|
CSPA803 |
Charaterization of Wireless Communication Channel in the
HF Frequency Band Zaiton
Sharif, Ahmad Zuri Sha’ameri and Fadzliana Saad Digital Signal Processing Lab, Fakulti Kejuruteraan
Elektrik, Universiti
Teknologi The estimation of channel impulse response (CIR) and
the channel frequency response (CFR) are performed using a cross correlation
method. Three different types of input
signals are used for this purpose.
There are Phase Shift Keying (PSK), Quadrature Phase Shift Keying
(QPSK) and Linear FM signal (LFM). Both
PSK and QPSK are modulated based on pseudo random sequence. A specific parameter of sampling frequency,
data bit rate and sequence length are chosen for this purpose. All of signals are subjected to various
gain of random noise (White Gaussian noise) generated. Then, the CIR and the CFR are respectively
estimated based on the correlation functions.
These signals are chosen due to their auto correlation characteristic
that approaches an impulse function.
Since the input-output relationship in time is a convolution between
the auto correlation of input and CIR, the cross correlation between output
and input signal is the estimate of the CIR.
The FFT of the CIR gives the CFR.
The estimation of CIR and CFR of these signals, with a given delay and
subjected to AWGN will be compared with their respective desired values using
a statistical method called mean square error (MSE). The relationship between SNR and MSE of
these signals gives an indication on the robustness of these signals with
respect to noise. |
|
CSPA805 |
Shadow Removing using Contourlet-based
Image Fusion Shirin
Mahmoudi Computer
Group, Engineering Department, Azad
University – Islamshahr branch, Islamshahr, The fusion of high-spectral but low-spatial
resolution multispectral and low-spectral but high spatial resolution
panchromatic satellite images is a very useful technique in various
applications of remote sensing, such as change detection. Recently, some
studies showed that wavelet-based image fusion method provides high quality
of the spectral content of the fused image. However, most of wavelet-based
methods have a spatial resolution of the fused result less than the Brovey,
IHS, and PCA fusion methods. However, wavelet transform could not efficiently
represent the singularity of linear/curve of the images. In this paper, we
introduce a new method based on the contourlet transform which represents
edges better than wavelets and remove the shadows of images. Since edges play
a fundamental role in image understanding, one good way to enhance spatial
resolution is to enhance the edges. So, we could find an automatic threshold
for discarding the weak edges of panchromatic satellite images using
information measure and then strong edges were fused in multi-spectral
images. We test the proposed approach on IRS satellite images and compared
our results with wavelet transform and IHS ones. |
|
CSPA806 |
Influence
of Statistical Information Criteria to One-Step-Ahead Prediction Error Mohd Hezri
Fazalul Rahiman, Mohd Nasir Taib and Faculty of
Electrical Engineering, Universiti Teknologi MARA, 40450 Shah Alam, MALAYSIA. This paper investigates the influence of popular
statistical information criteria to one-step-ahead prediction (1-SAP) error
of an ARX black-box model. The criteria investigated are the Akaike
Information Criteria (AIC), Akaike Final Prediction Error (FPE) and
Rissanen’s Minimum Description Length (MDL). The investigation will be based
on Pseudo-Random Binary Sequences (PRBS) data collected from an electrically
heated steam distillation essential oil extraction system. The data is the
steam temperature measured within the distillation column beneath the
material bed. By using MATLAB System Identification Toolbox, an ARX model
will be estimated and validated. Prior to model validation, all the
information criteria will be examined and the criteria that suggested the
most flexible model shall be selected for future works. The linear regression
will be minimized by using Levenberg-Marquardt algorithm. Evaluation of model
performance will be based on both graphical and statistical approaches such
as R2, adjusted-R2, residual
distribution, mean and variance. The results have shown that the selected
model based on MDL criterion is more parsimonious and flexible as compared to
the others. |
|
CSPA808 |
Realizing a Digital FIR Filter using Xilinx
FPGA Technology Abdul Hadi Abdul Razak and Mohd Faizul Idros Faculty of Electrical Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor,
MALAYSIA. In this day and age compression of data in digital signal
processing is a necessity. Compression formats such as MPEG, JPEG, MP3 and
others are widely used in mass media and computer technology. Usually compressions of data are being
developed using computer software, and RISC processor in a decoder. In every
format of compression data, Digital Finite Impulse response filter is use as
one of the main process in data compression including audio, image and video
compression. This paper describes the design of Transposed
Form FIR filter implemented in the Spartan-II and Virtex-E family of FPGAs.
The design is an 8-tap filter based on 16-bit input samples and 14-bit signed
coefficients. The basic building blocks of the filter are KCMs, Adders,
Registers, and a delay-locked loop. All the 14-bit coefficient factors are
stored with an 18-bit word size in the |
|
CSPA809 |
WiMAX
– Solution for BWA and Confirms to IEEE 802.16 Standard Muhamad
Radzi Abdul Rahman and Kamaruddin Mamat Faculty of
Information Technology and Quantitative Sciences, Universiti Teknologi MARA, 40450 Shah Alam, MALAYSIA. WiMAX is a wireless digital communications system,
also known as IEEE 802.16 that is intended for wireless "metropolitan
area networks". WiMAX can provide broadband wireless access (BWA) up to 30
miles (50 km) for fixed stations, and 3 - 10 miles (5 - 15 km) for mobile
stations. In contrast, the WiFi/802.11 wireless local area network standard
is limited in most cases to only 100 - 300 feet (30 - 100m). With WiMAX, data
rates are easily supported, but the issue of interference is lessened. WiMAX
operates on both licensed and non-licensed frequencies, providing a regulated
environment and viable economic model for wireless carriers. WiMAX can be
used for wireless networking in much the same way as the more common Wi-Fi
protocol. WiMAX is a second-generation protocol that allows for more
efficient bandwidth use, interference avoidance, and is intended to allow
higher data rates over longer distances. This paper investigates the
possibilities of using WiMAX as a solution for BWA and confirms to IEEE
802.16 Standard. The IEEE 802.16 standard defines the technical features of
the communications protocol. The WiMAX
Forum offers a means of testing
manufacturer's equipment for compatibility, as well as an industry group
dedicated to fostering the development and commercialization of the
technology. |
|
CSPA810 |
JSA based Job Shop Scheduling
Optimization P.V Senthiil 1 , V. Selladurai 2 1 Assistant
Professor, Department of Mechanical Engineering, 2 Professor,
Department of Mechanical Engineering, Timely and cost effective production is becoming
increasingly important in today’s global competitive market. The key problem
faced by today’s manufacturing industries is feasible allocation of various
jobs to available resources i.e. processors. This problem can be outwitted by
a proper scheduling technique, adopted by industries. A typical job shop
scheduling problem has a set of jobs to be processed in a set of processors,
with certain constraints and objective functions to be achieved. The most
commonly considered objectives are the minimization of make span (Cmax),
minimization of tardiness (T) which leads to minimization of late penalty cost,
and to maximize machine utilization. Machine shop scheduling can be done
using various techniques like standard dispatching rules, heuristic
techniques like Simulated annealing, Tabu Search, Genetic algorithm, etc. In this article a
typical job shop scheduling problem is solved using algorithmic simulated
annealing (SA) technique, a heuristic search algorithm. SA is a generic
neighborhood search algorithm used to locate optimal solution very nearer to
a global optimal solution. As a part of this article, a software based
program is developed in VB platform and realistic test instances were
performed. Experimental results obtained were further tuned by varying input
job parameters and optimal results were obtained. |
|
CSPA811 |
Detection of Fractured Bones in X-ray
Images using Image Segmentation Y. Zakariah, A.J. Nor’aini Faculty of Electrical Engineering, Universiti
Teknologi MARA, 40450 Shah Alam, MALAYSIA. Fractured bones are common affliction. Each year
many fractures are missed during x-ray diagnosis, resulting in
ineffective patient management and expensive litigation. The use
of computers has helped make computation work significantly faster, more
efficient, and with high accuracy. Since bone fracture can occur in many
ways, no one single algorithm can detect all the possible fractures
accurately. In this paper a method of detecting fractures especially for hair
line fracture is presented. Image processing technique involves image
segmentation that includes threshold and edge
detection using canny edge detector. It also involves filtering using median
filters. The basic processing techniques such as intensity transformation and
intensity manipulation are also used. Ten samples of x-ray images were used
for this purpose. Experiment performed shows that the method is
capable of accurately segmenting fracture as well as capable to enhance the
resolution of the fractured bone images from its original images. |
|
CSPA812 |
Implementation of Fuzzy
Logic Controller for Digital Pendulum Afaf Rozan Mohd Radzol, Kamarul Azlan bin Azman Faculty of
Electrical Engineering, Universiti
Teknologi MARA, Digital Pendulum System is an example of pole
balancing system and non-linear control theory having a complex dynamics
behavior that creates challenging control problem [5]. This paper is intended
to discuss the developed Fuzzy Controller to The Digital Pendulum Control
System. Swing is induced in the pendulum as the cart is moved back and forth
by the DC motor [5]. Generally the objective of implementing the fuzzy
controller to the system is to move the cart between positions from point to
point without letting the angle of the pendulum swing become too large and
comparing between the fuzzy controller method and PID controller method. The
effectiveness of the proposed control method is then shown in the
experimental results. |
|
CSPA813 |
Performance Assessment of Hybrid
Compression Schemes for Multispectral and Hyperspectral Data. R R. Sedamkar1, Bijith Marakarkandy2,
O. P. Jain3 1 NMIMS
University, 2
MPSTME-NMIMS, 3 L&T Infotec,
Mumbai, INDIA. Multispectral and Hyperspectral images are being increasingly
used in areas such as Biomedical Engineering and Forensic sciences apart from
traditional and key application areas viz. remote sensing and geosciences.
