTC Time
$GPGSA, $GPRMC, $GPVTG and $GPGSV (Coordinated hhmmss.sss Fix time to 1ms accuracy.
sentences at a regular time interval. A sample list of Universal Time)
NMEA sentences produced by the GPS receiver and A Data Valid.
Status Char
V Data invalid.
stored in a log file when travelled in a road is given in
Fig. 3. Latitude Float Degrees * 100 + minutes.
N/S Indicator Char N=north or S=south.
Longitude Float Degrees * 100 + minutes.
E/W Indicator Char E=East or W=West.
Speed over Speed Over Ground in
Float
Ground knots
Course over Course Over Ground in
Float
Ground Degrees
Date ddmmyy Current Date
Magnetic
Blank Not Used
Variation
E/W Indicator Blank Not Used

Mode Char A Autonomous
Fig. 3 Log file of floating car GPS data with $GPGGA, $GPGSA, Checksum *xx 2 Digits
$GPRMC, $GPVTG and $GPGSV sentences Message
<CR><LF> ASCII 13, ASCII 10
Terminator
G. B. Log File Pre-processing
The log file contains a number of different types of Hence the next step in the simulation process is to
sentences but the $GPRMC (recommended refine the log file by removing other sentences in
minimum sentence C) provides the essential GPS such a way that it contains the $GPRMC sentences
PVT (Position, Velocity and Time) data. All GPS only as shown in Fig 4. This refined log file now
receivers output this sentence along with some other contains the path of the probe car in terms of latitude
sentences. This data is used to locate moving objects and longitude at an interval of one second per
in terms of latitude and longitude. The moving object, sentence.
if attached with a GPS receiver, can be located with
the help of this NMEA sentence. The $GPRMC data In this context, replaying means picking the
format [13] is given in Table 1.An example of $GPRMC sentences one by one from the log file and
$GPRMC NMEA sentence is given below: plotting the latitude and longitude position of the
$GPRMC,120642.206,A,1118.4253,N,07742.4325, object continuously in a map. During replay, new
E,31.6,317.52,140510,,,A*62 $GPRMC strings are also generated for the moving
Where object as described below.
$GPRMC : Recommended Minimum sentence
C For instance, consider the following sentence in
120642.206 : Fix taken at 12:06:42.206 UTC refined log file.
A : Status A=active or V=Void.
1118.4253,N : Latitude 11 deg 18.4253' N
23
$GPRMC,120642.206,A,1118.4253,N,07742.4325, $GPRMC,111315.773,A,1120.8973,N,07743.0879,
E,31.6,317.52,140510,,,A*62 E,0.00,286.94,240410,,,A*6F
$GPRMC,111316.773,A,1120.8972,N,07743.0879,
E,0.00,286.94,240410,,,A*6D
GPRMC,111317.773,A,1120.8972,N,07743.0879,
E,0.00,286.94,240410,,,A*6C
$GPRMC,111318.773,A,1120.8972,N,07743.0878,
E,0.00,286.94,240410,,,A*62
$GPRMC,111319.773,A,1120.8972,N,07743.0878,
E,0.00,286.94,240410,,,A*63
$GPRMC,111320.772,A,1120.8972,N,07743.0877,
E,0.00,286.94,240410,,,A*67
$GPRMC,111321.772,A,1120.8974,N,07743.0872,
E,1.85,275.60,240410,,,A*6E
Fig. 4 Refined Log file of $GPRMC sentences $GPRMC,111322.772,A,1120.8979,N,07743.0860,
E,4.79,292.27,240410,,,A*6F
H. C. Replaying Multiple Instances of Several Log $GPRMC,111323.772,A,1120.8988,N,07743.0842,
Files E,6.96,294.28,240410,,,A*6A
$GPRMC,111324.771,A,1120.8996,N,07743.0819,
During simulation, the values in the date and time E,8.56,290.47,240410,,,A*60
fields in the above sentence are replaced with system The new $GPRMC sentences produced from the
date and time values as shown below. The micro- above sentences during simulation are listed below.
time in the time field is left unaltered because of its
insignificance. The check sum field is also left $GPRMC,040210.773,A,1120.8973,N,07743.0879,
unaltered as it is insignificant in simulation. The E,0.00,286.94,251110,,,A*6F
remaining values are treated as current values for $GPRMC,040211.773,A,1120.8972,N,07743.0879,
simulation. So the newly generated sentence for E,0.00,286.94, 251110,,,A*6D
simulation will be $GPRMC,040212.773,A,1120.8972,N,07743.0879,
E,0.00,286.94, 251110,,,A*6C
$GPRMC,151245.206,A,1118.4253,N,07742.4325, $GPRMC,040213.773,A,1120.8972,N,07743.0878,
E,31.6,317.52,221110,,,A*62 E,0.00,286.94, 251110,,,A*62
$GPRMC,040214.773,A,1120.8972,N,07743.0878,
The different fields of the newly generated E,0.00,286.94, 251110,,,A*63
sentence for the currently moving object can be $GPRMC,040215.772,A,1120.8972,N,07743.0877,
extracted and stored in the database. E,0.00,286.94, 251110,,,A*67
$GPRMC,040216.772,A,1120.8974,N,07743.0872,
E,1.85,275.60, 251110,,,A*6E
V. GPS TRACKING SIMULATION RESULTS $GPRMC,040217.772,A,1120.8979,N,07743.0860,
E,4.79,292.27, 251110,,,A*6F
Wonde-X series GPS receiver (ZX4125) was used $GPRMC,040218.772,A,1120.8988,N,07743.0842,
in order to produce log files once for each path. The E,6.96,294.28, 251110,,,A*6A
GPS receiver is fixed in a moving car and the NMEA $GPRMC,040219.771,A,1120.8996,N,07743.0819,
sentence generated by it are stored in the log file in a E,8.56,290.47, 251110,,,A*60
laptop kept in the moving car. The log-file is then
refined and replayed to produce simulation of one
moving object. During replay, new $GPRMC
sentences are produced out of the sentences present
in log file. For instance, the first ten sentences
originally available in the log file for a particular path
is given below.
24
Fig. 6 GPS database extracted from new $GPRMC sentences

Fig. 5a
produced by the simulator
Thus each moving object is associated with new
Thus, it is observed that the new $GPRMC
$GPRMC sentences generated out of old sentences
sentences generated during simulation resemble with
available in refined log file. The digital map built in the
the sentences generated by actual tracking. Besides,
earlier work [14] of authors is used to plot objects in a a replay of multiple instances of the previously
2D-plane. Multiple instances of the log-file are tracked path produces a GPS tracking simulation of
replayed in parallel to get simulation of several
moving objects on ground. The significant values are
moving objects as shown in Fig. 5a and 5b. The
stored in the GPS database.
values of fields in the new strings are stored in the
database as shown in Fig. 6. It can be noted that only VI. CONCLUSION AND FUTURE WORK
the significant values in the $GPRMC sentence is
stored in the database along with an object
This paper introduces a GPS tracking simulation
identification field, say, vehicle number.
process which is used to simulate a number of
moving objects by path replaying. It is found that the
generated sentences are similar to the data
generated in real time by a GPS receiver and they
are found to fit within standards. This GPS tracking
simulator has eliminated the limitations of the
previous work carried out in this area. The data
generated by this software are used to create a
database of sample GPS data which can be used by
researchers to develop and test GPS applications.
This generator can also be used by software
developers as well as the novice users of GPS to
learn navigation software. At present this system
simulates a fixed number of vehicles initially set
during the execution. In future, this work can be
extended to simulate random number of moving
vehicles at any point of time.
ACKNOWLEDGMENT
Fig. 5b
Fig. 5a, 5b Moving objects simulated and plotted in the map at The authors would like to thank Hashprompt Logistics
different instances of time. Management India Pvt. Ltd., for their support to
perform demonstration on GPS tracking using their
resources.
25
REFERENCES
[1] Yun Young-sun, Park Sung-min and Kee Chang-don, "Test of
GBAS Integrity Monitoring System Using GPS Simulator,"
Wuhan University Journal of Natural Sciences, vol. 8, no.2B,
pp. 697-704, 2003.
[2] Parkinson, B. W. and Spilker, J. J., "Global Positioning System:
Theory and Applications," American Institute of Aeronautics and
Astronautics, Washington, 1996.
[3] Interface control document, Navstar GPS Space Segment
(Navigation User Interfaces), 2000.
[4] Asoke K Talukder and Roopa R Yavagal, "Mobile Computing-
Technology, Applications and Service Creation," Tata McGraw
Hill Publishing Company, 2005.
[5] Goran M. Djuknic and Robert E. Richton., "Geolocation and
Assisted GPS," Computer, vol. 34, no. 2, pp.123-125, 2001.
[6] http://avangardo.com accessed on 10-10-2010.
[7] Sinivee V., "Simple yet efficient NMEA sentence generator for
testing GPS reception firmware and hardware," Novel
Algorithms and Techniques in Telecommunications and
Networking, Springer Netherlands, pp.207-210, 2010.
[8] Kou Yanhong, Yang Dongkai and Zhang Qishan, "GPS Satellite
Simulator Signal Estimation based on ANN," Journal of
Electronics(China), vol. 22 no.5, pp.458-464, 2005.
[9]http://www.gpspassion.com/forumsen/topic.asp?TOPIC_ID=114
933 accessed on 10-10-2010.
[10] Rajendran G., Arthanari. M., and Sivakumar M. "A Simplified
NMEA Sentence Generator for the Simulation of GPS
Tracking," Global Journal of Computer Science and
Technology, vol. 10, no. 14, November 2010.
[11] X. W. Dai, M.A. Ferman, “A simulation evaluation of a real-
time traffic information system using probe vehicles,” IEEE
ITSC, vol. 1, pp. 12-15, 2003.
[12] R. Kuehne, R.P. Schaefer and J. Mikat, “New approaches for
traffic management in metropolitan areas,” IFAC CTS, 2003.
[13] http://www.gpsinformation.org/dale/nmea.htm accessed
on 10-10-2010.
[14] Rajendran G., Arthanari. M., and Sivakumar M., "Customized
Digital Road Map Building using Floating Car GPS Data,"
International Journal of Computer Science and Information
Security, vol. 8, no. 3, pp. 21-29, June 2010.
26
Object Oriented Design of E-learning System

for Distance Education
#1 #2 #3
S. Mukhopadhyay , S. Dan , A.K.Singh ,
#1
Department of Computer Sc.Unversity of Burdwan, India
#2
Registrar, University of Burdwan, India
#3
Department of Computer Sc. Netaji Mahavidayalaya(B.U) Arambag , India
is a continuous process where the learner deprived to