Multispectral and Hyperspectral imagery has four dimensions 2 spatial, 1
spectral and 1 temporal, therefore these data have very high volumes. Data
compression schemes have become essential for reducing storage requirements
and network traffic for economical distribution of this data. Lossless
compression methods are a good option, but they cannot offer a satisfactory
compression ratio for hyperspectral and Multispectral images, therefore lossy
compression schemes which preserve very useful information have to be
considered moreover the time required to compress and extract information
from the data is also crucial. In this paper we propose a hybrid scheme to
compress the hyperspectral and multispectral data. We use Principal Component
Analysis (PCA) for spectral decorrelation and dimensionality reduction and
the Principal components (PCs) are coded using variable length codes. We have
experimented on Landsat and AVIRIS images .The decompressed images were
evaluated objectively using standard performance metrics viz. Maximum
Absolute Error (MAE), Mean square error (MSE), Peak Signal to Noise Ratio
(PSNR), and performance of transformation is reported using merit of
transform. Distribution of Hyperspectral and Multispectral data over the
internet will take considerable time particularly if the medium over which
the image is being transmitted is slow. Many a times the user is interested
only in certain bands and not the entire data set. In such cases the proposed
method will enable transmission of only those compressed Principle components
(PC’s) which introduce the least error in the required bands along with the
transformation Matrix and Mean vector. The user can re-create almost an exact
version of the required bands using the proposed scheme |
|
CSPA814 |
Biometric Verification using Brain
Signatures Hema C.R,
Paulraj M.P, Abdul Hamid Adom, Sazali Yaacob, Nagarajan R., Harkirenjit Kaur Universiti In this paper we investigate the use of brain
signatures as a possible biometric verification technique. Research on brain
signals has shown that each individual has a unique brain wave pattern.
Biometry using brain signals is an emerging field with very little work been
done in this area. EEG signals generated by mental tasks are acquired to
extract the distinctive brain signature of an individual. EEG signals during
mental tasks were acquired from five subjects. In this paper we propose a
novel algorithm for identification using power spectrum and Elman neural
networks. The performance of the neural network is appreciable with an
average accuracy of 79 to 97.4%. |
|
CSPA815 |
Stereo Vision System for Assistive Smart
Wheelchair Hema C.R,
Paulraj M.P, Abdul Hamid Adom, Sazali Yaacob, Nagarajan R., Mohamad Sulaiman
Ibrahim Universiti Maneuvering and controlling an electrically powered
wheelchair poses a complex and hardly manageable task for severely
handicapped and elderly people with heavily reduced physical and mental
abilities. With an increase in the number of senior citizens there is an
increase in the demand for user friendly wheelchairs. This paper proposes a
vision based obstacle location system for wheelchair using stereo vision.
Neural network are used to estimate the distance of the obstacles.
Experimental results validate the proposed method. |
|
CSPA816 |
Domestic Solid Waste Sorting using Vision
Sensor Hema C. R,
Paulraj M. P, Sazali Yaacob, Abdul Hamid Adom, Nagarajan R., D. Maheswaran,
R. Jaii Ganes University Recognizing the problem of domestic waste in the
community is one the most important aspects of community life, especially as
domestic waste management impacts significantly on the environment. For
effective waste management, domestic waste must be initially separated into
plastic, glass, metal, and paper. Glass and metal objects can easily and
economically be separated from the waste stream by air classification or magnets,
but the weight and other properties of plastic and paper articles make them
more difficult to distinguish for purposes of separation. In this paper we
propose a novel algorithm using vision to sort paper and plastic waste. A
neural network classifier is designed to classify the paper and plastic
waste. Classification accuracy of 96.4 % were obtainable. |
|
CSPA817 |
Particle Swarm Neural Network Based
Classification of EEG Mental Tasks Hema C.R.,
Paulraj M.P, Sazali Yaacob, Abdul Hamid Adom, Nagarajan R. Universiti Brain machine interface (BMI) provides a digital
channel for communication in the absence of the biological channels. BMI are
used to rehabilitate patients with neurodegenerative diseases, a condition in
which all motor movements are impaired including speech leaving the patients
totally locked-in. BMI are designed using the electrical activity of the
brain detected by scalp EEG electrodes. Classification of EEG signals
extracted during mental tasks is a technique for designing a BMI. In this
paper five different mental tasks from two subjects were studied,
combinations of two tasks are studied for each subject. A PSO ELMAN neural
network is used for classification of the EEG signals. Band power features of
the EEG signals are used for the classification. Classification accuracies
varied from 79.5% to 91% for the 10 different task combinations for each of
the subjects. The results obtained validate the performance of the PSONN
algorithm for mental task classification. |
|
CSPA818 |
Ultrasonic Process Tomography Imaging
Sensor: An Analysis on Transceivers Sensing Method M.H.
Fazalul Rahiman1, Z. Zakaria1, R. Abdul Rahim2 1 Tomography
Imaging Research Group, Universiti 2 Process
Tomography Research Group (PROTOM) Control and
Instrumentation Engineering Department, Faculty of Electrical Engineering, Universiti
Teknologi Direct analysis of the internal characteristics of
process plant has become a widespread need in order to improve the design and
operation of the equipment especially in liquid/gas two phase flows. It has
been proven that the operation efficiency of a process is closely related to
accurate measurement and control of hydrodynamic parameters such as flow
regime and flow rate. In recent years, the applications of tomography
techniques as a robust non-invasive tool for direct analysis of the
characteristics of multiphase flows have increased. In this study, ultrasonic
tomography imaging sensor has been utilised to visualize the distribution of
liquid/gas in a vertical column. The sensing element consists of 32 units of ultrasonic
transceivers to cover the column cross-section. The motivation of the paper
is to analyse the performance of the transceiver methods in visualizing
bubble hold-ups in vertical column. Some analyses have been carried out using
several phantoms and the system was found excellent in visualizing the
internal characteristics and provides the concentration profile for the
corresponding phantoms. |
|
CSPA820 |
Model Order Selection Criterion for
dengue infections using Autoregressive Model H. Abdul Rahim1,
F. Ibrahim2 and M. N. Taib3 1Department
of Control and Instrumentation, Faculty of Electrical Engineering, Universiti
Teknologi 2Department
of Biomedical Engineering, Faculty of Engineering, 3Faculty of
Electrical Engineering, Universiti
Teknologi MARA, 40450 Shah Alam, This paper describes the development of linear
autoregressive (AR) models to diagnose the progression of dengue infection
based on hemoglobin. Three different AR model order selection criteria namely
Final Prediction Error (FPE), Akaike’s Information Criteria (AIC) and
Lipschitz number have been evaluated and analyzed. The results showed that
Lipschitz number has better accuracy compared to FPE and AIC. Finally based
on Lipschitz number, appropriate model orders have been selected for diagnose
the progression of dengue infection based on hemoglobin status. Further work is to apply this appropriate
model orders to nonlinear Autoregressive (NAR) model. |
|
CSPA821 |
Nonlinear Autoregressive Modeling to
Diagnose the Progression of Dengue Infection Based on Hemoglobin H. Abdul
Rahim1, F. Ibrahim2 and M. N. Taib3 1 Department
of Control and Instrumentation, Faculty of Electrical Engineering, Universiti
Teknologi 2Department
of Biomedical Engineering, Faculty of Engineering, 3Faculty of
Electrical Engineering, Universiti
Teknologi MARA, 40450 Shah Alam, This paper proposed a nonlinear AR (NAR) based on
Artificial Neural Networks (ANN) is applied to the modeling of a predicting
the hemoglobin status for dengue infection.
The type of order selection criteria involves the final prediction
error (FPE), Akaike’s Information Criteria (AIC), and Lipschitz number. This model is divided two approaches which
are unregelurized approach and regularized approach. The results show that
using NAR model with regularized approach yield better accuracy by 80.60% to
predict the hemoglobin status for dengue infection compared with AR model. |
|
CSPA822 |
Fourier Descriptor for Human Shape
Recognition Nooritawati
Md Tahir, Aini Hussain, Dept of
Electrical, Electronics and System, Faculty of Engineering, Universiti
Kebangsaan The aim of this study is to investigate Fourier Descriptor
(FD) as feature vectors for shape representation and recognition since FD is
the best known boundary based shape descriptor and has proven to outperform
most other boundary based methods in terms of accuracy. Furthermore, FD is
also invariant to geometric transformations and has good noise tolerance. The
main concern regarding FD is the number of terms that need to be maintained
from the original Fourier transform for effective representation and
description. A system that computed FDs of human and non human from their
silhouettes; normalized the descriptors and further applied as feature
vectors for recognition is developed. Initial results of experiment showed
that using adequate number of both low and high frequency components could
represent the shape based on high recognition rate achieved. The process of
shape recognition using FDs looks promising. |
|
CSPA823 |
Automated Interior Parked Vehicle Heat
Discharger using Fuzzy Logic Shaiful
Izzuddin bin Mohamad, Rosni Abu Kassim and Faculty of
Electrical Engineering, Universiti In this study, an automated heat discharger system
for a parked vehicle using fuzzy logic is developed. The Automatic Heat
Discharger acted as a device in controlling and maintaining the temperature
condition in a parked vehicle using two sensors and two actuators to simulate
the real world scenario. The sensors applied are the Negative Coefficient
Temperature (NTC) and light dependent sensor (LDR) whilst the actuators are
an exhaust fan and curtain as shade for protection of the vehicle interior
against hot weather and sun exposure. Additionally, the development of fuzzy
logic control algorithm is performed using the fuzzyTECH 3.5 environment. Initial
results proved that the developed system is capable to maintain the
temperature of a suitably sized of interior parked vehicle and perfectly
control the temperature by discharging the heat away from the vehicle
interior. |
|
CSPA825 |
Capacity and Performance Analysis of MIMO
systems with Receive Antenna Selection A. S.
Hiwale1 and A. A. Ghatol2 1 Dept. of
Electronics and Telecommunication Engineering, 2 Dr.
Babasaheb Ambedkar Technological University, Lonere,
Mangaon, Raigad, The use of multiple-antennas and number of
transceiver RF chains increase the hardware complexity of MIMO systems.
Antenna selection is the promising technique to reduce the complexity of MIMO
systems that employ less number of RF chains compared to number of available
antennas. Antenna selection at the receiver or transmitter or at both
transmitter/receiver greatly reduces complexity of MIMO systems while
retaining the benefits of full complex MIMO systems. Various selection
algorithms are proposed by the researches based on maximizing signal-to-noise
ratio (SNR) or maximizing capacity. This paper discusses some practical
selection rules and gives the exact analysis of capacity and error performance
of MIMO systems when using a subset of available antennas at the receiver.
The numerical results obtained using MATLAB for channel capacity and
performance shows that with two transmit antennas and selecting 2 out of 4
receive antennas yields capacity per channel use of 6 b/sec/Hz compared to 7 b/sec/Hz of a full complex system.
Similarly, a loss in signal-to-noise ratio (SNR) of about 3 dB is exhibited
at error probability of 10-5 in MIMO with antenna selection (2 out of 4
receive antennas) compared to MIMO system using all available antennas. |
|
CSPA826 |
ASTHRA - “An Automated Robotic Vehicle
for the Disabled patients using Embedded Systems” Natarajan
R., Dinesh Singh J. Department
of Information Technology, Chennai –
600 044, Tamil Nadu, ASTHRA is an automated, dynamic memory mapped
robotic vehicle, which enables the disabled to carry on their locomotion with
ease and confidence. Also Asthra acts as an automated patient monitoring unit
which enables to transfer vital information during emergency as well as
normal conditions. The vehicle mainly consists of five basic components
including a embedded web servers, dc motors with driver circuit, an
ultrasonic transmitter and receiver to detect obstacles, a radio frequency
remote and programmable and control memory unit. The dc motors are placed at
the hub of the vehicle’s tire and their chief aim is to drive or render
sufficient power to pull the weight of the patient to move about. The most
important feature of Asthra is the embedded web servers. The use of the
embedded web servers is to transfer information’s regarding the patient’s
temperature, pulse, ECG and other important necessary features that need to
be transmitted during times of emergency. We maintain a local intranet by
which the doctor could just login onto a website maintained at the intranet
and get the required details about each and every patient. Also alerting
mechanisms are added which enable to alert doctors regarding the condition of
patients under chronic health conditions. In our case, a RF remote is
provided as an interface between the user and the vehicle. The dc motors are
provided with a driving circuit which is connected to a PIC microcontroller
board. The microcontroller board enables dynamic memory mapping by utilizing
the “train and execute” mode of operation. Utilizing the ultrasonic obstacle
detection circuit, the microcontroller unit programs itself as per the
dimensions of a particular place. Thus Asthra combines upon various features
which render effective locomotion to the disabled as well as act as an
effective patient monitoring system. |
|
CSPA827 |
An Efficient Rate Scalable Technique for
Image Coding and Transmission N.B.
Chopade1 and A. A. Ghatol2 1Department
of Electronics and Telecomm., Engineering, 2Dr.
Babasaheb Ambedkar Technological University, Lonere 402103 Dist: Recently analog images such as Metrological,
Astronomical, Seismic, X-rays, MRI, ECG, EMG, and EEG data are replaced by
digital images for storage & retention. It has been estimated that every
year 80 billion new digital images are produced. Multimedia data (still
images, video & audio) accessed by Internet and mobile users also forms a
considerable portion of the database. Multimedia data in uncompressed form
requires considerable storage space, transmission bandwidth &
computational time. Compression & coding of digital images allows storing
an image data in secured and confidential manner. It also reduces the storage
& transmission cost. The necessity of image compression is continuously
growing in the recent past. The Image compression methods essentially include
transform of image, quantization and encoding. One of the promising & prominent
approaches in this domain is discrete wavelet based image compression, which
uses the concept of multirate DSP and subband coding. The popular JPEG
standard uses DCT for transforming image, which suffers from blocking
artifacts & mosquito noise. The JPEG2000 standard uses Discrete Wavelet
Transform (DWT) for image transformation. This paper describes wavelet based
coding method Set Partitioning in Hierarchical Trees (SPIHT) to encode &
compress the digital image data. It uses recursive set partitioning procedure
to sort subsets of wavelet coefficient by maximum magnitude with respect to
threshold. Because of low complexity & simplicity, this method has very
fast encoding and decoding capability which makes it very efficient in
multimedia transmission. The numerical results obtained using MATLAB shows
that the output image has high value of peak signal-to-noise ratio (PSNR),
smaller mean square error (MSE) with good compression ratio (CR) for low bit
rate. Index Terms - discrete wavelet transform (DWT), set partitioning in
hierarchical trees (SPIHT), peak signal-to–noise-ratio (PSNR), mean square error (MSE), compression ratio (CR). |
|
CSPA831 |
Design of Low-Power Field Programmable
Analog Array based on OTA in 0.35 µM CMOS Technology Rajendra Kanphade,
Dr. M. Shojaei-Baghini, Manish Patil, Priti Gawande, Nitin Ingole, Manisha
Chhangani, Amit Khade, Jagvir Verma, Sanjeev Ranjan, D. G. Wakde VLSI & Embedded System Design Center, SSGM Analog processing may be utilized for applications
requiring low precision, smaller area and lower power than their digital
counterparts, in large part due to their not requiring anti aliasing filters
and signal converters. Fast analog integration in deep sub-micron CMOS
technologies has become a very important issue. The need for fast redesigns
for different technologies of existing analog building blocks becomes crucial
in the IC industry. FPAA is an emerging technology with its own challenges
due to the inherent complexity of programmable and adjustable modular analog
design. In this work a design of an OTA-based FPAA in 0.35 µm mixed signal
CMOS process using Cadence Custom IC Design technology for low-power
applications is presented. This FPAA aimed for low-power applications and
will implement adjustable transconductors, programmable continuous time
filtering and analog multiplication. The most important specifications we
have considered include tuning range, programming versatility, consumed area
and power dissipation. We have also presented a new efficient interconnection
scheme among CABs. Many current mode building blocks can be used for
realizations of linear and non-linear analog systems. Because of the
structural simplicity and convenient controllability of an OTA, it can be
found to be a better candidate. One of the features of OTAs is that their
transconductance can be adjusted and programmed, which is an important factor
in the design of FPAAs. |
|
CSPA832 |
Neural Network Based Detection of Voice Disorders
using Energy Spectrum and Equal-Loudness Contours Paulraj M P1,
Sazali Yaacob 1, S.N. Sivanandam 2, M. Hariharan1 1 School of
Mechatronic Engineering, Universiti 2 Department
of Computer Science and Engineering, Impairment of vocal function can have a major impact
on the quality of life, severely limiting communication at work and affecting
all social aspects of daily life. In
the recent years, voice diseases are increasing dramatically due to unhealthy
social habits and voice abuse. Acoustic analysis is a non-invasive technique
to detect and diagnose the voice disorders. In this paper, a simple feature
extraction method based on band energy spectrum and weighing factor of its
center frequency derived from Equal-loudness contours is proposed. A simple
Elman recurrent network model is developed for testing the proposed features.
The simulation results indicate that the proposed algorithm can distinguish the
voice as pathological or non-pathological voice and provides the mean
classification accuracy of above 90%. The proposed method has the potential
for diagnosing the voice disorders. |
|
CSPA833 |
Adaptive Predictive Fuzzy Logic Attitude
Control R. Nagarajan1,
Paulraj M P1, Sazali Yaacob1, Zaridah Mat Zain1,
Warren Soh Kay Hoh2, Ahmad Sabirin Arshad2, 1 School of
Mechatronic, Universiti 2 Astronautic
Technology (M) Sdn Bhd, No 2 Jalan Jururancang U1/21, Section U1 This paper proposed one axis and three axis attitude
control of a micro-satellite in space. In order to reduce the effect of
unpredictable time delay and variations in other parameters, an adaptive
predictive form of a Mamdani fuzzy logic controller (FLC) is introduced.
Magnetic torquing is considered for generating control torques on this
satellite. This principle is given by the interaction between Earth magnetic
field and magnetic field generated by coils on the satellite for actuation.
To reduce the effects of this unfavorable behavior, a predictive adaptive
form of a Mamdani type fuzzy logic controller (FLC) is introduced. Predictive
controller is needed to compensate the effects of dead time which occurs in
the micro-satellite control system. The predictor estimates the required
control at the next sampling time and applies to the system at the current sampling
time. The response of the satellite is compared with a known reference model
and an adaptive algorithm to yield the error between both outputs towards
zero. The design schemes of modeling adaptive and predictive FLC are
described as follow: basic FLC, predictive FLC and adaptive and predictive
FLC. The simulation results verified the efficiency and accuracy of the
proposed controller and show that three axis control can be achieved with
magnetorquers as actuators in low Earth orbit. |
|
CSPA837 |
A Novel Motion Sensing Algorithm for
Motion-Based Mobile Phone Interface Kok Chen
LIM, Yoong Choon CHANG and Tse Kian NEO This paper introduces a novel motion sensing
algorithm for motion-based mobile phone interface design. In our motion
sensing algorithm, First-Order Image Derivative (FOID) is used as a gradient
measure for featured blocks selection. Optimum-search Block Matching
Algorithm (OBMA) is then proposed for motion estimation. With OBMA, the
average search step in block matching is reduced by up to 30% as compared to
Full-search Block Matching Algorithm (FBMA). The computational intensity of
our algorithm is further reduced by up to 70% with the introduction of Smart
Global Motion Estimation (SGME). Experiment results show that our algorithm
is more efficient to be used in camera-enabled mobile devices. In order to
test on the effectiveness our algorithm has been programmed and tested on the
Nokia N73 phone using Java J2ME programming language. |
|
CSPA0844 |
Implementation of Repetitive Control
Algorithm in Reducing Vibration using MATLAB/Simulink Nor Salwa Damanhuri1, Jari Hatonen2, 1 Faculty of
Electrical Engineering, UiTM 2 School of
Automatic and Control Systems Engineering, The University of Repetitive control algorithm is a new technique to
control systems with periodic disturbances. The purpose of implementing
repetitive control algorithm is to reduce the vibration effect on the system
by making the algorithm converge to zero tracking error. Vibration is a major
problem in the engineering area. There have been major research conducted
using various methods in order to minimize the vibration effects to the
control system. This paper investigates the application of a repetitive
control algorithm technique to control systems with periodic disturbances. In
analyzing the repetitive control algorithm, the FIR assumption is a criterion
that needs to be satisfied so that the algorithm converges to zero tracking
error. The repetitive controller is designed by using the MATLAB/Simulink
software. Then, the results obtained from the MATLAB/Simulink are evaluated. |
|
CSPA847 |
Artificial Intelligent Based Friction
Compensation using Multilayer Feedforward Network Wahyudi and
Nur Muhammad Zaiyad Muda @ Ismail Department
of Mechatronics Engineering, Faculty of Engineering, International
Islamic University Friction is one of the major contributing factors
for problem associated with accuracy in motion control systems. However, it is not easy to overcome the
problem caused by the friction. It is a nonlinear system so that linear
control theory can not handle the friction effectively. Moreover friction
also is characterized by uncertainty due to lubrication condition and/or
payload change. Among the successful
method to overcome for the effect of friction is model-based friction
compensation which relies on the accuracy of the friction model. In this paper, model-based friction compensation
based on multilayer feedforward network (MFN) friction model to compensate
the friction is introduced and evaluated experimentally on a rotary
experimental motion system. The experimental results show that the use
MFN-based friction model is effective to compensate for the effect of
friction in the motion control system. |
|
CSPA848 |
Algorithm Research on Wavelet Support
Vector Machine Lin Jipeng1,
Liu Junhua2 1 School of
Electronics and Control, Chang’an University, 2 School of
Electrical Engineering, A wavelet kernel for support vector machine (SVM)
based on wavelet dual frame theory and conditions of constructing SVM kernel is
presented, which not only increases the precision and convergent rate of the
model, but also is especially suitable for local signal analysis,
signal-noise separation and detection of jumping signals with the
characteristics of multi-scale interpolation and sparse variation, thus
enhances the generalization ability of the support vector machine (SVM),
recognition efficiency and computation burden is alleviated. According to the
wavelet kernel function and the regularization theory, a least square wavelet
support vector machine (LS-WSVM) is proposed to greatly simplify the solving
process of SVM. The LS-WSVM is then applied to the nonlinear system
identification to test the validity of the wavelet kernel function, In
function SINC regression simulation, the max abs error (MSE) is no more 0.18.