Abstract: E-Learning, an important component of E- attend formal mode of education.
Education, is a novel system for communication
between learners and teachers. In this paper we have
E-Learning mechanism involves a variety of
tried to design an Object Oriented UML based E-
Learning system for distance education of an institute professionals such as Instructional designers, Course
with the objective to enhance the quality of teaching Writers/ Content Creators, Reviewers, Graphic
learning process in distance mode. The various objects Designers and Knowledge Organizers/Library, and
participating in the system have been identified and
shown by USE CASE diagrams and the object oriented Information Professionls. Vental and Prakash [4]
development process, object to object communication have described the following figure of e-education
among the objects, is shown by sequence diagram. environment for higher education.
Key words: E-Education, Distance Education,
Digital library, O-O Design
Course Instructional
Writers Designers
Content Organiser
I. Introduction Knowledge Partners in
E-learning System
Content
Organisers/Library Creators
and Information
E-Learning allows learners to study from home, Graphic
Visualize
minimizes the distance between learners and
Graphic
teachers, and enables institutes to provide a high Designer
quality education with minimum cost through distance

mode. E-Learning System/
Digital Libraries E-Universities
Distance education is really a golden opportunity for

those who are not in a position to continue their Internet Resources Non-Print Media
BBS Microfilms
education in normal mode, attending classes Network Floppies
Newsgroups Slides
regularly, due to some reasons. Some disadvantages Electronic
Conferences
Audio Tapes
Video Tapes
of distance education which does not include direct Discussion Forums Data Cartridges
CD-ROMS
provision of regular interaction between a student
and a teacher, no question of revision, etc may be
solved by e-learning. In e-learning, a component of e-
education, instructions are imparted with the help of
computer and communication technology. The web is a powerful information delivery
mechanism. It has lots of information, which can get
The ultimate objective of e-learning is to guide the
to the learner fast. The crucial issue is there must be
learners in every corner including remote place of the
a real value to the information that is being posted on
country and thereby providing education for all.
the web. The web is conceptually a kind of library, so
E-Learning covers a wide set of application and called Digital Library, a storehouse of information to
processes. It involves delivery of content of resources go to, in order to get something to read and
for learning via different communication protocols. It something to do.
27
Information collection, organization and II. Objective of E-learning

dissemination greatly affected by technology thereby
making/the predictions of Lancaster’s paperless The main objective of e-learning is to guide the
society to reality. Digital Library collection contains student / learner of every corner of the country. E-
fixed , permanent documents. Not only that current learning provides education all times so there is
Libraries have more dynamic collections, but digital notes question of day and night . Students are free to
environment will enable of quick handling. Digital access their notes / tutorials and assignments any
Libraries [1] are based on digital technologies. These time.
will break all the physical barrier of data transfer. It
can store the large amounts of information in various
form i.e. text, audio, video and graphic material. III. Benefits of E-learning in Distance Education
Learners can make effective search for the
information in digital libraries with sophisticated The benefits of e-learning in distance education is
search engines. Learner can browse required summaried below
material and can easily download into his/her system.
Overall the Digital Libraries will greatly support the e-
Learning environment.
• Minimize the distance between the learner
and teacher . e-learning allows people to
study from home.
Typical e-learning web content one normally finds on • e-learning provides a high quality and
efficient educations with minimum cost.
the web:
• e-learning provide education in 24X7 mode.
Read text: press button for next page.
Read text: make choice from numbered list; receive IV. Objects for e-Learning system and Use cases
score.
Read question: answer; get feedback; read next

Learner: End users of the system as shown in Use
question.
Case 1 of learner.
Read lots of text; answer questions at end.
Course Designer: Responsible for designing course
Take test immediately; learn score; get feedback. curriculum.
These examples are simply imitating the commonly Content writer: Responsible for writing the course
accepted notion of what education looks like-which materials for a particular course.
does not involve the way people really learn. “People
E-Course library/ Digital Library: Output of Course
really learn by doing a task they care about, failing
Designers and Content writers.
and redoing it until they get it right”.
Administrative Manager: He/She is in-charge of over
Object oriented system design using UML [2,3] is an
all administrative functions.
important research activity now-a-days. Different
steps of O-O design are design of classes, Briefly we describe the use cases as follows:
identification and construction of different objects,
and ultimately establishing object to object Use Case 1: Learner-Different Cases are Enrollment,
communication. View course, Browse, and Examination.
28
Use Case 4: E-Course Library/ Digital Library - Five

possible Use Cases are provided here, namely,
Login, course, Internet resource, non print media and
Enroll Update Library. Learners have to login to the system
first. Digital Library contains several course materials
in digital form may be available in CDs/Pen drives or
through Internet. Digital Library must be updated
Exam regularly.
Bro
inatio
Login Course
Vie
Use Case 1
Use Case 2: Course Designer- Four different cases

for course designer are Update course, setting Update Inte
question papers, updating model questions, monitor Library
rne
users ( learners) performances.
No
Update
Use Case 4
Monitor
Question for User
Examinat Use Case 5: Administrative Manager- Six different
Cases have been proposed for administrative
manager. These are Guideline to be given to the
learners, Form Design, supervise the course may be
Upd named as Course Monitoring, fees collection,
controlling examination activities, presenting results.
ate
Use Case 2
Guidel
Use Case 3: Content Writer- Two cases are related ine
to Content Writer. These are write contents of the
course and updating the contents as and when Form
needed. C
Design
Con
Update Fee
Collectio
Result
Exam
Control
Use Case 5
Use Case 3
29
V. Sequence Diagram Sequence Diagram demonstrates the dynamical

behaviour of objects in the system in a use case by
Lea E- Cont Cour Admi describing the communication among objects by
nistr message passing. In our proposed system content
VI. Conclusion:
E-Learning is an important component of E-

Education and education for all is possible if e-
Learning is implemented in distance education. O-O
implementation of E-Learning is simple and easy to
implement for software development point of view.
References
[1] A. M. Midda and S. Mukhopadhyay, “Information and

Communication Technology in E-Education”, Journal of Library
and Information Technology, Vol 2, No. 1, 2006 ,pp 42-47.
[2] S Banerjee, D E Booth, S Ghosh, S Mukhopadhyay, “A

prototype Design for Intellectual property Right Management in E-
Commerce-A UML Based Approach”, Journal Of Computer
Society of India, Vol 36, No 4. , 2006, pp 12-17.
[3] A.K Singh and S Mukhopadhya, “E-Learning in Distance

Education using UML”, Int. Journal of Information and Computing
System, Vol 12, No 2, 2009, pp 30-35.
[4] R. M. vental, K Prakash, “Introducing

Electronic Information Resources , through E- learning Mechanism
: A Study with Reference to Distance Education in its Environment”
University News, Vol.42(14), April 05-11,2004
[5] L. Vernal, M.U. Paily, “ICT in Teacher Education” , University

News , Vol.42 (39), September 27-03 October , 2004.
writer constructs the contents of digital library.

Learners log into the e-course library, login is verified
and a confirmation is passed to the learners. Then
learners can browse from the E-Library and access
tutorial sessions. If a learner does not find the desired
information, reports to the content writer. Then
content writer communicates with course designer for
update the library. Request may also come from
administrative manager from time to time to the
course designer for updating the library. Course
designer then updates the course and instructs the
content writer for updating. Content writer then
updates the e-Library and reports to the designer.
Course designer reports to the Administrative

Manager. Finally Administrative Manager informs the
learners about updation.
30
ON THE RELATIVE CHARACTER GRAPH

OF A FINITE GROUP
Dr.K.T.NAGALAKSHMI#1 Dr.A.V. JEYAKUMAR#2
Department of Mathematics Department of Mathematics
K.L.N.College of Information Technology Periyar Maniammai University
Pottapalayam-630 611,Sivagangai District Vallam,Tanjore District
Tamil Nadu, INDIA Tamil Nadu, INDIA
Abstract—This paper,in a sense, is a sequel to an For the basics of character theory of G, we shall
earlier construction by T.Gnanaseelan of a graph Γ (G, refer to [4] . The universal notations such as IrrG,
H) for any finite group G and a subgroup H using χ H, θ G and [ χ , ψ ] will stand for the complete set
(complex) irreducible characters (See [3]). We
of irreducible characters of G, the restriction of a
construct another graph Ω (G, H), which is structurally character χ of G to a subgroup H, the induction of a
quite different from Γ (G,H). However, we prove that
character of H to G and the scalar product
for the special case of the sequence of subgroups Sn-1
⊂ Sn ⊂ Sn+1 (where Sn is the symmetric group on n 1
χ (s) ψ (s −1 )
letters), Γ (Sn, Sn-1) and Ω (Sn+1, Sn) are indeed O (G )
∑
S ∈G
(of course, all
isomorphic.
representations are finite dimensional taken over the
/ ).
complex field C
Key words: Character Theory, Graph Theory, Group
Theory, and Representation Theory.
I. INTRODUCTION 1) Definition : The relative character graph Γ (G,

From the time of R.Brauer, various finite graphs H) of G with respect to a subgroup H has IrrG as its
have been constructed using mostly irreducible vertex set and two distinct χ , ψ in IrrG are adjacent
characters (both complex and p – modular) of a finite if and only ifχ H and ψ H have atleast one element of
group G. These graphs in general give a pictorial IrrH in common. This is equivalent to saying that [ χ ,
representation of the intricate nature of irreducible
ψ ]H > 0. Clearly Γ (G,H) is a simple graph in graph -
characters of G. For instance, the famous Brauer
graph has as vertex set the full set of complex theoretic sense, that is, it has no double edges and
irreducible characters of G and two distinct vertices self-loops.
are incident if and only if their reduction mod p To begin with the following easy observations can
(where p is a prime dividing O (G)) contains atleast
be made.
one p-modular irreducible character in common.
2) Γ (G, H) is the null-graph if and only if H = G.
Quite recently, T.Gnanaseelan in his Ph.D. thesis
[17] has constructed a new finite graph for any 3) If H is the trivial subgroup, then Γ (G,H) is a
subgroup H of G, which he calls the relative character complete graph, but even for certain types of non-
graph of G over H and denotes by Γ (G, H). In the trivial subgroups H, Γ (G,H) can be complete. For
next section, we shall define Γ (G,H) and recount instance, if H is a cyclic group generated by x and if
some of the salient properties as proved in [ 17] . all the matrices ρ i (x), where ρ i runs through a full
set of inequivalent irreducible representations of G,
have 1 as eigen value, then Γ (G, H) turns out to be
II. THE GRAPH Γ (G, H) complete. (This can be of some interest in the
representation theory of finite Chevalley groups).
31
4) If H and K are two subgroups of G such that K namely {1}, { θ 2, θ 3} and { θ 4}. Let 1G, χ 2, χ 3, χ 4
⊂ H, then Γ (G,H) is a subgraph of Γ (G,K). and χ 5 be the irreducible characters of S4 with
5) If, x ∈G then Γ (G, H) = Γ (G, Hx), where Hx degrees 1, 1, 2, 3 and 3 respectively.Then 1GH = 1G +
is the conjugate of H under x. χ 2, θ G2 = θ 3G = χ 3 and θ G4 = χ 4 + χ 5. We
6) Γ (G, H) is connected if and only if core GH = conclude that Ω (L,G) has three connected
(1). components, namely, {1G, χ 2}, { χ 3} and { χ 4, χ 5}.
Naturally, Γ (G,H) and Ω (G,H) are not
7) A connected graph Γ (G, H) is a tree if and isomorphic.
only if G is a Frobenius group NH, and the kernel N is
a unique elementary abelian normal p-Sylow We do not propose to study the graph Ω (G,H)
subgroup for some prime p with order pm and O(H) = systematically here such as the connectivity
m
P - 1. properties etc, as was done by Gnanseelan for his
graph Γ (G,H). However, taking cue from the last
example, we make some beginning which partially
compares with results of Gnanseelan mentioned in
III. THE GRAPH Ω (G, H).
[7].
We shall now construct another finite graph, again
2) Proposition : If H is a normal subgroup of G,
with reference to a subgroup, which, in a sense, will
then Ω (G,H) is disconnected. (If H is trivial Ω
be dual to the graph Γ (G, H).
(G,H), is just a dot, whereas Γ (G, H) is complete).
1) Definition : The vertex set of Ω (G, H), is IrrH
Proof : Let O1, O2, . . . ., Os be the complete set of
and two distinct θ and φ are adjacent if and only if distinct orbits of IrrH under the conjugate action of G
the induced characters θ and φ have atleast one
G G
an IrrH. Then by Clifford’s theorem it is clear that two
element of IrrG in common,in other words[ θ G, φ G]G distinct θ, φ, ∈ IrrH are adjacent in Ω (G,H) if and
> 0. The structural properties of Ω (G, H) differ in only if θ and φ lie in the same orbit Oi for some i.
many ways to those of Γ (G, H). To begin with, even Since H ∆ G, s >1 and hence Ω (G,H)
the vertices sets of both the graphs are different. is disconnected.
But there is one situation wherein the vertices sets Infact, Ω (G,H) has exactly s components each of
of both Γ (G, H) and Ω (G, H) coincide. This which is complete.
occurs when we have a sequence of groups and
subgroups of the form H ⊂ G ⊂ L. Here the vertices We shall push a little further in the compare and
sets of both Γ (G, H) and Ω (L,G) coincide. But contrast syndrome vis-à-vis Γ (G, H) of Gnanaselan.
these graphs need not be the same, as can be seen
from the following example. Consider the sequence H As per (1.3), Γ (G, (1) ) is complete But in Ω
⊂ G ⊂ L , where, L is the symmetric group S4, G is (G,H) completeness can be achieved in quite
the Alternating subgroup A4 and H is the subgroup different situations. First we shall recall Mackey’s
consisting of the two elements {(1), (12)(34)}. Since subgroup theorem.
core GH = maximal normal subgroup of G contained
3) Theorem (Mackey [1]) : If H is a subgroup of
in H is trivial, using the criterion for connectivity
obtained in [3], Γ (G,H) is a connected graph. Using G and if θ, φ, ∈ IrrH, then (θ ,
G
φ G) =
Clifford’s theorem, it is quite easy to see that Ω ∑ (θ, x φ) H ∩ x , where D is a set of double coset
H
(L,G) is not connected. To see this, let 1H, θ 2, θ 3 x∈D
and θ 4 be the four distinct irreducible characters of representatives of H in G.

A4 of degrees 1, 1, 1 and 3. The conjugate action of
We now prove the following theorem providing a
S4 on IrrA4 breaks these 4 characters into 3 orbits,
sufficient condition for the completeness of Ω (G,H)
32
4) Theorem ; Let H be a subgroup of G such that upper triangular matrices, (q = pn for a fixed prime p)
for some x ∈D , the set of (H, H) - double coset and also G = PSL (2, q) and B = Borel subgroup.
representatives of H, H ∩x H = (1). Then Ω (G,H) For a brief reference of the representation theory of
G, we refer to [1]. We straightaway go to the
is complete. following.
Proof : By Mackey’s subgroup theorem θ G, φ G = 7) Proposition ( [1 ], p.22): Let φ, θ ∈ , IrrB.

∑ (θ, φ) H ∩
x
x
H
for any two θ, φ, ∈ IrrH,. By Then [φ
G
, φG ] = | w ∈ W | φ w = φ | Where W is
x∈D
the weyl group of G and φ w
denotes the characters
assumption, H ∩x H = (1) for some x ∈D. Hence
obtained by the conjugate action of w on φ.
(θ, φ )(1) > 0, which implies ( θ G, φ G) > 0. Since this
x
is true for any two distinct θ, φ, ∈ IrrH, it follows 8) Proposition: When G = GL (2, q), the graph
that Ω (G,H) is complete. Ω (G, B) is connected.
5) Corollary : If H is non-normal and prime cyclic, Proof : As already mentioned, W consists of just
two elements namely, 1 and the element =
then Ω (G,H) is complete.
0 1
Proof is clear. σ =   (7) gives all the non-principal
1 0
6) Corollary : If G = NH is a Frobenius group characters θ m, n of Ω which can be pulled back to
with complement H, then Ω (G,H) is complete. give all non-principal irreducible characters of B. It
can be easily seen that θσ m , n = θ n ,m . (Note
Proof : By definition, H ∩a H = (1) for all a ∉H .
The result now follows from the above theorem.
that only when m ≠ n the pull backs are irreducible.)
Having made the above observations, it is perhaps

Hence by (7)
worthwhile to point out the significance of both Γ (G,
H) and Ω (G, H) both graph theoretically and group-
( θm , n , θ n , m ) = |w ∈W| θ mw ,n θ n ,m |
G G
= =1
theoretically. As was mentioned earlier, Gnanaseelan
has highlighted many graph – theoretic properties of
Γ (G, H). Further study of Γ (G,H) in graph –
theoretic aspects such as colouring domination etc
Hence (T( θ m , n ), T ( θ n , m ) is non zero for all non -
may throw more light on the study of character theory
of G itself. It is interesting to note that the vertex sets principal irreducible characters of B and therefore by
of Γ (G,H) and the Brauer graph are one and the the definition of the graph Ω (G, B) all the non-
same and therefore comparisons are possible. principal characters of B are adjacent. Since 1GB must
Turning to our Ω (G,H) we observe that this graph have some irreducible other than 1G we have Ω (G,
has a strong bearing on the character theory of B) is connected.
algebraic groups and the related finite groups. The n
The case G = PSL (2, q), (q - p for a prime p), the
reasons are obvious. The study of the character
Projective Special Linear Group of rank 1. The
theory of these groups heavily rests on the so-called
subgroup that we choose is as before the Borel
‘Harish – Chandra Induction’ and all the fascinating
subgroup B - TU, where T consists of the 2 x 2
theory governing these aspects can be naturally fitted
matrices of determinant 1 and U denotes a p -
graph theoretically into our graph Ω (G,H). We
Sylow subgroup (the unipotent subgroup). We have
propose to take up this study systematically in our
future works, but for the moment, we shall briefly q −1
O (T) = when p is odd and q-1 when p is 2,
outline the construction of Ω (G,B), where G = GL 2
(2,q) and B = the Borel subgroup, the subgroup of
33
q (q 2 − 1) I (1B) = {1G, χ 2 }, (1)

O(U) = q and O (G) = when q is odd and
2
q(q2-1) when p is 2. I ( α1 ) = { χ 2 , χ 3 , {ψ j } q −3 , { ψ 'k } q −3 } (2)
1≤ i ≤ 1≤ k ≤
2 2
For the complete table of (Complex) irreducible
representations of G, we refer to [2] (of course, we I ( α2 ) = { χ2 , χ 4 , {ψ j } q −3 , { ψ 'k } q −3 } (3)
1≤ j≤ 1≤ k ≤
should pick out the irreducibles of G from this table, 2 2
which is quite easy). Now B is a Frobenius group with
kernel U and complement T. q −1
I( φ i) = ψj (for some j) (2 ≤ i≤ ) (4)
We should distinguish two cases now, the p odd 2
case and the p = 2 case.
We have now proved the following theorem,

I. The odd prime case. whose proof can be gleaned from the equations (1),
(2), (3) and (4).
Case 1. q ≡ -1 (mod 4)
9) Theorem : The graph Ω (G,H) is
q+5 connected.We shall draw the graph Ω (G,B) for the
In this case, these are distinct case p = 7.
2
irreducible cases of G which are denoted as follows :
1B
χ1 (= the trivial character), χ 2 (= the
Steinberg character of degree q), χ 3, χ 4, the
q −1 q −3
irreducible characters of degree , the
2 4
irreducible characters ψ1 , ψ 2 , .... ψ q −3 each for
4 α1 α2
q −3
degree q+1 and the irreducible characters
4
ψ '1 , ψ ' 2 , .... ψ ' q −3 each of degree q-1.
4
From the character theory of

Frobenius groups, we can easily see that B has
φ1 φ1
q+3
exactly irreducible characters denoted by
2
{Φ i } q −1 ( φ 1 = trivial character ≠ 1B), α 1
1≤ i ≤
2 Case 2 : q ≡ 1 (mod 4)
q −1
and α 2 each of degree . q+5
2 The irreducible characters of G are
2
We can very easily calculate the character denoted by χ 1, χ 2 (= Steinberg), χ 3, χ 4, of
θ G
, for any θ ∈ IrrB . We write down the induced q +1 q −5
degrees , the irreducible characters
covers I ( θ ) as follows : 2 4
34
ψ 1, ψ 2, . . . . , ψ q −5 , each of degree q+1 and the

4
IV. THE GRAPHS Γ (G,H) AND Ω (G, H) FOR G – THE
q −1
characters ψ '1 ,... ψ q −1 each of degree q-1. SYMMETRIC GROUP SN.
4 4
We have already noted in (2) that even in the
The irreducible characters of B in this case also special case of groups given by the sequence H ⊂
G ⊂ L, Γ (G,H) and Ω (L,G) need not be
are 1B, α 1, α 2 and {φ i } q −1 .
1≤ i ≤ isomorphic. However, it is interesting to note that in
2
the case of the sequence of groups given by Sn-1 ⊂
The induced covers in this case are given below. Sn ⊂ Sn+1, Γ (Sn, Sn-1) and Ω (Sn+1, Sn) are
indeed isomorphic. The rest of the paper deals with a
proof of this statement.
I (1B) = {1G, χ 2 }, (1) It is well known that the study of irreducible

representations of Sn can be made through partitions
I ( α 1) = { χ 2 , χ 3 , {ψ j } q −5 , { ψ 'k } q −1 } (2) of n.
1≤ i ≤ 1≤ k ≤
4 4
1) Partitions of N :
A decreasing sequence of positive integers

i.e λ ≥ λ ≥ . . . ≥
λ = ( λ 1, λ 2, . . . ., λ r)
I ( α 2) = { χ 2 , χ 4 , {ψ j } q −5 , { ψ 'k } q −1 } (3) 1 2
1≤ j≤
4
1≤ k ≤
4 λ r ≥ 1 is called a partition of n if λ 1+ λ 2+ ....+ λ r =
n. We denote this by the symbol λ λ n The
i.e
λ i’s are called the parts of the partition λ , and the
q −1 integer r is called the number of parts or length of λ .
I( φ i) = ψj (for some j) (2 ≤ i≤ ) (4)
2
Two partitions λ = ( λ , . . . . ., λ ) and =
1 r
µ (µ 1 , . . . .
, µ s) of n are said to be equal if r = s and λ 1 = µ i
We have
10) Theorem : The graph Ω (G, B) is connected. for all i.
We shall draw the graph Ω (G, B) for the case when We shall omit several details regarding the well-
p = 5. known theory of representations via partitions as
developed by Frobenius and Schur and just state the
1B basic facts needed in the sequel.
2) The set of all conjugacy classes of Sn is

naturally bijective with the set of all partitions of n.
Since the set of all conjugacy classes of Sn is
naturally bijective with the set of all complex
α1 α2 irreducible characters of Sn, it immediately follows
that the set of all complex irreducible characters of
Sn is naturally bijective with the set of all partitions of
n.
3) (Frobenius - Schur) : λ of
To every partition
φ1
n, one can associate an irreducible module V λ for
Sn. The family (V λ ) λ n is a complete set of
35
mutually inequivalent irreducible representations of Proof : Let [ λ ] ↑ and [ µ ] ↑ have a common