In logistic chaos sequence prediction, the MSE is no more than 8×106.
Meanwhile, the MSE is not increased while increasing the predicting length.
This hints it has better generalization ability. |
|
CSPA851 |
Digitally Compensated 4-20 mA Pressure
Transmitter Sidney F. Da Luz1,
Rodrigo Reina Muñoz2, Jorge Ramirez3, Edgar Charry4,
Jaime Lasso V4., Edgar Charry R1. 2Engineering
and Santo André, SP, BRAZIL. 3Electrical
Department- Instituto Superior Politécnico José Antonio Echeverría, CUBA. 4MEMS
- Microssistemas Híbridos de Pressão Indústria e Comercio Ltda., BRAZIL. This paper concerns the implementation of a 4-20 mA
pressure transmitter using digital calibration and compensation techniques.
Digital compensation is presently used as it allows cost reduction in
comparison with analogue techniques like those using laser trimming. As
digital processor was used a commercial chip that provides all functionality
needed to obtain the transmitter, including amplification, A/D and D/A
conversion, an EEPROM to store coefficients and a CPU to control the
activities flow to adjust the sensor’s output signal. The algorithm provided
with the DSP was modified to obtain compensation using current as the main
variable, instead of voltage, as originally used in pressure transducer
applications by the chip’s supplier. It was possible to produce some transmitters
prototypes digitally calibrated and compensated with accuracy under 0.25% FSO
(Full Span Output) in a temperature range of 0oC to 80oC
using 4-20mA current output. |
|
CSPA852 |
Designing a Digital Display Panel for
Motorized N.Sulaiman,
M.R.Mohamed, K.Amat Faculty of
Electrical & Electronic Engineering, Universiti Malaysia Pahang, 25000 Kuantan, Pahang,
MALAYSIA. Electric Vehicle (EV) which is using rechargeable
battery to operate now becoming an alternative type of transportation. The EV
now is in the phase of challenging the conventional vehicle in a factor of
speed, running time, energy efficiency and pollution. Motorized Battery
Operated Vehicle (MBOV) is an electric vehicle that uses battery to operate.
It is designed previously just to make it move forward & backward and
without any display panel. Since it is powered by battery, it is vital that
the battery performance must be monitored. The battery can be ruined by
excessive discharging and charging. MBOV uses 2 batteries with 12 Volts each.
In order to monitor and control the performance of MBOV efficiently, the
digital display panel needs to be designed. It will display four items; (1)
Initial & running value of battery voltage, initial & running value
of current, the speed of vehicle and the remaining minutes before the vehicle
need to be recharged. All the values will be displayed in LED 7 segments. To
display the battery voltage and current, the Analog-to-Digital Converter
(ADC) is used to convert the analog signal to digital signal. Then the
digital signal will be displayed at LED 7 segments by using Binary-to-Decimal
Converter (BCD). The speed is measured by the rate of change in position of
vehicle’s wheel and motor at certain time. The dot/bar display driver is used
to monitor the changing process from the highest to the lowest level of the
time left. Here the initial value of time is 110 minutes base on 24 Volt of
voltage. The time will be reduced until 28 minutes. Every 2 Volts of charging,
the remaining minutes will be displayed in LED 7 segments. Meanwhile, the
voltage will be changed from 24 Volts (Maximum) to 18 Volts. The current are
ranging from 0.0 A to 9.99 A. The results of the research show that all the
parameters needed are displayed correctly such as initial and running voltage
and current, the speed and the remaining time. For future work, it is
recommended to use LCD to replace LED 7 segments and optimize the size of the
circuit by using Printed Circuit Board (PCB). |
|
CSPA866 |
A Practical Approach for Human Motion
Understanding Altahir A.
Altahir, Vijanth S. Asirvadam and Patrick Sebastian Department
of Electrical and Electronics Engineering Universiti
Technologi Petronas, 31750, Tronoh, Bandar Seri Iskandar, Perak D. Ridzuan,
MALAYSIA. With the current advancement in video technology,
video surveillance has become an integral part of daily life. Usually a human
operator watches them to catch events of interest in the scene, but this is a
tedious and time consuming process in which requiring constant attention, and
leads to inadequate surveillance capability. Therefore, there is an urgent
need for automated systems for analysis of surveillance video streams. This
paper presents a novel operational computer vision framework for
understanding human motion from video streams. The system captures monocular
video on a scene containing two humans and performs analyzing and mining
operations based on five measurements to identify the motion. The five
measurements are extracting pixel coverage of a particular object, time
domain, frequency distribution of pixels of interest, crossed distance per
frame and characterizing the events for each one of human subject.
Experimental results shows by mining the measurements mentioned above a
better understanding of human motion can be extracted. |
|
CSPA867 |
Investigation of Ambiguity Resolution for
Single and Dual Frequency Relative Positioning Norsuzila
Ya’acob, Mardina Abdullah and Mahamod Ismail Department
of Electrical, Electronics and System Engineering, Universiti
Kebangsaan Global Positioning System (GPS) is currently one of the
most popular global satellite positioning systems due to global availability
of signal and performance. High positioning accuracy is now achievable using
carrier phase based relative positioning technique, in which two or more
geodetic quality GPS receivers are deployed. Baseline length – dependent
errors due to ionospheric refraction complicates ambiguity resolution and
limits the accuracy attainable in GPS baseline determination. In this study,
the effects of initial phase ambiguity at GPS and modelling of ionosphere on
base components were studied. The solutions that are related to frequency and
solutions were done by using single frequency and dual frequency data. Data from Real Time Kinematic (RTK) and
Malaysian Active Station (MASS) of JUPEM (Jabatan Ukur dan Pemetaan |
|
CSPA868 |
Universal Infrared Receiver for PC’s
Media Player Z. Abdul
Halim and Y.H.Ying School of
Electrical and Electronic Engineering, University
Sains 14300
Nibong Tebal, Seberang Perai Selatan, Infrared data transfer is becoming increasingly more
important. Television remote control uses infrared light, so do PC mice,
keyboard, printers and other peripherals. This is because the infrared
devices offer lower production cost, wide operating range and good
communication security. The PC infrared receiver presented in this project
was a practical application that represents a continuation of the series of
articles on PC interface and Visual Basic. Visual Basic 6 was used to decode
the received infrared signal and perform PC’s functions such as window media
player. Serial port DB-9 (COM1) was used as the interface between computer
and the hardware of this project. By end of this project, total three brands
of remote controls were able to use to remote the PC such as Philips, Sony
and JVC. |
|
CSPA869 |
Attitude Control of Small Satellites with
Reaction Wheels Zuliana Ismail,
Rizal Zahari, Renuganth Varatharajoo Department
of Aerospace Engineering, University
Putra This paper presents the attitude control for small
satellites by using three reaction wheels. To avoid any wheel’s angular
momentum reaching the saturation level rapidly, the secondary actuators such
as magnetorquers are significantly required. Since reaction wheels are
powered by electrical power, it is slightly inappropriate to adopt multiple
reaction wheels for the small satellite missions because of the power
constraints. However, the power
consumption could be minimized by having the best orientation of the reaction
wheels on-board satellite. The purpose of this paper is to investigate the
most suitable reaction wheels configuration in an orbiting satellite.
Numerical simulation is performed by using the MatlabTM/Simulink for a
reference mission. Based on simulation results, the best reaction wheels
configuration that produced the minimum torque which leads to a low power
consumption was identified. |
|
CSPA880 |
Application of Taguchi Method in the
Optimization of RF-Low Noise Amplifier Design Parameters S.F. Wan
Muhamad Hatta, Department
of Electrical Engineering, University This paper outlines the Taguchi optimization
methodology, which is applied to optimize the Radio Frequency Low Noise
Amplifier (RF-LNA) design performance. The LNA design incorporates the
forward body biasing scheme with the power constrained simultaneous noise
input matching (PCSNIM) and the gain enhancement technique in order to boost
design performance. The design parameters evaluated are the gate inductor
value, Lg, the additional PSCNIM-capacitor, Cex, which is located between the
source terminal and the gate terminal of the common source amplifying
transistor, and the feedback capacitor, CF which has been included to realize
the gain enhancement technique. An orthogonal array, signal-to-noise (S/N)
ratio and Pareto analysis of variance (ANOVA) are employed to analyze the
effect of these design parameters. Through statistical analysis of the
experimental results, the optimal variable combination for high gain and low
noise performance are low Lg value and CF values. It has also been realized
that a low value of Cex would give a good power gain performance while a
larger value gives a desired low noise figure. Using the Taguchi method for
design of experiment (DOE), other significant effects such as the interaction
among the design parameters are also investigated. The study shows that the
Taguchi method is suitable to solve the designing problem with minimum number
of trials. |
|
CSPA884 |
Performance Comparison of LQR and PID
Controller for a Ball-beam System A. N. K.