Sn over C/. irreducible constituent. (Note that we have simplified
the notation, since there is no danger of confusion).
4) The Branching Theorem ( [ 5 ]): If λ is a
partition of n, we shall denote by [ λ ] the unique (i) Suppose that |r-s| > 1.
irreducible representation V λ associated to λ . We
Assume that r = s + k, k > 1. That is, λ has
shall simultaneously denote the irreducible (complex)
s + k parts and µ has s parts with k > 1. By
character associated to V λ also by [ λ ]. Given a
Branching Theorem, the partition corresponding to
subgroup H of a finite group G, and given θ ∈ IrrH
any irreducible constituent of [ λ ] ↑ will have atleast
and χ ∈ IrrH respectively, we shall denote by θ ↑ GH s + k parts and the partition corresponding to any
the induced character of θ to G and by χ ↓GH the irreducible constituent of [ µ ] ↑ will have atmost s+1
restriction of χ to H. parts. But s + 1 < s + k as k > 1. Hence no
irreducible constituent of [λ]↑ equals any
5)Branching theorem for Sn.
irreducible constituent of [ µ ] ↑ , which proves (i).
Let S ⊆ S ⊆ S in a natural way. Given
(ii) Assume that | λ - µ | > 1 for some i, 1 ≤i≤
λ = ( λ 1, . . . . , λ r ) n, let λ j ± denote the partition
n-1 n n+1
r.
i i
( λ 1,...., λ j ± 1,..... λ r) of n ± 1 according as λ

j-1
> First let r = s. Suppose µ i - λ i > 1.
λ j or λ j > λ j+1.
Let λ = ( λ ,. . . . ., λ ,. . . ., λ ) and
1 i r
Then we have
µ = ( µ , ..... λ + k , ....., µ r), ki > 1.
[λ]↓ ⊕ j [λ ] and
Sn j− 1 i i
Sn −1 =
λ j > λ j+1 Then, in the partitions corresponding to each
irreducible constituent of [ λ ] ↑ , the ith part is either
[ λ ] ↑Snn +1 ⊕ j [λj+ ]
S
=
λ j−1 > λ j
λ or λ +1 whereas the ith part of the partitions
i i
corresponding to each irreducible constituent of
6) Lemma : [ µ ] ↑ contains entries which are atleast λ i+ ki.
Let λ = ( λ 1, . . ., λ r) and µ = ( µ 1, . . . ., Hence no two irreducible constituents of [ λ ] ↑ and
µ s) be two partitions of n. Then [ λ ] ↑

Sn +1
and λ [ µ ] ↑ will coincide.
Sn
have a common irreducible constituent if and only if

Similar argument holds when λ -
i
µi > 1
i) |r-s| ≤ 1, for some i. When r = s + 1, λ r = 1 and µ r = 0
ii) | λ - µ | ≤ 1 for every i (taking µ s+1 = 0 when and when s = r + 1, µ s = 1 and λ s = 0, we can
i i
prove (ii) using similar arguments, which completes
r = s+1 and λ r+1
= 0 when s = r +1 the proof of (ii).
respectively.
(iii) Suppose that | λ -
i
µ i |=1for k i’s, k > 2. Since
iii) |λ -µ | = 1 for exactly two distincts i’s
i i
Σλ i = Σµ i = n, if λ i = µ i -1, then there exists
= 0 for the remaining i’s. some j
36
such that λ =
j
µ j +1 to maintain parity. By µ j. Therefore, ( λ 1, . . . . , λ j-1, µ j +1 . . . . . . . , λ k +1, . .
.. λ s) is a constituent of [ µ ] ↑ . Again λ k-1 > λ k.
Branching theorem, only one part of [ λ ] and [ µ ] will
Therefore ( λ 1, . . . . λ j-1, µ j +1 . . . . λ k-1, λ k+1, . . . .
be
increased by at a time, which constitute the parts of

λ s) is constituent of [ λ ] ↑ as well. This completes
the proof of the lemma.
[λ]↑ and [µ ]↑ respectively. Therefore no 7) Lemma : Let λ = (λ , . . . . λ r) and µ = (µ ,

1 1
irreducible
µ s) be partitions of n. Then [ λ ] ↓
Sn
. . . . , Sn −1 and
constituent of [ λ ] ↑ will coincide with any irreducible [ µ ] ↓ Snn −1 have a common irreducible constituent if
S
constituent of [ µ ] ↑ . That is, [ λ ] ↑ and [ µ ] ↑ will

and only if
not i) |r-s| ≤ 1,
have a common constituent. Using similar ii) |λ -µ |

i i
≤ 1 for every i (taking µ s+1
= 0 when
arguments
r = s+1 and λ = 0 when s = r + 1
r+1
we can prove the result when µ i = λ i - 1. Hence the respectively).
result. Conversely, let the three conditions hold. iii) | λ - µ | = 1 for exactly two distinct i’s
i i
Case (i) : Let | r – s |=1. It is enough to prove the = 0 for the remaining i’s.
result for r = s+1. Then λ = ( λ , . . . . λ ) and µ = ( µ ,
1 s+1 1
. . . . . , µ s). We can take µ s+1 = 0 and hence by Proof :Let [ λ ] ↓ and [ µ ] ↓ have a common
(ii) λ = 1. By iii) there exists some j such that irreducible constituent (as before, we have simplified
s+1
the notations).
µj = λ +1 and for all other i, λ =
j i
µ i. That means,
(i) Suppose that |r-s| > 1.
µ = (λ , . . . . . λ
1 i-1
, λ , . . . . λ , 1 ) and µ =
i s
Assume that r = s+k, k>1. By Branching
(λ , . . . . λ j-1 , λ j+1 , . . . . λ s, 0). Now, theorem, the partition corresponding to any
1
λ j−1 ≥ λ j + 1. That is λ j−1 > λ j . Therefore irreducible constituent of [ λ ] ↓ will have atleast s+k-
1 parts and the partition corresponding to any
( λ 1, . . . ., λ j+1, . . . λ s, is a constituent of [ λ ] ↑
clearly, replacing 0 by 1 in irreducible constituent of [ µ ] ↓ will have almost s
parts. Since k > 1, s + 1 < s + k. Hence no
1) [ µ ] ↑ we get ( λ , . . . ., λ j-1, λ j+1, . . . . irreducible constituent of [ λ ] ↓ is equal to any
1
λ s, 1) irreducible constituent of [ µ ] ↓ which proves (i).
2) is a constituent of [ µ ] ↑ as well.
Case (ii) : (ii) Assume that | λ - µ | >1 for some i, 1 ≤i≤ r.
i i
r = s. Then λ = ( λ 1, . . . . , λ s) and = ( µ 1, . . . . , µ s). By Let r = s Suppose µ i - λ i > 1. Let λ = ( λ 1, . . . .,

(iii), there exists some j and k such that λ j = µ j+1 λi , . . . . , λ r ) and µ = ( µ 1, . . . . . µ i + ki, .....,
and µ k = λ k +1 and λ j = µ i for all other i’s. That is, µ r), k > 1. Then in the partitions corresponding to
i
λ = ( λ 1, . . . . , λ j-1, µ j + 1, . . . , λ k, . . . . . , λ s) and each irreducible constituent of [ λ ] ↓ , the ith part is
µ = ( λ 1, . . . . , λ j-1, µ j, . . . . , λ k +1, . . . . λ s) Then λ j-1 > either λ or λ -1 whereas the ith part of the
i i
37
partitions corresponding to each irreducible Therefore, ( λ , . . . . , λ , λ , . . . .0) is a constituent

1 j-1 j
[ µ ] ↓ contain entries which are
of [ µ ] ↓ clearly the above is a constituent of [ λ ] ↓
constituent of
atleast λ i + ki - 1. But ki > 1. as well.
Case (ii) :
Hence no two irreducible constituents of [ λ ] ↓ Let r = s. Then λ = ( λ 1, . . . . , λ s) and µ = (µ , . . . . ,

1
and [ µ ] ↓ will coincide. µ s). By (iii), there exists some j and k such that
λ j = µ j-1 and µ k = λ k - 1 and λ i = µ i for all other
Similar argument holds when λ i - µ i > 1 for some
i’s. That is, λ = ( λ , . . . . , λ , µ - 1, . . . , λ k, . . . . .,
1 j-1 j
1. The cases r = s +1 and s = r + 1 can be dealt with
similarly, completing the proof of ii). λ s) and µ = ( λ 1, . . . . , λ j-1, µ j, . . . . , λ k -1, . . . . λ s)
Then λ j-1 > µ j. Therefore, ( λ 1, . . . . , λ j-1- 1, µ j
(iii) Suppose that | λ - µ i |=1 for k i’s, k > 2.
Since
i
Σλ i = Σ µ i = n, if λ i = µ i + 1, then ......., λ k -1, . . . . λ s) is a constituent of [ µ ] ↓ .
there exists some j such that λ = µ -1 to
Again λ k-1 > λ k. Therefore ( λ 1, . . . . λ j-1, µ j . . . .
j j
maintain parity. By Branching theorem, only one λ k-1 -1, λ k, . . . . λ s) is a constituent of [ λ ] ↓
part of [ λ ] and [ µ ] will be increased by 1 at a time, also.
which constitute the parts of [ λ ] ↓ and [ µ ] ↓ This completes the proof of the lemma. We are
respectively. Hence if k > 2, there will be parts now in a position to prove our main theorem.
say λ , µ in [ λ ], [ µ ] respectively (k ≠ 1,2) such
k k 8)Theorem : The two graphs Ω (Sn+1, S ) and
that λ ≠ µ k and these parts will not be touched
n
k Γ (Sn, Sn-1) are isomorphic.
upon restriction. Thus in each part of [ λ ] ↓ and
[ µ ] ↓ , there will be some λ ≠ µ k. That is, [ λ ] ↓ Proof :Let [ λ ] and [ µ ] belong to IrrSn and let
k
and [ µ ] ↓ will not have any common constituent. [ λ ] ↑ and [ µ ] ↑ contain an irreducible character of
Similarly we can prove this result when µ i = λ i-1. S

n+1
in common. Then by Lemma 5 and Lemma 6
Hence the result. [ λ ] ↑ and [ µ ] ↑ contain an irreducible character of

Sn-1 in common. Again by the two lemmas, the
Conversely, let the three conditions hold.
converse statement also holds. Hence Γ (S , S )
n n-1
Case (i) :
and Ω (S
n+1
, S ) are isomorphic graphs.
n
Let | r – s | =1. It suffices to prove the result for
r = s+1. Then λ = ( λ , . . . . λ ) and µ = ( µ , . . . . ., V.CONCLUSION
1 s+1 1
µ s). We can take µ s+1 = 1 and hence by (ii) The graphs Γ (G, H) and Ω (L,G) are generally
λ s+1 = 0. By iii) there exists some j such that different for H ⊂ G ⊂ L, as we observed earlier.
Taking the cue from our result for Sn, it would be
µj = λ -1 and for all other i, λ =
j i
µ i. That is,
interesting to find conditions (in general) on H, G and
λ = ( λ 1, . . . . . λ i-1 , λ i , . . . . λ s, 1 ) and L so that Γ (G,H) and Ω (L,G) are isomorphic.
µ = ( λ 1, . . . . . λ j-1 , λ j-1 , . . . . λ s, 0). Now,
Also a deep study of Ω (G,H) in the case of
λ j−1 ≥ λ j - 1. since otherwise in λ = ( λ 1, . . . . λ j-1, algebraic groups and suitable subgroups may throw
more light on the representation theory of such
λ j, . . . .), contradiction. Hence λ j-1 > λ j.
groups.
38
[9] A.Turull,“ The Schur Indices of the Irreducible Characters of the

Special Linear Groups,” J.Algebra., 235(1), pp. 275-314,2001.
REFERENCES
[10] M.Guralnick,r, K.Magaard, J.Saxl,, and P.H.Tiep,“ Cross
Characteristic Representations of Sympletic and Unitary Groups,”
[1] C.W.Curtis,, and I.Reiner,“Methods of Representation Theory,”
J.Algebra., 257(2),pp.291-347,2202.
Volume I., Wiley Interscience Publications., Newyork,1981.
[11] R.A.Ferraz, “Simple Components and Central units in Group
[2] i.M.Isaacs,“Character Theory of Finite Groups,” Academic
Algebras,” J.Algebra., 279(1), pp. 191-203,2004.
Press., New York,1976.
[12] G.James,and A.Kerber, “The Representation Theory of the
[3] C.Musili,“Representations of Finite Groups,” Hindustan Book
Symmetric Group” J.Algebra., 284(1),pp. 326-332,2005.
agency., New Delhi,1993.
[13] M.Guralnick,r, K.Magaard, J.Saxl,, and P.H.Tiep,“ Cross
[4] J.P.Serre,J “Linear Representations of Finite Groups,” Springer
Characteristic Representations of odd Characteristic Sympletic
Verlag., Newyork,1971.
Groups and Unitary Groups,” J.Algebra., 299(1), pp. 443-
446,2006.
[5] Narsingh Deo, “Graph Theory with applications to Engineering
and Computer science,”Prentice-Hall,1974.
[14] F.Szechtman, “Irreducible Characters of Sylow Subgroups of
Symplectic and Unitary Groups,” J.Algebra., 303(2), pp. 772-
[6] F.Harary,F., 1988, “Graph Theory,” Narosa Publishing House,
730,2006.
New Delhi.
[15] I.M.Isaacs,“Character Kernels and Degree Ratios in Finite

[7] N.Chigira,,Y.Takegahara, andH.Yamaki, “On the number of
Groups,” J.Algebra., 322(6), pp. 2220-2234.2009.
homomorphisms from a finite group to a general linear group,”
J.Algebra., 232(1), pp, 236-254,2000.
[16] H.Fukushima, “Irreducible Products of Characters of Solvable
Groups,” J.Algebra., 321(1), pp. 312-315,2009.
[8] H.Mitsuhashi, “The q-analogue of the Alternating group and its
Representations,” J.Algebra., 240(2), pp. 535-558,2001.
[17] T.Gnanaseelan, “Studies in Algebraic groups and
representations,” Ph.D.Thesis,Madurai Kamaraj University,1999.
39
DETECTION OF DIABETIC RETINOPATHY

USING RADIAL BASIS FUNCTION
Mr. R. Vijayamadheswaran#1, Dr.M.Arthanari#2, Mr.M.Sivakumar#3
#1
Doctoral Research Scholar,Anna University, Coimbatore.
#2
Director,Bharathidhasan School of Computer Applications, Ellispettai, Erode.
#3
Doctoral Research Scholar, Anna University, Coimbatore.
Abstract — Retinal exudates classification and distribution of exudates and microaneurysms and
identification of diabetic retinopathy to diagnose the eyes hemorrhages[3], especially in relation to the fovea
using fundus images requires automation. This research can be used to determine the severity of diabetic
work proposes retinal exudates classification. retinopathy
Approach: Representative features are obtained Hard exudates are shinny and yellowish
from the fundus images using contextual clustering (CC)
intraretinal protein deposits, irregular shaped, and
segmentation methods. The number of features obtained is
found in the posterior pole of the fundus [4]. Hard
two. The radial basis function (RBF) network is trained by
the features. Final weights are obtained and subsequently exudates may be observed in several retinal vascular
used for testing. pathologies. Diabetic macular edema is the main
cause of visual impairment in diabetic patients.
Results: The presence of exudates is identified Exudates are well contrasted with respect to the
more clearly as the CC uses neighbourhood information. background that surrounds them and their shape and
By knowing the outputs of RBF during testing, accurate size vary considerably [5]. Hard and soft exudates
diagnosis and prescription for treatment of the affected
can be distinguished because of their color and the
eyes can be done. One hundred fundus images are used
sharpness of their borders. Various methods have
for testing. The performance of RBF is 96%(48 images are
classified). been reported for the detection of Exudates. Efficient
algorithms for the detection of the optic disc and
Conclusion: Simulation results show the retinal exudates have been presented in [6][7].
effectiveness of RBF in retinopathy classification. Very
large database can be created from the fundus images Thresholding and region growing methods
collected from the diabetic retinopathy patients that can be were used to detect exudates [8][9], use a median
used for future work filter to remove noise, segment bright lesions and
dark lesions by thresholding, perform region growing,
Keywords: Diabetic retinopathy, fundus image,
then identify exudates regions with Bayesian,
exudates detection, radial basis function, contextual
Mahalanobis, and nearest neighbor (NN) classifiers.
clustering
Recursive region growing segmentation (RRGS).[10],
have been used for an automated detection of
I. INTRODUCTION diabetic retinopathy Adaptive intensity thresholding
and combination of RRGS were used to detect
Diabetic Retinopathy (DR) cause blindness exudates,[11], [12], combine color and sharp edge
[1]. The prevalence of retinopathy varies with the age features to detect exudate. First they find yellowish
of onset of diabetes and the duration of the disease . objects, and then they find sharp edges using various
Color fundus images are used by ophthalmologists to rotated versions of Kirsch masks on the green
study eye diseases like diabetic retinopathy [2]. Big component of the original image. Yellowish objects
blood clots called hemorrhages are found. Hard with sharp edges are classified as exudates.
exudates are yellow lipid deposits which appear as
bright yellow lesions. The bright circular region from
where the blood vessels emanate is called the optic
disk. The fovea defines the center of the retina, and
is the region of highest visual acuity. The spatial
40
segmentation is typically used to locate data in a

vector. The result of contextual data segmentation is
a set of regions that collectively cover the entire data.
Each value in a data is similar with respect to some
characteristics. Adjacent regions are significantly
different with respect to the same characteristics.
Several general-purpose algorithms and techniques
have been developed for data segmentation.
Contextual clustering algorithms segments a data
into one category (ω0) and another category (ω1).
The data of the background are assumed to be
drawn from standard normal distribution.
II. MATERIALS AND METHODS 1. Define decision parameter Tcc (positive) and
weight of neighbourhood information β
This research work proposes contextual (positive). Let Nn be the total number of data
clustering (CC) and Radial Basis Function (RBF) in the neighbourhood. Let Zi be the data.
network. CC is used for feature extraction. The 2. Initialization: classify element of data with
extracted features are input to the RBF network. In Zi>Tcc to ω1 and element of data to ω0. Store
the classification to C0 and C1.
order to achieve maximum percentage of
3. For each element of data ‘i’, count the
identification of the exudates, proper data input for number of data ui, belonging to class ω1 in
RBF, optimum topology of RBF and correct training the neighbourhood of data ‘i’. Assume that
of RBF with suitable parameters is a must. the element of data outside the data area
belong to ω0.
A large amount of exudates and non 4. Classify element of data with
exudates images are collected. Features are β N
extracted from the images using contextual clustering zi + (u i − n ) > Tα to ω1 and other
segmentation. The features are input to the RBF and Tcc 2
labeling is given in the output layer of RBF. The element of data to ω0. Store the classification
labeling indicates the exudates. The final weights to variable C2.
obtained after training the RBF is used to identify the 5. If C2 ≠C1 and C2 ≠ C0, copy C1 to C0, C2 to C1
exudates. Figure 1 explains the overall sequence of and return to step 3, otherwise stop and
proposed methodology. return to C2.[13]
IV.RADIAL BASIS FUNCTION

III.CONTEXTUAL CLUSTERING
Radial basis function neural network (RBF) is
Image segmentation is a subjective and a supervised neural network. The network has an
context-dependent cognitive process. It implicitly input layer, hidden layer (RBF layer) and output layer.
includes not only the detection and localization but The 2 features obtained are used as inputs for the
also the delineation of the activated region. In network and the target values for training each
medical imaging field, the precise and computerized exudate is given in the output layer.
delineation of anatomic structures from image data
sequences is still an open problem. Countless Training RBF is done as follows:
methods have been developed, but as a rule, user
1. Finding distance between pattern and
interaction cannot be negated or the method is said centers .
to be robust only for unique kinds of images. 2. Creating an RBF matrix whose size will be
(np X cp).(Figure 2) , where np = number of
Contextual segmentation refers to the pattern used for training and cp is number of
process of partitioning a data into multiple regions. centers which is equal to 10. The number of
The goal of segmentation is to simplify and / or centers chosen should make the RBF
change the representation of data into something that network learn the maximum number
is more meaningful and easier to analyze. Data oftraining patterns under consideration.
41
3. Calculate final weights which is inverse of

RBF matrix[14][15][16][17][18] multiplied with
Target values.
4. During testing, the performance of the RBF Fig.3 Hardexudates
network, RBF values are formed from the
features obtained from CC and processed
with the final weights obtained during
training. Based on the result obtained, the
image is classified to have a type of exudate
or not.
V. EXPERIMENTAL WORK
Color retinal images obtained from