Nasir, M. A. Ahmad and N. Hambali Faculty of
Electrical and Electronics Engineering, Universiti Malaysia Pahang (UMP), 25000 Kuantan,
Pahang, MALAYSIA. This paper presents investigations into the
performance comparison of LQR and PID controller for a ball-beam system. A
ball-beam balancer system with the beam supported in the middle and rotates
against its central axis is considered and a linear time invariant dynamic
model is derived in this study. Simulation results of the response of the
ball-beam system are presented in time domain. Two system responses namely
the ball position and angular beam are obtained. The performances of the LQR
and PID controller are examined in terms of input tracking capability and
time response specifications. A comparative assessment of both control
schemes to the system performance is presented and discussed. |
|
CSPA893 |
A Modified Evolution Equation for Particle Swarm Optimization in
Solving Class Scheduling Problem Mohd Azri Abdul Aziz, Mohd Nasir Taib, Faculty of Electrical Engineering, Universiti Teknologi MARA (UiTM), 40450 Shah Alam, The solution of a class scheduling problem is
generally represented by the arrangement of events and its associated
parameters (lecturer, students, etc.) in timeslot and room, while the
evolution equation for PSO is in a numerical format. Several advantages in
terms of a fast execution of algorithm can be obtained if the format for both
is compatible. It will also ease the process of modeling the class scheduling
problem using Particle Swarm Optimization technique. In order to make the
format for both evolution equation and solution of the problem is compatible,
this paper will propose some modification on the PSO’s evolution equation.
The proposed modified equation will still maintain two important elements of
PSO which are the self-learning (cognitive) and social interaction. The
modified evolution equation was then used in the experiment to prove that it
can simulate the swarm intelligence behaviour as close as to the original
equations. The result shows that the implementation of the modified evolution
equation has been successfully demonstrated for the purpose of solving a
class scheduling problem. |
|
CSPA894 |
PSO with Guided Local Search in Determining the Arrangement of Event
for Class Scheduling Problems Mohd Azri
Abdul Aziz, Mohd Nasir Taib Faculty of
Electrical Engineering, Universiti
Teknologi MARA (UiTM), 40450 Shah Alam, The arrangement of events in class scheduling
problem determine the solution of the problem. The movement of at least one
event to a new slot will create new different arrangement and therefore
changing the solution to the problem. A totally random selection of which
event to be moved and where to move will prone to the duplication of event
arrangements. This will therefore drag the time in finding the best event
arrangement or solution of the problem. This selection method will make the
movement even worst if it violates the hard constraint. This paper will
propose a PSO with guided Local Search in selecting the event to be moved and
also the new slot it can be moved to. This is done to ensure the duplication
of solution is avoided and thus will speed up the process in searching for
the best solution. The checking for hard constraint violation will be done in
the Guided Local Search algorithm, which means the solution passed to the
evaluation process is a hard constraint free and non-duplicated solution. In
the proposed algorithm, PSO will be responsible in selecting the area where
the particle
will move
while the Guided Local Search will complete the search by finding the exact
location of the particle. |
|
CSPA908 |
PSA Approach for Image Search in Fabric
Images Khor Siak
Wang1, Fatimah Ahmad2 1Faculty of
Information and Communication Technology Universiti
Tunku Abdul Rahman, 46200, PJ, 2Faculty of
Computer Science and Information Technology, Universiti
Putra The ability to search a collection of fabrics to
find a particular combination of texture is increasingly important especially
in the area of fashion and interior design. Different designers working in
similar fields may have different ideas on their products such as pattern of
clothes, textures of textiles, placements and colour combinations of floor
tiles, wall papers etc. Instead of producing these works on papers, which may
suffer from the risks of corruption, more designers have now relied on the use
of storage technology to store their works in graphic forms. This does not
only conserve space but also enables them to retrieve their works for
reference whenever they intend to. When the number of stored images grows,
searching for a desired piece of design work is likely to be time-consuming
and painstaking. A large collection of images containing different designs
pattern and textures of certain fashion designs demands for efficient search
system. Of all the visual contents identifiable from a fabric image, texture
is considered to be the commonest visual attribute that aids in image
retrieval. Common approaches for texture-based image retrievals are largely
centered around frequency-based models, which are notably suffered from the
problem of poor retrieval accuracy. In this paper, a novel approach for
texture-based image retrieval has been proposed. This approach, known as
power spectrum analysis, or PSA for short, is based on the idea of dividing
an image into six equally sized partitions where the power spectrum of each
partition is computed to be used to aid in the image retrieval process.
Benchmarked using the popular performance measurement technique, Recall and
Precision, adopting the PSA technique for retrieving images based on their
texture contents has produced remarkable retrieval accuracy, almost three
times more accurate over any other frequency-based models. |
|
CSPA910 |
Keystroke Pressure-Based Typing
Biometrics Authentication System Using ANN and SVM-Based Classifiers Hasimah Ali1,
Momoh J. E. Salami2, Wahyudi2, Mohamad Arinal1 1School of
Mechatronics Engineering, Universiti 2Department
of Mechatronics, IIUM Security of an information system depends to a large
extent on its ability to authenticate legitimate users as well as to
withstand attacks of various kinds. Confidence in its ability to provide
adequate authentication is, however, waning. This is largely due to the
wrongful use of passwords by many users. In this paper, the design and
development of keystroke pressure-based typing biometrics for individual
user’s verification which based on the analysis of habitual typing of
individuals is discussed. The paper examines the use of maximum pressure
exerted on the keyboard and time latency between keystrokes as features to
create typing patterns for individual users. Combining both an Artificial
Neural Network (ANN) and Support Vector Machines (SVMs) are adopted as
classifiers to verify the authorized and unauthorized users based on
extracted features of typing biometric. The effectiveness of the proposed
system is evaluated based upon False Reject Rate (FRR) and False Accept Rate
(FAR). A series of experiment shows that the proposed system that used
combined classifiers produces good FAR and FRR. |
|
CSPA919 |
Backstepping
Controller Design for a Quadrotor Unmanned Air Vehicle A.
F. Mohamed, Atheer L. Salih, and M. Moghavvemi This paper presents the design of a
backstepping controller for a quadrotor vertical take-off and landing (VTOL)
unmanned air vehicle. This controller sets the helicopter to track three
certain positions (x, y, z) and the yaw angle to their desired values. In
addition, it stabilizes the pitch and roll angles. The model of the system is
presented in three interconnected subsystems; 1. an underactuated subsystem
S1 representing the under-actuated subsystem which gives the dynamic relation
of the horizontal positions (x, y) with the pitch and roll angles, 2. a
fully-actuated subsystem S2 that provides the dynamics of the vertical
position z and the yaw angle (z and ψ), and 3. rotors subsystem S3 which
gives the dynamics of the propeller forces (F1, F2, F3 and F4). To stabilize
the overall system, the backstepping controller designed based on the
Lyapunov stability theory. Various simulations of the model show that the
control law stabilizes the quadrotor with good tracking. It generates the
control signals u = [u1, u2, u3, u4] for DC motors of the quadrotor
helicopter ensuring that the position {x (t), y (t), z (t), ψ (t)} track
the desired trajectories {xd(t), yd(t), zd(t), ψd(t)} respectively. The
control commands stabilize the pitch and the roll angles {Ø(t), θ(t)} as
well. However, the last step design has been modified to include a
dynamic of the four rotors, which is used to generate the lift forces. To
verify the performance of the proposed backstepping controller, a classical
PD controller is presented in this paper and the results of the two
controllers are compared. The simulation results show the good performances
and the differences between the tow types of the controllers. |
|
CSPA963 |
Study the Effect of Drain Induced Barrier
Lowering (DIBL) on P-Type MOSFET Device Hashimah
Hashim, Ahmad Sabirin Zoolfakar and M.H.F.M. Safaia Faculty of
Electrical Engineering, Universiti
Teknologi MARA (UiTM), 40450 Shah Alam, This paper presents the study of DIBL on PMOS device
by using SILVACO TCAD simulator. The study is based on the effect of DIBL due
to short channel transistor. The aim is to reduce the effect of DIBL in short
channel device. From the conducted experiment, there are two approaches to
reduce DIBL effect. The two approaches are reducing the quantity of HCL and
reducing the temperature of implantation process. Three different values of
drain voltage, VD are biased in PMOS device to investigate the role of this
voltage on DIBL effects. The DIBL value can be calculated and simulated from
the software. From the simulation results that have been conducted shows that
the value of DIBL is reduced by 1.5% when the quantity of HCL is reduced. |
|
CSPA968 |
An Overview of Dual Band Antenna for WLAN
Application Mohammad
Tariqul Islam1,2, Tan Yee Mun1, Chan Yee Kit1,
Norbahiah Misran2 1Faculty of
Engineering & Technology, 2Institute
of Space Science (ANGKASA), University
Kebangsaan A comprehensive overview of recent developments of
new and innovative antenna designs is presented in this paper. A theoretical
investigation is carried out on microstrip patch antenna for WLAN application
in this research. Investigation is carried out on the basis of design
perspectives, size, impedance bandwidth and cross polarization level of
antenna. Approximate equations for designing patch antenna are illustrated.
In addition, the design flow chart and guidelines for developing dual band
microstrip antenna are also provided. |
|
CSPA974 |
Sample Rate Converter from CD to DAT
using Truncated Lagrange Interpolation Aseim
Najem, Ahmed Elbarsha EE-Department,
In sampling rate conversion between two different
formats, the desired output sample values are the interpolated sample values
at noninteger (or, integer) multiples of the original sampling period. In
this letter, we present a sampling rate conversion from 44.1-kHz compact disc
to 48-kHz digital audio tape formats. This method involves upsampling the
input signal by two, and then passing the interpolated signal through a
fractional delay filter (FDF) based on truncated Lagrange interpolation. This
structure will be more efficient than when we use fractional delay filter
based on Lagrange interpolation, because the proposed filter is
computationally efficient; where, it is based on polynomial formulas, which
have common terms for all coefficients, and maintains high quality of
signal-to-noise ratio as same as in the other method. |
|
CSPA979 |
Dynamic Modelling of a Two-link Flexible
Manipulator M. A. Ahmad1,
Z. Mohamed2 and A. N. K. Nasir1 1Faculty of
Electrical Engineering, Universiti 2Faculty of
Electrical Engineering, Universiti
Teknologi This paper presents dynamic modelling of a
manipulator with two-link flexible manipulator based on closed-form equations
of motion. The kinematic model is
based on standard frame transformation matrices describing both rigid
rotation and modal displacement, under small deflection assumption. The
Lagrangian approach is used to derive the dynamic model of the structure.