Aravind Hospitals, Madurai (India). According to
the National Screening Committee standards, all
the images are obtained using a Canon CR6-45
Non-Mydriatic (CR6-45NM) retinal camera. A
modified digital back unit (Sony PowerHAD
3CCD color video camera and Canon CR-TA) is
connected to the fundus camera to convert the
Figure 2 shows sample images of eyes in good
fundus image into a digital image. The digital
condition. Figure 3 shows sample images of eyes
images are processed with an image grabber
with hard exudates.
and saved on the hard drive of a Windows 2000
based Pentium -IV.
The Sample images of normal (Figure 2) VI. RESULTS AND DISCUSSION

and abnormal types (Figure 3)are given.
For template matching and comparison

Fig. 2 Normal fundus images purposes, representative exudates are isolated
from the original retinopathy images in order to
create exudates templates which are presented
in
Figure 4.
Fig. 4: Segmented pictures of hard exudates
(a)Sample Hard Exudates
42
Figure 4a shows the sample templates out of one

hundred templates collected. Each template has
varied scattering of the exudates. Figure 4b shows ,
the segmented exudates by the CC method. The
(b) Segmented hard exudates by CC black indicates the background of the image and the
white shows the hard exudates. CC does effective
segmentation. Statistical features for the hard
43
exudates templates are found. The statistical features Fig.5c Plane 2 of the image in Figure 5a
considered are ‘Convex Area’, ’Solidity’,’ Orientation’
and ‘Filled Area’. In Fig. 5a, different stages of
outputs of CC are given. The entire image processing
included here involves normalizing, histogram
equalization, segmentation.
Fig.5a A portion of the original true color image
Fig.5d Plane 3 of the image in Figure 5a
Figure 5a presents a portion of the original

diabetic retinopathy image in true color. The
Fig.5e Plane 2 of the image segmented by using
plane-1 information of the original image is
CC
shown in Figure 5b. The plane-2 (Figure 5c) and
plane-3(Figure 5d) are shown. Identification of
exudates is done using plane-2 information.
Fig.5b Plane 1 of the image in Figure 5a
The hard exudates are found scattered in

the retinopathy image. The segmented image of CC
shows more noise. Noise is present in CC
segmented image. Figure 6a presents 9 pixel values
summed versus the window number during scanning
the image to be segmented. The average summed
number is above 1500 which is an indication of slight
white background appearance as can be seen from
Figure 5c (plane 2).
44
Fig.6a Total pixel values an object is greater than a value of 500, then it is
2500
treated as optic disc. Using the boundingbox concept,
this object is filled with black. Hence the remaining
2000 objects could be either the noise or the exudates.
Figure 7 shows the edydisc removed by applying
Summation of the window
1500 statistical features.
1000
Fig.7 Eyedisc removed
500
0
0 2 4 6 8 10 12
3 X 3 window position 4
x 10
Fig.6b Mean of each window
300
250
200
Mean of the window
6000
5000
150
4000
Statistical area of the objects
3000
100
2000
1000
50
0
0 50 100 150 200 250 300
Object numbers
0
0 2 4 6 8 10 12 The sample outputs of statistical area of
3 X 3 window position x 10
4
the imfeature is shown in Table 1.
The mean (Figure 6b) and the contextual values
(Figure 6c) are shown.
Fig.6c Contextual values calculated for each window

300
250
Contextual value of window
200
150
100
50
0
0 2 4 6 8 10 12
3 X 3 window position 4
x 10
The property of imfeature is applied to the segmented

image. The area of the labeled objects in the
segmented image are obtained. The optic disc in the
image is removed by using a threshold. If the area of
45
Implementation of RBF VII. CONCLUSION

Table 1 Data for training RBF
Training Inputs Target
outputs
Area Filled Solidity Orientation Labeling
The main focus of this work is on segmenting
Area
55 59 0.6111 -16.5837 1 the diabetic retinopathy image and classify the
exudates. Segmentation is done using contextual
59 59 0.7024 -29.5294 1 clustering and classification of the exudates is done
using radial basis function (RBF) network. The
61 61 0.5980 43.1644 1
performance classification of exudates by using RBF
64 64 0.5161 -4.1202 1 and CC is better than that of using only CC. The
proposed RBF classifies the segmented information
69 70 0.6970 20.1090 1
of the image into hard exudates or not.
75 75 0.7732 7.9202 1 1. All the fundus images in this work have to be
transformed to a standard template image condition.
78 80 0.6393 82.4571 1
This corrects in the illumination effect on the images.
89 91 0.5973 84.0033 1 2. Only when the fundus image is taken with good
quality, detection of exudates is more accurate.
100 101 0.5587 -39.8444 1
104 108 0.7324 -12.7048 2

REFERENCES
109 109 0.8790 42.4872 2

[1] XU Jin, HU Guangshu, HUANG Tianna, HUANG Houbin
139 139 0.7128 81.1306 2 CHEN Bin “The Multifocal ERG in Early Detection of Diabetic
Retinopathy” - Proceedings of the 2005 IEEE Engineering in
165 165 0.9016 45.6726 2 Medicine and Biology 27th Annual Conference, Shanghai,
China, September 1-4, 2005
[2] Akita K. and H. Kuga. A computer method of
167 180 0.5860 55.3490 2 understanding ocular fundus images. Pattern Recognition,
15(6):431–443,1982.
214 219 0.7431 40.4485 2 [3] Lili Xu and Shuqian Luo, A novel method for blood vessel
detection from retinal images, BioMedical Engineering OnLine
251 251 0.6452 80.2676 2 2010, 9:14 doi:10.1186/1475-925X-9-14
5108 5117 0.8913 91.3917 2 [4] Walter, T.; Klevin, J.C.; Massin, P.; et al. A Contribution of
Image Processing to the Diagnosis of Diabetic Retinopathy —
Detection of Exudates in Color Fundus Images of the Human
Retina. IEEE Transactions on Medical Imaging 2002, 21,
1236–1243
The RBF is trained with the data given in
Table 1. Each row made up of 4 variables. A [5] Akara Sopharak and Bunyarit Uyyanonvara, “Automatic
Exudates Detection From Non-Dilated Diabetic Retinopathy
labeling is given in the last column. Eighteen Retinal Images Using FCM Clustering Sensors 2009, 9, 2148-
patterns are considered for training RBF. These 2161; doi:10.3390/s90302148
18 patterns are taken from the segmented [6] Xiaohui Zhang, Opas Chutatape School Of Electrical &
Electronic Engineering Nanyang Technological University,
image given in Figure 5e. Additional hard Singapore, Top-Down And Bottom-Up Strategies In Lesion
exudate images can also be considered from Detection Of Background Diabetic Retinopathy. Proceedings
Of The 2005 IEEE Computer Society Conference On
which additional patterns can be obtained. A Computer Vision And Pattern Recognition (CVPR’05), 2005.
topology of 4(nodes in the input layer) X 5(nodes in the hidden [7] Vallabha, D., Dorairaj, R., Namuduri, K., and Thompson,
H., Automated detection and classification of vascular
layer) X 1(node in the output layer) is used for training RBF. abnormalities in diabetic retinopathy. Proceedings of Thirty-
The final weights obtained is used for Eighth Asilomar Conference on Signals, Systems and
Computers. 2:1625–1629, 2004.
classification of the segmented exudates from
[8] Liu, Z.; Chutatape, O.; Krishna, S.M. Automatic Image
the noise present in the segmented image. Analysis of Fundus Photograph. IEEE Conf. on Engineering
in Medicine and Biology 1997, 2, 524–525.
[9] Christopher E. Hann, James A. Revie, Darren Hewett,
Geoffrey Chase and Geoffrey M. Shaw, Screening for
Diabetic Retinopathy Using Computer Vision and
Physiological Markers, Journal of Diabetes Science and
Technology Volume 3, Issue 4, July 2009
46
[10] Sinthanayothin, C.; Boyce, J.F.; Williamson, T.H.; Cook,

H.L.; Mensah, E.; Lal, S.; et al. Automated Detection of
Diabetic Retinopathy on Digital Fundus Image. J. Diabet.
Med. 2002, 19, 105–112.
[11] Usher, D.; Dumskyj, M.; Himaga, M.; Williamson, T.H.;
Nussey, S.; et al. Automated Detection of Diabetic
Retinopathy in Digital Retinal Images: A Tool for Diabetic
Retinopathy Screening. J. Diabet. Med. 2004, 21, 84–90.
[12] Sanchez, C.I.; Hornero, R.; Lopez, M.I.; et al. Retinal
Image Analysis to Detect and Quantify Lesions Associated
with Diabetic Retinopathy. IEEE Conf. on Engineering in
Medicine and Biology Society 2004, 1, 1624–1627.
[13]Contextual clustering for analysis of functional fMRI data,
IEEE Transaction on Medical imaging, 20(2001),pp403-411
[14] Purushothaman S., Sambasiva Rao Baragada,
Ramakrishna S., Rao M.S., Polynomial Discriminant Radial
Basis Function for Steganalysis, International Journal of
Computer Science and Security
[15] García M, Sánchez CI, Poza J, López MI, Hornero R.,
Detection of hard exudates in retinal images using a radial
basis function classifier, Ann Biomed Eng. 2009
Jul;37(7):1448-63. Epub 2009 May 9.
[16] J. Anitha; C. Kezi Selva Vijila; D. Jude Hemanth,

Automated radial basis function neural network based image
classification system for diabetic retinopathy detection in
retinal images (Proceedings Paper) , Second International
Conference on Digital Image Processing, Proceedings Vol.
7546, 26 February 2010, DOI: 10.1117/12.852746
[17]García M, Sánchez CI, López MI, Abásolo D, Hornero R.

,Neural network based detection of hard exudates in retinal
images, Comput Methods Programs Biomed. 2009
Jan;93(1):9-19. Epub 2008 Sep 7.
[18] M. García, M. I. López, R. Hornero, A. Díez and J. Poza,