Links are modelled as Euler-Bernoulli beams with proper clamped-mass boundary
conditions. Explicit equations of motions are detailed by assuming two modes
of vibration for each link. Moreover, effects of payload on the response of
the flexible manipulator are discussed. Extensive results are included that
validate the theoretical derivation. |
|
CSPA981 |
Integrated
Gas Sensor Based On Infrared Absorption For Monitoring Indoor Air Quality Marianah Masrie1, Anuar
Ahmad2, Ruslan Rolon1, Ramli Adnan1, 1Department of
System Engineering, Faculty of Electrical Engineering, Universiti Teknologi Mara, Shah Alam, Selangor 2Faculty of
Engineering, Universiti Industry Selangor, Selangor Non-dispersive infrared gas sensor has become one of
the methods for gas detection in indoor environment. Most of the sensors have
the capability of sensing only single gas and the light source is usually a
thermal source that has many disadvantages. This paper discussed the
development of an integrated gas sensor based on infrared absorption
technology to measure the concentration of carbon dioxide (CO2) and carbon monoxide (CO) to 1 part
per million (ppm) level for indoor environment. PbSe Photoresistor was used
as a sensor to measure the two gases at the absorption band spectrums of 4.3
µm for CO2 and 4.6 µm for CO molecules in the air. The infrared
lights are emitted by two light emitting diodes (LEDs). These light sources
are much smaller, high speed, low power consumption and does not need optical
filters. Some experimental results will also be presented to show the
performance of the develop prototype system. |
|
CSPA982 |
Analysis of the Execution Time for
Point-to-Point Communication Operations in a Cluster-based Parallel System Mohamed
Faidz Mohamed Said1, Mohd Nasir Taib2 and Saadiah Yahya1 1 Faculty of
Information Technology & Quantitative Sciences 2 Faculty of
Electrical Engineering Universiti Teknologi MARA, 40450 Shah Alam, MALAYSIA Point-to-point communication is fundamental to all
communication subsystems in cluster computers. It forms a basis for efficient
message implementation, especially those that involve software applications.
A point-to-point communication service involves a sender and a receiver. The
model for point-to-point communication involves three parameters: sending
latency, receiving latency, and network latency. The network latency is the
time required to transmit the message across the network and generally
several significant factors may contribute to it, including network
bandwidth, underlying switching mechanism, and blocking time. This research
attempts to measure the blocking time effect accurately by conducting a
benchmarking test in the controlled parallel computing environment. |
|
CSPA983 |
MPI Overlap Performance Results in a
Beowulf Cluster Mohamed
Faidz Mohamed Said1, Mohd Nasir Taib2 and Saadiah Yahya1 1 Faculty of
Information Technology & Quantitative Sciences 2 Faculty of
Electrical Engineering Universiti Teknologi MARA, 40450 Shah Alam, MALAYSIA Performance of Message Passing Interface (MPI)
operations was measured on a small-scale Beowulf cluster system. The library
used for the parallel programming is mpich implementation version 1.2.0. The
tests were run using Linux operating system due to its sophisticated
multitasking virtual memory kernel. When choosing communication operations,
many of the design choices for parallel software architecture are concerned
with the detailed behaviour of its low-level primitives. There are few
studies using these primitives that quantify the potential benefit of overlap
of message transfer with computation. This research attempts to address with
an empirical method to quantify the potential overlap and examine the
possible performance benefit. The results demonstrate that a high potential
tolerance to network latency and bandwidth exist. This allows for a
significant relaxation of network requirements without a consequent
degradation of application performance. |
|
CSPA984 |
Symptoms and Signs in Dengue Patients Utilizing
The Self Organizing Map Tarig
Faisal1, Fatimah Ibrahim1, Mohd Nasir Taib2 1 Department
of Biomedical Engineering, Faculty of Engineering, 2 Faculty of
Electrical Engineering, Universiti Teknologi Mara, 40450 Shah Alam, Selangor,
MALAYSIA. Early clinical diagnosis is needed in order to
identify the level of the disease in dengue patients. However, the physician
cannot distinguish between the dengue fever DF and the dengue haemorrhagic
fever (DHF) at the early stage. Therefore, many statistical studies were
conducted to define the significant symptoms and signs for classifying the
dengue patients at the early stage. However, none of those studies utilized
the unsupervised neural network for visualizing, analyzing and understanding
the complication of symptoms and signs in the dengue patients. Thus, this
study was conducted to investigate the significant symptoms and signs in
dengue patients before the day of defervescence. The self organizing map
(SOM) was utilized to visualize and understands the symptoms and signs of 195
dengue patients. The results show that the bleeding tendency and Abdominal
Epigastic pain are the significant common signs which can be use to classify
the dengue patient in all days. Moreover, Flash Face is the significant sign
presented in days 3 and day 2. Conjunctivitis and Petechiea Rash was
presented on day 2 as a significant sign for the dengue patient
classification. |
|
CSPA990 |
Reduction of Crosstalk by Adding Grounded
Tracks on Printed Circuit Board Ahmed
Alferjani, Ahmed Elbarsha Electromagnetic Interference (EMI) is a growing
problem with new electronic devices using high-speed switching circuits and
transmitting wide-band signals on densely packed printed circuit boards
(PCB’s). Cross-talk can result in the degradation of analog signals, or the
loss of digital information. This paper presents a technique based on
additional traces grounded by vias, which the cross-talk can be minimized.
The circuit analysis code SPICE is used to analyze a model of coupled lines.
The simulation method proposed in this work is time-domain methods with input
signal is trapezoidal signal. |
|
CSPA993 |
Parallel Algorithm for Blind Source
Separation Ibrahim
Elimam Abdalla1, Abubakr Elsidig Mirghani El Hussien1,
Aisha Hassan Abdalla2 1 Al Neelain
University, 2
International Islamic University In Blind Source Separation a mixture of sources,
sounds or any type of signals are to be separated without any prior
information about these sources and the way they have been mixed. Most of the
methods and algorithms that have been developed to tackle this problem are
computationally expensive. Specially those dealing with Convolutive Blind
Source Separation. Here, a new parallel, computationally none expensive
Convolutive Blind Source Separation algorithm is presented and evaluated.
Then parallelization of an already existing Blind Source separation algorithm
is also presented and tested. |
|
CSPA995 |
Hybrid Color Image Segmentation
Algorithm Azilah
Saparon and Sasni Ismail Faculty of
Electrical Engineering, Universiti
Teknologi MARA, Shah The objective of color image segmentation is to
separate spatial regions of an image on the basis of similarity within each
region and distinction between different regions without a priori knowledge
about specific surfaces. It is the first step in image analysis and pattern
recognition. Several techniques can be
used for color image segmentation such as empirical evaluation of various
color spaces, clustering in feature space, and physics-based modeling. This paper
presents some analyses on several color image segmentation techniques and
amongst them are edge detection, histogram threshold and region-based
approaches. This is done by testing
some edge detector algorithms and comparing their limits and their contribution
to image segmentation. The same
process is done on region based algorithms and histogram threshold
approaches. Based on the findings of
each technique, this work is aimed to identify which of the approaches works
best at achieving a system that can reliably split up objects under active
contour or boundary and variable color conditions so that we can have
accurate segmentation results. A procedure to integrate the techniques has
been identified which is the main contours of the image are used to
identify the different regions present in the image and then the segmentation
stage is performed based on the active region model which allows us to take
region and boundary information into account in order to segment the whole
image. This is known as Hybrid approach and from the results; Sobel edge
detector and |
|
CSPA700 |
Classification
Of Electricity Demands Based On Fuzzy Clustering Z Zakaria1 MIEEE, M H Sohod2
MIEEE, 1 Faculty of
Electrical Engineering, Universiti
Teknologi MARA, Shah 2 Engineering
Department, TNB Distribution, Tenaga
Nasional The electricity
industries need new business strategies for providing value added service to
consumer. Analysis of consumer behavior in handling load usage is an
important errand since it could provide the demand characteristic of a
particular consumer. This information could be used in making new marketing
strategies, electricity pricing and policy making. A demand characteristic
could be extracted from the monthly billing data but sometimes it is
insufficient. Besides that, another
method that could be applied to analyze the electricity demands would be by
installing time interval meter, quarter-hourly, half-hourly or hourly at each
of the point’s demands. However, this method is expensive and costly due to
equipment, maintenance and processing cost. Therefore, load profile acquired
by classifying the load curves is seen as the other alternative approach for
providing a satisfactory and most cost-effective approach in analyzing the
consumers’ demand characteristics. This paper investigates the capability of
fuzzy clustering in classifying electricity daily load curve from different
feeders in a particular distribution network. Fuzzy clustering based on fuzzy relation has been employed in this work.
Further, a factor analysis technique namely Principal Component
Analysis (PCA) is used to examine the daily load curve to determine the most
valuable features of the load data prior to clustering process. Results
obtained demonstrate the ability of the proposed method in classifying
electricity demands according to the energy consumption. |
|
CSPA807 |
Evaluation
of Optimized NNARX Structure in Modeling of Steam Distillation Essential Oil
Extraction System Mohd Hezri Fazalul
Rahiman, Mohd Nasir Taib and Faculty of
Electrical Engineering, Universiti Teknologi MARA 40450 Shah This paper evaluates the Neural Network
AutoRegressive with eXogenous (NNARX) structure in modeling the steam distillation
essential oil extraction. The model order will be selected based on
Rissanen’s Minimum Description Length (MDL) information criterion. In the
training of NNARX model, both unregularized and regularized models will be
assessed. There are three regularization levels of the weight decay that will
be implemented in this work. The number of hidden neuron and iteration will
be optimized before the training session. The testing of the trained model
will be based on R2,
adjusted-R2, NMSE, RMSE,
residual histogram and correlation tests. All results will be compared and
evaluated with respect to the testing data.
|
|
CSPA819 |
Characterization of Pharmaceutical Film
using Microwave Non-Destructive Testing Technique Nor Khaizan
Anuar1,3, Mohd Nasir Taib2,3 Wong Tin Wui1,3
and Deepak K. Ghodgaonkar4 1Particle
Design Research Group, Faculty of Pharmacy, Universiti Teknologi MARA, 40450 Shah Alam Selangor,
MALAYSIA. 2Faculty of
Electrical Engineering, Universiti
Teknologi MARA, 40450 Shah 3Non-Destructive
Biomedical and Pharmaceutical Research Centre, Universiti Teknologi MARA, 40450 Shah Alam Selangor,
MALAYSIA. 4Dhirubhai
Ambani Institute of Information and Communication Technology, DA-IICT
Near The applicability of microwave non-destructive
testing (NDT) technique as an optional tool in drug content measurement of
pharmaceutical films was investigated. Hydroxypropylmethylcellulose (HPMC)
was selected as the model matrix polymer with polyethylene glycol (PEG) 3000
as the plasticizer and loratadine as the model drug. Both blank and
drug-loaded films were prepared using the solvent-evaporation method. The
results indicated that the microwave NDT technique is potentially useful in
quantification of drug content via the measurement of power transmission
coefficient (nPTC) (Linear regression values: r value ≥ -0.91 and p < 0.05) of the pharmaceutically
prepared film. Practically, the significance of correlation between the nPTC
values and drug content was not affected by the composition and storage
humidity of films, as well as, the microwave frequency band employed in the
characterization of the films. |
|
CSPA835 |
Automatic Motorbike Engine Fault
Diagnosing System using Functional Link Neural Network in Wavelet Domain Paulraj M
P, Mohd Shukry Abdul Majid, Sazali Yaacob, M. Hariharan and Mohd Zubir Md Zin Universiti The sound of working vehicle provides an important
clue for engine faults diagnosis. Endless efforts have been put into the
research of fault diagnosis based on sound. It offers concrete economic
benefits, which can lead to high system reliability and save maintenance
cost. A number of diagnostic systems for vehicle repair have been developed
in recent years. Expert system for identification and diagnosis of motorbike
engine faults has become very complex and critical task. Artificial Neural Network is a very
demanding application and popularly implemented in many industries including
condition monitoring via fault diagnosis. This paper presents a feature
extraction algorithm that extracts the features from the motorbike engine
noise signal using discrete wavelet transform. The engine noise signals are
decomposed into 5 levels using Daubechies (“db4”) wavelet family. From the
decomposed signals, the energy features are extracted and used to train the
functional link neural network model for classifying the engine faults. The
expert opinion about the faults present in the motorbike is also used to
model the functional link neural network. The developed functional link
neural network model gives a classification accuracy ranging from 80 to 90
percent. |
|
CSPA836 |
The Effect of Acoustic Environment on
Classrooms Speech Intelligibility Paulraj M
P, Sazali Yaacob, Ahmad Nazri, M Thagirarani Acoustic
Applications Group, Universiti Classrooms provide an environment to students to
obtain knowledge through hearing and understanding the lectures delivered by
teachers where a very high level of acoustical quality is required. Most of
the time, noisy and reverberant classrooms may act as a barrier to this
learning process. It is essential that that architects and design engineers
consider the acoustical properties of any venue early in the design process.
In this paper, it is proposed to develop a method that uses neural network to
predict RASTI in university classroom at all listener’s position. In the
proposed technique, a set of Malay words arranged in CVC format is complied
and a speech signal database is created. In a classroom, the speech signal is
presented at a level of 49 to 72 dB with +/- 2dB and the noise at levels of
55, 60 and 63 dB are electrically mixed to yield a signal-to-noise ratio of
-16 to +16 dB. The sound pressure levels are then measured at different
classroom positions. From the measured sound pressure level, RASTI at various
listeners’ positions are acquired. From the experimental data, two simple
neural network models are developed to predict the RASTI in a classroom.
After extensive experimentation and validation of the proposed method, the
results indicate that the proposed technique is shown to be effective for
predicting speech intelligibility using RASTI. |
|
CSPA839 |
Real-Time Digital Tracking Control of
Hydraulic Cylinder Ruhizan
Liza Ahmad Shauri1, Ramli Adnan1 and Mohd Marzuki
Mustafa2 1Faculty of Electrical
Engineering, Universiti
Teknologi MARA, 40450 Shah Alam, 2Dept. of
Electrical, Electronic and Systems Engineering, Faculty of Engineering, Universiti
Kebangsaan Hydraulic cylinder has been widely used as actuator
in industrial applications due to its high mass-force ratio. The nonlinear
properties of hydraulic cylinder had challenged researchers to design a
suitable controller for position control, motion control, and tracking
control. Most researchers have used advanced control approach to improve the
motion or tracking control. Based on these problems, we had done a real-time
digital tracking control studies on hydraulic cylinder using
trajectory-adaptive zero phase error-tracking control (ZPETC) without
factorisation of zeros polynomial algorithm. The control strategy uses a
recursive least square parameters estimation algorithm to compute the optimum
set of controller parameters offline prior the actual control operation by taking
advantage of the available known reference input. The experimental results
obtained show significant tracking performance. |
|
CSPA840 |
Trajectory-Adaptive Zero-Phase
Error-Tracking Control with Advanced Learning Scheme Ramli Adnan1, Mohd Marzuki Mustafa2
and Abd Manan Samad3 1Faculty of
Electrical Engineering, Universiti
Teknologi MARA, 40450 Shah Alam, 2Dept. of
Electrical, Electronic and Systems Engineering, Faculty of Engineering, Universiti
Kebangsaan 3Dept. of
Surveying Science and Geomatics, Faculty of Arc., Planning and Surveying, Universiti Teknologi MARA, 40450 Shah Alam, Selangor,
MALAYSIA. Hydraulic actuators have been widely used in
industrial applications due to its high force-mass ratio and faster response.
Many researchers have used advanced control strategies to control hydraulic
cylinder for application in motion control or tracking control due to its
natural nonlinear property. With intention to improve the motion or tracking
performance effectively, this paper propose a control strategy called the
trajectory-adaptive zero-phase error-tracking control (ZPETC) with advanced
learning scheme. The proposed controller improves the king performance due to
automatically online tuned controller parameters using current reference
trajectory and advanced learning scheme. The simulation results using a
non-minimum phase model from true plant have shown improved tracking
performances. |
|
CSPA841 |
Feature Extraction Based On Energy
Features from Frequency Response of the Vocal Tract Model Using Bandwidth
Approach Paulraj
M.P., Universiti Automatic speech recognition (ASR) has made great
strides with the development of digital signal processing hardware and
software especially using English as the language of choice. In this paper, a
new feature extraction method is presented to identify vowels recorded from
80 Malaysian speakers based on parameters obtained from Vocal Tract Model
based on Bandwidth (BW) approach. Bandwidth approach identifies frequency
ranges from the first peak of every vowel frequency responses. Mean and
maximum energies were calculated from these Bandwidth frequency ranges.
Classification results from Bandwidth Approach were then compared with the
first 3-formant features using Linear Predictive method. A neural network
model was developed to classify the vowels. Results shows that Bandwidth
approach obtained 5.49% higher classification rate than formant features. |
|
CSPA842 |
Translation of Sign Language to Voice
Signal through Head and Hand Gestures Recognition Paulraj M
P, Hazry Desa, Hema C.R., Mohd Hafiz Fazalul Rahiman, M. Hariharan, Wan Mohd
Ridzuan Wan Ab Majid Universiti Deaf and hard of hearing people often have problems
being able to understand and lip-read other people. Often deaf and hard of
hearing people feel left out of conversation or they are actually ignored by
other people. This is very difficult for deaf people to deal with, as they
want to join in but may feel embarrassed to ask people what they are talking
about, to ask for things to be repeated again and again. The sign language is
the fundamental communication method between those who suffer from hearing
defects. In order for an ordinary person to communicate with deaf people a
translator is usually needed to translate sign language into natural
language. In this paper, a simple feature extraction method based on
centroid-magnitude and Discrete Cosine Transform (DCT) are proposed for
extracting the features from the video image of sign language. 10 different
gesture signs are considered in this work.
Two neural network models are developed for the recognition of 10
different gestures. The simulation result shows that neural network provides
a recognition rate of 89%. |
|
CSPA845 |
Estimation of Paulraj M P2,
Fadzilah Hashim1, R. Badlishah Ahmad1, M. Hariharan2
1 School of
Computer and Communication Engineering 2 School of
Mechatronic Engineering University The fusion of high-spectral but low-spatial
resolution multispectral and low-spectral but high spatial resolution
panchromatic satellite images is a very useful technique in various
applications of remote sensing, such as change detection. Recently, some
studies showed that wavelet-based image fusion method provides high quality
of the spectral content of the fused image. However, most of wavelet-based
methods have a spatial resolution of the fused result less than the Brovey,
IHS, and PCA fusion methods. However, wavelet transform could not efficiently
represent the singularity of linear/curve of the images. In this paper, we
introduce a new method based on the contourlet transform which represents
edges better than wavelets and remove the shadows of images. Since edges play
a fundamental role in image understanding, one good way to enhance spatial
resolution is to enhance the edges. So, we could find an automatic threshold
for discarding the weak edges of panchromatic satellite images using information
measure and then strong edges were fused in multi-spectral images. We test
the proposed approach on IRS satellite images and compared our results with
wavelet transform and IHS ones. |
|
CSPA846 |
Artificial Intelligent Based Friction Identification
using Multilayer Feedforward Network Wahyudi,
Ismaila b. Tijani and Nur Liyana Tajul Lile Department
of Mechatronics Engineering, Faculty of Engineering, International
Islamic University It is well known that friction is one of the major
contributing factors for problem associated with accuracy in motion control
systems. Therefore, friction
identification is an important process to be done for compensating the
negative effect of the friction especially when the model-based friction
compensation is used. Many sophisticated friction models have been proposed
by researcher. However the selection of the appropriate friction model for a
certain system will be trouble some in practical application. To overcome
this problem, in this paper, a novel friction identification method based on
the multilayer feedforward network (MFN) is proposed for motion control
system driven by an electric motor. Experimental results show that the
proposed method is effective as friction model. |
|
CSPA862 |
CDMA Coverage and Capacity Optimization Puneet
Kumar and Ankit Dua Included in the primary objective to have usable
coverage are the objectives to insure reliable call set-up, handoff between cells/sectors,
minimize dropped calls and maximize (And balance) call capacity per sector
and carrier. RF power parameters and antenna configurations support the
coverage and call set-up objective. Adjustment of handoff parameters and
channel power adjustments supports the handoff reliability objective. Dropped
call analysis supports the dropped call objective and further exposes
coverage and RF call processing parameter adjustment needs. Continuing live
optimization (analysis of service measurements) supports the final objective
of capacity optimization through longer term studies of larger volumes of
measurement data. |
|
CSPA863 |
RF Regulatory System Puneet
Kumar and Ankit Dua Dept. of
Electronics and Communication The Federal Communications Commission (FCC)
regulates the use of the RF spectrum in the |
|
CSPA871 |
The Development of an Inexpensive
Portable Potentiostat Z. Abdul Halim1, O. Sidek2, M.