Utility of a Radial Basis Function Classifier in the Detection of
Red Lesions in Retinal Images, World Congress On Medical
Physics And Biomedical Engineering, September 7 - 12,
2009, MUNICH, GERMANY IFMBE Proceedings, 2009,
Volume 25/11, 21-24, DOI: 10.1007/978-3-642-03891-4_6
47
Peristaltic flow of a Williamson fluid

in an asymmetric channel through
porous medium
A.Kavitha1 R.Hemadri Reddy*2 S.Sreenadh1 R.Saravana1
1
Department of Mathematics, Sri Venkateswara University, Tirupati-517 502
2
* School of Advanced Sciences, VIT University, Vellore, TamilNadu
release of liquid and solid gastric content into the
Abstract:Peristaltic transport of a Williamson fluid small intestines where nutrient uptake occurs. The
in an asymmetric channel through porous MRI image of the stomach has revealed its
medium is studied under long wavelength and asymmetry nature. Another physiological system
low Reynolds number assumptions. The namely uterus is also modeled as an asymmetric
nonlinear governing equations of the peristaltic channel by Eytan and Elad [17]. These facts will
flow are solved using perturbation method. The explain the necessity of considering the physiological
solution for the stream function is obtained by system to be asymmetric ducts also. Motivated by
neglecting inertia and curvature effects. The these facts, it will be interesting to study the peristaltic
velocity distribution, the volume flow rate and the transport of Williamson fluid through an asymmetric
pressure rise are also determined. channel filled with porous material.
In this paper peristaltic pumping of Williamson
Key words: Williamson fluid, Reynolds number, fluid through a porous medium in an asymmetric
Peristalsis, Velocity and Pressure rise. channel with flexible walls is investigated. Using
the wave frame of analysis, boundary value
I. Introduction problem is solved and the results are discussed
Peristalsis is a well-known mechanism for through graphs.
pumping biological and industrial fluids. Even though
it is observed in living systems for many centuries; the
mathematical modeling of peristaltic transport has II. Mathematical Formulation
begun with the important works by Fung and Yih [1] Let us consider the peristaltic transport of an
using laboratory frame of reference and Shapiro et incompressible Williamson fluid in a two dimensional
al.[2] using wave frame of reference. Many of the channel of width d1 + d 2 . The flow is generated by
contributors to the area of peristaltic pumping have
either followed Shapiro or Fung. Most of the studies sinusoidal wave trains propagating with constant
on peristaltic flow deal with Newtonian fluids. The speed c along the channel walls. The geometry of the
complex rheology of biological fluids has motivated wall surfaces are defined as
investigations involving different non-Newtonian  2π 
Y = H1 = d1 + a1 cos  ( X − ct ) (upper wall),
fluids. Peristaltic flow of nonNewtonian fluids in a tube λ 
was first studied by Raju & Devanathan [3].
 2π 
Peristalsis is a mechanism adopted by many Y = H2 = −d2 − b1 cos  ( X − ct ) +φ (lower wall),
physiological systems and mechanical peristaltic  λ 
pumps. Most of the physiological systems may be (1)
approximated as symmetric ducts. In view of this, where a 1 and b1 are the amplitudes of the
Peeyush Chandra [4], Sarojamma et al [5],
Ramachandra Rao and Usha [6], Misra & Pandey [7], waves, λ is the wave length, d 1 + d 2 is the width of
Vajravelu et al. [8-11], Subba Reddy et al. [12, 13]
the channel, c is the velocity of propagation, t is the
and Srinivas et al.[14,15] made detailed studies on
peristaltic pumping through tubes and channels. time and X is the direction of wave propagation. The
Brassuer and Anupampal (vide Chengel & Cimbala, phase difference φ varies in the range 0 ≤ φ ≤ π in
[16]), made experiments on the mechanical which φ = 0 corresponds to symmetric channel with
functioning of the stomach using MRI (Magnetic
Resonance Image). They observed that the stomach waves out of phase and φ = π the waves are in
is a mixer, a grinder, a storage chamber, and a
sophisticated peristaltic pump that controls the
48
phase, further a1 , b1 , d1 , d 2 and φ satisfies the and low Reynolds number, neglecting the terms of
order δ and higher, equations (5) and (6) take the
condition a
2
1 + b 21 + 2a1b1 cos φ ≤ (d1 + d 2 ) 2 . form
∂p ∂   ∂u   ∂u 
= 1 + We     − σ 2 (u + 1) (7)
∂x ∂y   ∂y   ∂y 
∂p
=0 (8)
∂y
The corresponding boundary conditions in wave
frame of reference are given by
u= -1 on y=h1(x)
u= -1 on y=h2(x) (9)
Elimination of pressure from equations (7) & (8) yields
dp ∂   ∂u   ∂u 
Figure 1.Physical Model = 1 + We     − σ 2 (u + 1) (10)
dx ∂y   ∂y   ∂y 
Introducing a wave frame ( x , y ) moving with velocity The volume flow rate q in a wave frame of reference
is given by
c away from the fixed frame ( X , Y ) by the h1 ( x )
transformation, q= ∫ udy
x = X −ct , y =Y ,u =U −c ,v =V and P(x) = P(X,t). h2 ( x )
(2) (11)
and defining The instantaneous flow Q(x, t) in a fixed frame is
h1(x) h1(x) h1(x)
x y u v c h h λ d
x= , y= ,u= ,v= ,t = t , h1 = 1 , h2 = 2 , τxx = τxx, τxy = 1 τxy Q(x,t) = ∫ (u +1)dy = ∫ udy + ∫ 1dy=q+(h -h )
λ d1 c c λ d1 d2 µ0c µ0c h2 (x) h2 (x) h2 (x) 1 2
d d ρcd Γc d γ&d 2
τyy = 1 τyy, δ = 1 , Re= 1 ,We= , P= P, γ& = 1 1
µ0c λ µ0 d1 cλµ0 c (12)

The time average flux Q over one period T (= λ / c )
(3)
of the peristaltic wave is
and using the above non-dimensional quantities, the T 1
1 1
resulting governing equations become (Nadeem[18]), Q= ∫ Qdt = ∫ ( q + h1 − h2 )dt = q + 1 + d (13)
∂u ∂v T 0 T 0
+ =0 (4)
∂x ∂y
 ∂u ∂u  ∂p 2 ∂τ xx ∂τ xy III. Perturbation solution
δ Re u + v  = − −δ − −σ 2 (u +1) (5) Since, equation (10) is non-linear; its exact solution
 ∂x ∂y  ∂x ∂x ∂y may not be possible. Therefore, we expand u, P and
 ∂v ∂ v ∂p ∂τ ∂τ q as
δ 3 Re u + v  = − − δ 2 xy − δ yy (6)
u = u0 + Weu1 + 0(We 2 ), P = P0 + WeP1 + 0(We 2 ),
 ∂x ∂y  ∂y ∂x ∂y
q = q0 + Weq1 + 0(We 2 ), (14)
where
∈ ∂p
σ2 =
k
, Da = 2 , ∈= porosity and k = permeability where P = , Substituting above expressions in
Da a ∂x
∂u , equation (10) and boundary conditions (9), we get the
τ xx = −2 [1 + Weγ& ] following system.
∂x
 ∂u ∂v  ∂v 0
τ xy = − [1 + Weγ& ]  + δ 2  , τ yy = −2 [1 + Weγ& ] , System of order We
 ∂x ∂x  ∂y
dp0 ∂ 2u0
= 2 − σ 2 (u0 + 1),
1
 2 2
 (15)
 ∂u   ∂u 2 ∂v  2  ∂v 
2 2
γ& =  2δ 2   + +δ  + 2δ    . dx ∂y
  ∂x   ∂y ∂x   ∂y   and the respective boundary conditions are
Here δ, Re, We represent the wave, u 0 = −1 for y = h 1 , u 0 = −1 for y = h2
Reynolds and Weisseing numbers, respectively.
(16)
Under the assumptions of long wavelength δ << 1
49
System of order We1 and the volume flow rate q1 is given by
σ h1 σ h2
2
A −2σ h1  e −e  A −2σ h1  e −σ h1 − e−σ h2 
dp1 ∂ 2u1 ∂  ∂u0  q1 =  + e   − E ( h1 − h2 )
 − σ u1
e
= +  2
(17) D  σ  D  σ 
dx ∂y 2 ∂y  ∂y 
 eσ h1 − eσ h2 
+ Ee −σ h1   +H
u 1 = 0, for y = h 1 , u 1 = 0, for y = h 2 (18)  σ 
0 (23)
Solution for system of order We
Solution of Eq. (15) satisfying the boundary where
conditions (16) can be written as ( B+C) e−2σh1 eσh1 −eσh2 + F−G e−σh1 eσh1 −eσh2 +( B+C) e−σh1 −e−σh2 
H=   ( )    
p  1 −σ h  σ y  D  σ   σ  D  σ 
−σ y 
u0 = 0  (1 − ke 1)  e + ke −1 −1
σ 2  eσ h1 ( e −e ) − dx 3σ ( e −e )
  F 2σh1 2σh2 dp0 k2 −2σh1 −2σh2
−
  2σe2σh1
(19)
where


 σ h2 σ h1 
e −e 

 , and the volume flow From equation (23), we get
dp1 σ
=
3
( q1 − H ) D
k =   
 eσ ( h2 −h1) − eσ ( h1−h2 ) 
dx I
  (24)
 
where
rate q 0 is given by
I = ( eσh1 −eσh2 ) e−2σh1 +( eσh1 −eσh2 )( e−σh1 −e−σh2 ) −σD( h1 −h2 ) + De−σh1 ( eσh1 −eσh2 )
q0 = 2
dp0  1  −σh1
  3 e (1−ke )( e −e ) −k( e −e ) −σ( h1−h2)−( h1 −h2)
−σh1 σh1 σh2 −σh1 −σh2 Substituting equations (21) and (24) in to equation
 dx σ  (14) and using the relation (14), we get
dp σ3    L  (
(20) = ( q+( h1 −h2 ) ) −σ6We2 q2 +h12 +h22 −2hh
1 2 +2q( h1 −h2 ) 2 
From Equation (20), we get
dx J    J 
25)
dp0 σ ( q0 + ( h1 − h2 ) ) 3
where J=
I
, Integrate above equation over
= −σh1
dx e (1− ke−σh1 )( eσh1 − eσh2 ) − k ( e−σh1 − e−σh2 ) −σ ( h1 − h2 ) D
one wavelength, we get
(21)
(21) 1 1
σ3    L 
∆P=∫ dx =∫ ( q+( h1 −h2 ) ) −σ6We2 q2 +h12 +h22 −2hh
dp
1 1 2 +2q( h1 −h2 ) 2 dx
Solution for system of order We 0
dx 0
J   J 
Substituting the zeroth-order solution (19) into
(17), the solution of the resulting problem satisfying (26)
the boundary conditions take the following form.
where L = A1 + A2 + A3 − A4 + A5 + A6 − A7 − A8
( A+ B+C) e−σh1 + D( E + F −G) e−σh1  σy  A+ B+C −σy 1  dp1 
u1 =  e − e − 2 
 D   D  σ  dx 
 e −2σh1 e σ(2 h 2 −h1 ) − e σh 2 
2
A1 =
2
3σ 3
D
(1 − ke ) −σ h1 2
 
2  dp0  2 −2σy
−Fe−2σh1e2σy +   ke
3σ3  dx   eσ h1 − eσ h2 
 ,
(22)  σ 
where A = E ( eσ h1 − eσ h2 ) ,B =
F (e −2σ h1 σ ( 2 h2 − h1 )
e − eσ h2 , ) A2 =
2
k2 e ( − eσ h2 
σ ( h2 − 2 h1 )
 −e
σ ( h1 − 2 h2 )
)  e
σ h1
3σ 3 D σ 
( )
2
2  dp 0  2 σ ( h2 − 2 h1 )
C= − e ( 1 2)
σ h −2h
k e
3σ 3  dx  σ h1
 e −e  σ h2
A3 = 3 (1 − ke −σ h1 ) e −σ h1 
2 2
σ ( h2 − h1 ) σ ( h1 − h2 )
−e ,
D= e , E = 1  dp1  , 3σ  σ 
σ 2  dx 
σ h1 σ h2
2 2 e −e 
2
F = 2  dp0  1 − ke−σ h 2 , G =
2  dp0  2 −2σ h1 A4 = 3 k 2 e−2σ h1  
3   ( )   k e
3σ 3  dx  3σ  σ 
1
3σ  dx 
50
A5 =
3σ
2
3
D
(1 − ke ) ( e −σ h1 2 −2σ h1 σ ( 2 h2 − h1 )
e − eσ h2 )
 e −σ h1 − e −σ h2 
 
 σ 
−σ h1
− e −σ h2 
A6 =
2
3σ 3 D
(
k2 e
σ ( h2 − 2 h1 )
−e
σ ( h1 − 2 h2 )
)  e σ


A7 =
6σ
2
4 (1 − ke )
e −2σ h1 ( e 2σ h1 − e2σ h2 ) ,
−σ h1 2
A8 = 4 ( e −2σ h1 − e −2σ h2 )
k2
Fig.2. The variation of ∆P with Q for different values of
3σ
σ
for a fixed a = 0.8, b =0.5, φ =0, d =2; We = 0.03
IV. Results and Discussions
From equation (26) we have calculated the

pressure difference as a function of Q for different
values of permeability parameter σ and different
phase differences φ for a fixed a=0.8, b=0.5, d=2,
We=0.03 and is shown in figures (2) and (3).We
observe that the larger the parameter σ the greater
the pressure rise against which the pump works. We
observe that for a given ∆P , the flux Q increases
with increasing σ .For free pumping there is no
difference in flux Q for increase in σ .We also
Fig.3. The variation of ∆P with Q for different values of σ
observed that the pressure rise increases with
increasing phase difference φ . for a fixed a = 0.8,b =0.5, φ = π /6, d =2; We = 0.0
The variation of pressure rise with time
averaged flow rate is calculated from equation (26) for
different values of We and different phase differences
for a fixed a=0.8,b=0.5,d=2, σ =1.5 and is shown in
figures (4) to (6). We observe that the larger the We,
the smaller the pressure rise against which the pump
works. We observe that for a given ∆P ,the flux Q
decreases with increasingWe. For a given flux Q the
pressure rise decreases with increase We.
The variation of ∆P with time averaged flow
rate is calculated from equation (26) for different
values of the phase difference φ , for a fixed a=0.8,
b=0.5,d=2, σ =0.5, We=0.03 and is shown in figure
(7). We observe that the larger the phase
difference φ , the greater the pressure rise against
which the pump works. We observe that for a Fig.4. The variation of ∆ P with Q for different values of We
given ∆P , the flux Q increases with increasing φ . for a fixed a = 0.8,b =0.5, φ = 0, d =2; σ = 1.5.
For a given flux Q the pressure rise increases with
increasing φ .
51
REFERENCES
[1]. Fung, Y.C. and Yih, C.S. Peristaltic transport,
Trans ASME, E.J. App. Mech, 35, pp.669-675,
1968.
[2]. Shapiro, A.H., Jaffrin, M.Y and Weinberg, S.L.
Peristaltic pumping with long wavelengths at
low Reynolds number, J. Fluid Mech. 37,pp.
799-825, 1969.
[3]. Raju, K.K. and Devanathan, R. Peristaltic
motion of a non-Newtonian fluid part –I Rheol.
Acta. 11,pp. 170-178, 1972.
Fig.5. The variation of ∆P with Q for different values of We [4]. Peeyushchandra,G.Radhakrishnamacharya,

Viscous Flow through a Permeable Tube of
for a fixed a = 0.8, b =0.5, φ = π / 6 , d =2, σ = 1.5. Varying Cross-section with Reference to
Proximal Renal Tubule Physiological Fluid
Dynamics-I, SCPFD Publications, Madras
PP.113,1985.
[5]. Sarojamma, G., Haritha, P. and Nagarani, P.
Peristaltic Transport of a Viscous Fluid in a
Channel under the Influence of a Transverse
Magnetic Field, Proc. of AP Academy of
Sciences, Vol. 7, No. 1, pp.25-32, 2003,
[6]. Ramachandra Rao, A., and Usha, S. Peristaltic
transport of two immiscible viscous fluid in a
circular tube, J.Fluid Mech., 298,pp.271-285,
1995.
[7]. Misra, J.C. and Pandey, S.K. Peristaltic
transport of a non-Newtonian fluid with a
peripheral layer, Int. J.Engg Sci., 37,pp. 1841-
Fig.6. The variation of ∆P with Q for different values of We 1858, 1999.
for a fixed a = 0.8, b =0.5, φ = π / 3 , d =2, σ = 1.5. [8]. Vajravelu, K. Sreenadh, S. and Ramesh Babu,
V. Peristaltic pumping of a Herschel-Bulkley
fluid in a channel, Appl. Math.&Computation,
169,pp. 726-735, 2005a.
[9]. Vajravelu, K. Sreenadh, S. and Ramesh Babu,
fluid in an inclined tube, Int. J. Non-linear Mech.
40,pp. 83-90, 2005b.
[10]. Vajravelu, K. Sreenadh, S. and Ramesh Babu,
fluid in contact with a Newtonian fluid,
Quarterly of Appl. Math.64,No.4,pp.593-604,
2006.
[11]. Vajravelu. K, Sreenadh. S, Hemadri Reddy. R,
and Murugeshan.K, Peristaltic Transport of a
Casson fluid in contact with a Newtonian Fluid
Fig.7. The variation of ∆P with Q for different values of φ in a Circular Tube with permeable wall,
for a fixed a = 0.8, b =0.5, We =0.03, d =2, σ = 0.5. International Journal of Fluid Mechanics
Research, 36 (3),pp. 244-254,2009.
52
[12]. Subba Reddy, M.V. Manoranjan Mishra,
Sreenadh, S. and Ramachandra Rao, A.
Influence of lateral walls on peristaltic flow in a
Rectangular ducts, Journal of Fluids
Engineering, 127,pp. 824-827, 2005.
[13]. Subba Reddy.M.V, A.Ramachandra Rao and
S.Sreenadh, Peristaltic motion of a power-law
fluid in an asymmetric channel, International
Journal of Nonlinear Mechanics, 42,pp. 1153-
1161, 2006.
[14]. Kothandapani. M, Srinivas. S , Non-linear
peristaltic transport of a Newtonian fluid in an
inclined asymmetric channel through a porous
medium, Physics Letters A, Volume 372, Issue
8, pp. 1265-1276,2008.
[15]. Srinivas.S, Kothandapani. M, Peristaltic
transport in an asymmetric channel with heat
transfer International Communications in Heat
and Mass Transfer, Volume 35, Issue4,pp.514-
522,2008.
[16]. Brasseur, J.G., Corrsin, S. and Lu, Nan Q. The
influence of a peripheral layer of different
viscosity on peristaltic pumping with Newtonian
fluids, J.Fluid Mech., 174,pp.495-519, 1987.
[17]. Eytan, O. and Elad, D. Analysis of Intra –
Uterine fluid motion induced by uterine
contractions, Bull. Math. Bio.,61,pp.221-238,
1999.
[18]. Nadeem.S ,Akram Safia, Influence of inclined
magnetic field on peristaltic flow of a Williamson
fluid model in an inclined symmetric or
asymmetric channel, Mathematical and
Computer Modelling, Volume52,Issue1-2,
pp.107-119, 2010
[19]. Nagarani, P., Sarojamma, G. and Girija.
Jayaraman, Effect of boundary absorption in
dispersion in Casson fluid flow in a tube, Annals
of Biomedical Engineering, Vol. 32, No.
5,pp.706-719, 2004.
53

IJITCE Jan

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IJITCE Jan

Uploaded by

Copyright:

Available Formats

INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY & CREATIVE ENGINEERING (ISSN:2045-8711)

VOL.1 NO.1 JANUARY 2011

2. Crowd Safety: A Real Time System For Counting People ……….[6]

3. A Novel design of Electronic Voting System Using Fingerprint…..[12]

4. GPS Tracking Simulation By Path Replaying…………………………[20]

5. Object Oriented Design of E-learning System for Distance Education…[27]

6. On The Relative Character Graph of A Finite Group………………..[31]

7. Detection of Diabetic Retinopathy Using Radial Basis Function….[40]

8. Peristaltic flow of a Williamson fluid in an asymmetric channel through porous

Measurement of Potassium Levels in the

Introduction: The main objective of this research is to

Figure.1 consists of two sensors (i.e.

Procedure for determining the concentration of

Prepare two standard concentrations C1 and C2

8. As written in the software, when input changes the

9. Here when enable button is active, it is indicated in

10. If we want to end the process, enable is disabled or

11.Appache Web Server software is used to configure

12.LCD display is used for standalone results.

S.no. Soil Measurement Present Fig: 5: On line measurement of potassium sample

[2]Advanced sensor technologies, Inc.Toll free: 1-888-wow-ASTI

[3]“Computer networks” By Andrew S.Tanenbaum, prentice hall

[10] G.A.Rechnitz and N.C.Kenny, Anal.Lett. 2,395(1969)

[11] G.kakabadse, Ion-selective Rev., 3,127(1982).

[12] R.A.Durst, “Ion Sensitive Electrodes”, [13] [13]

[14] Iodide ion selective electrode manual of pH products company

[15] A. Hameed, M. A. Goddal, Z. H. Yamini, A. H. Yahya,

[16] Australian Journal of Basic and Applied Sciences, 4(5): 922-931,

[17] K. Hiiro, A. Kawahara and T. Tanaka, Anal. Chim. Acta,110, 321

[18]K. Hiiro, A. Kawahara and T. Tanaka, Nippon Kagaku Kaishi,1980,

[19]K. Hiiro, A. Kawahara and T. Tanaka, Bunseki Kagaku, 31, E33

[20]K. Hiiro, T. Tanaka, A. Kawahara and S. Wakida, Anal. Sci., 2,145

[21]K. Hiiro, S. Wakida, T. Tanaka, A. Kawahara and M. Yamane,

[22] Denshi Kogyo Geppo, 27, 38 (1985).

CROWD SAFETY: A REAL TIME SYSTEM

at Bhavani where a flow rate of people is very high

Fig. 1. a) Entry gate

Fig. 2. b) Passage of people through wooden barriers (Back

When a person enters through the door where

and serial programming and re-programming with

Percentag Advanced Video and Signal Based

Science from BITS, Pilani. He holds a

Sivakumar M has 10+ years of experience

A Novel design of Electronic Voting System

• Constant spending funds for the elections staff

So, the proposed electronic voting system has to

III FUNDAMENTALS OF FINGERPRINT

The types of information that can be collected from

Fig 3 Typical Structure for Fingerprint System [16]

2) AFIS (Automatic Fingerprint Identification/ significant percentage of fingerprint images is of poor

Pores extraction: Pores are extremely fine details 2 2

From the results of Table 6 it is declared that Mr.

V. CONCLUSIONS AND FUTURE DIRECTION

Fig.7 A Novel Design for E-Voting Process

GPS Tracking Simulation by Path Replaying

The GPS receivers produce GPS data in the form

III. PREVIOUS W ORK researchers to develop and test GPS applications.

A keyboard or mouse controlled NMEA sentence

A work in this area has recently been done by the

E/W Indicator Blank Not Used

Fig. 6 GPS database extracted from new $GPRMC sentences

Object Oriented Design of E-learning System

is a continuous process where the learner deprived to