Ravichandran3 1 School of
Electrical & Electronic Engineering, Universiti
Sains 2
Collaborative Microelectronic Design Excellence Centre (CEDEC), Universiti
Sains 3 Department
of Medical Microbiology and Parasitology, Health Campus, Universiti Sains Malaysia, 16150 Kubang Kerian,
Kelantan, MALAYSIA. This paper presents a design and construction of a
small, simple and inexpensive programmable potentiostat. The proposed
potentiostat system consists of a signal input part including digital analog
converter, a signal detection part including transimpedance amplifier based
circuit and 8 bit analog digital converter and a signal control part
including 89C52 microcontroller. The system is cost effective especially for
individual or small group experiments. It is also suitable for independent
student project involving field portable electrochemical instrumentation. The
system has been tested using potassium ferricyanide. Data management are
developed using Visual Basic software. Data are collected every one second
and can be visualized in Microsoft Excell for further analysis. |
|
CSPA872 |
Producing Bipolar SPWM Signal Using Low
Cost Microcontroller M. F. Abas,
A. Irawan, N. M. Saad Universiti Due to the overwhelming concern in energy
efficiency, the resistive speed control is now being overshadowed by the
inverter. Inverter technology is now being developed and applied in most
speed controlled ac motored application such as air-conditioning systems,
freezer and refrigerator. Speed control is achieved by varying the output Vac
frequency. The motor speed is directly proportional to the Vac frequency,
thus the higher the Vac frequency, the faster the speed of the motor.
Inverter technology can be mainly classified into three. The first is the
modified-square wave inverter, SPWM inverter and multi-stage inverter. Among
the three, the most difficult signal to produce using a low-cost
microcontroller is of the SPWM inverter. This paper will give introduction to
SPWM, several approached in producing the bipolar SPWM signal and the result
achieved from the approaches. |
|
CSPA873 |
Image Processing of a Banana: Area
Determination via Edge Detection using MATLAB Syed
Khaleel Ahmed, Aidil Azwin Zainul Abidin, Zaipatimah Ali, Wong Bing Yit, and
Zainul Abidin Md Sharrif Department
of Electrical Engineering, Universiti
Tenaga Nasional, The ability to sort agricultural produce based on
quality is very important. It enables the producer to pluck it at the optimum
instant, the seller to obtain maximum value, and the consumer to get the best
quality. This paper addresses one way of determining the face-area of a
banana from its image. The basic idea centers on determining the edges.
Several methodologies of edge-detection are tested, and an optimal solution
reported. A regular digital camera is used for acquiring the image, and all
manipulations are performed in MATLAB. |
|
CSPA874 |
A MATLAB based Face Recognition System using
Image Processing and Neural Networks Jawad Nagi1,
Syed Khaleel Ahmed1, Farrukh Nagi1 and Farrukh Nagi2 1 Department
of Electrical and Electronics Engineering Universiti
Tenaga Nasional, Km7, 2 Department
of Mechanical Engineering Universiti
Tenaga Nasional, Km7, Automatic recognition of people is a challenging
problem which has received much attention during recent years due to its many
applications in different fields. Face recognition is one of those
challenging problems and up to date, there is no technique that provides a
robust solution to all situations. This paper presents a new technique for
human face recognition. This technique uses an image-based approach towards
artificial intelligence by removing redundant data from face images through
image compression using the two-dimensional discrete cosine transform
(2D-DCT). The DCT extracts features from face images based on skin color.
Feature-vectors are constructed by computing DCT coefficients. A
self-organizing map (SOM) using an unsupervised learning technique is used to
classify DCT-based feature vectors into groups to identify if the subject in
the input image is “present” or “not present” in the image database. Face
recognition with SOM is carried out by classifying intensity values of
grayscale pixels into different groups. Evaluation was performed in MATLAB
using an image database of 25 face images, containing five subjects and each
subject having 5 images with different facial expressions. After training for
approximately 850 epochs the system achieved a recognition rate of 81.36% for
10 consecutive trials. The main advantage of this technique is its high-speed
processing capability and low computational requirements, in terms of both
speed and memory utilization. |
|
CSPA881 |
Application of Fuzzy Logic Control for
Roof-Top Bus Multi-Circuit Air Conditioning System H. Nasution1,2,
M. N. Musa2, H. Abdullah2, M. N. W. Hassan2,
M. A. Baharain2, and M. K. Mansour3 1 Department
of Mechanical Engineering, Faculty of Industrial Technology Universitas
Bung 2 Department
of Thermo-Fluids, Faculty of Mechanical Engineering Universiti
Teknologi 3 Department
of Mechanical Engineering, Faculty of Engineering A multi-circuit roof-top bus air-conditioning (AC)
system has been developed. Depending
on the cooling load, either one or more circuits may automatically be
switched on. The speed of one of the running compressors may automatically be
varied in order to fine-tune the response of the system to the continuous
variation of the cooling load. This
paper presents an algorithm developed based on fuzzy logic control (FLC) that
enables selected compressors to run at the appropriate operating speed. The main objective of the experimental work
is to evaluate the energy saving obtained when the fuzzy algorithm, through
an inverter, continuously regulates the compressor speed. The experiments were
conducted, without external load, with temperature set-points for the
conditioned space of 22, 23 and 24oC. Measurements were taken at a time interval
of one minute. The experimental results show that significant energy savings
of approximately 51.39 to 60.62%, 42.30 to 64.35% and 39.14 to 56.91% were
obtained, with respect to the three temperature set-points. Another objective of the study is to
determine the amount of energy saved when a conventional thermostat control
mechanism is applied to the AC system. The thermostat control imposes an
on/off cycle on one selected compressor that runs at the nominal frequency of
50 Hz. Measurements were taken under
the same respective condition of the three temperature set-points. Comparison between the two results shows
that the use of FLC gives more energy saving than the use of the on/off
control. |
|
CSPA882 |
Study of the Gate-Induced Drain Leakage
Current on N-MOSFETS N. Soin and
C. L. Yan Department
of Electrical Engineering, Engineering Faculty, As MOS devices continually scaled down, the hot
carrier effect becomes more significant due to large electric fields and
causes several types of leakage currents. The gate-induced drain leakage
(GIDL) current has been identified as the major drain leakage phenomenon in
off-state MOSFETs. The gate-induced drain leakage current (GIDL) in typical
n-channel MOSFET that occurs in accumulation mode is due to the band-to-band
tunneling (BBT) which is attributed not only to the transverse electric
field, ET, but also to the lateral electric field, EL, in the gate-to-drain
overlap region. The main sources of these electric fields are from the
gate–source (Vgs) and drain–source (Vds) voltages. This paper presents the
simulation study of the impact of gate induced drain leakage (GIDL) current
on the overall leakage of Submicrometer VLSI circuits. It also involves the
analysis of the behavior and mechanism of GIDL of nMOSFETs. Furthermore the
parametric study of the n MOSFET has also been carried out in order to
analyze the GIDL phenomenon. |
|
CSPA885 |
Efficient Hardware Realization for Discrete Pascal Transform using
Matrix Factorization Narison Ronnarongrit, Sorawat Chivapreecha and Kobchai Dejhan Faculty of Engineering and King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, This paper presents the
discrete Pascal transform for digital signal processing, the operations of discrete
Pascal transform is based on Pascal transform matrix which perform as the
operator for signal transformation. The Pascal transform matrix that is used
in the discrete Pascal transform which is proposed in this paper is divided
to 2 types. Generally, the matrix transformation is necessary to have many
multipliers and adders, which is depending on the dimension of used matrix
operator. The factorization of Pascal matrix into binary matrices will allow
the transformation using Pascal matrix to operate without multipliers and
only adders are used. Therefore, the hardware realization for transformation
circuits can be efficient designed by using the butterfly unit for discrete
Pascal transform to establish the whole structure. Moreover, the hardware structure
of two dimensional discrete Pascal transform will be proposed for 2-D signal
processing case. |
|
CSPA886 |
Investigation of Frequency Characteristic in Discrete Pascal Transform
and Its Applications Sorawat Chivapreecha, Ussanai Nithirochananont and Kobchai Dejhan Faculty of Engineering and King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, This paper presents the
investigation of frequency or filtering characteristic that is hidden in the
discrete Pascal transform. The Pascal matrix that used for signal
transformation by discrete Pascal transform in this paper is divided into 2
types, the first type has hidden high-pass filtering characteristic and
another type has hidden low-pass filtering characteristic. This paper will
show the method to investigate those characteristics by the transfer function
of discrete Pascal transform systems will be formulated and used for analysis
the frequency characteristic of such systems. Moreover, the relationship
between the dimension of Pascal matrix and frequency characteristic will be
shown. Finally, the result from the investigation will be applied for both
1-D and 2-D signal filtering applications. |
|
CSPA887 |
Modified Pascal Matrix for Biquad Digital Filter Design and Its Filter
Structure Realization Sorawat Chivapreecha1, Aungkana Jaruvarakul2 and
Kobchai Dejhan1 2 Department of Industrial Electrical Technology, Faculty of Engineering King Mongkut’s Institute of Technology North Bangkok, Bangkok 10800, 1 Faculty of Engineering and King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, This paper presents a
new design method for designing a biquad digital filter which can give five
frequency responses simultaneously, those are lowpass, highpass, bandpass,
bandstop, and allpass filtering. The proposed method is based on Pascal
matrix operation for bilinear s-z transformation, the Pascal matrix is used
for transforming the normalized analog biquad transfer functions to digital
biquad transfer functions. Moreover, the previous Pascal matrix will be
modified to give only one matrix equation for proposed design method instead
of six matrix equations in general case. Finally, the filter structure
realization is proposed for biquad digital filter implementation, which is
correspondence to the proposed design matrix equation. |
|
CSPA888 |
On A Generalized Efficient Heart Sound Classifier System for
Diagnosing Heart Disorders Prashant Kumar, Kishan Kumar Venkatesan Philips Medical Systems, Plilips Innovation Campus, The work aims towards
the development of an automated classifier system for diagnosing heart
disorders. The phonocardiogram signal is first segmented into its primary
components followed by the estimation of the boundaries of primary heart
sounds S1 and S2. The extraction of features is then carried out in the time
domain which is used to train a Multilayer Perceptron neural network using
Back Propagation algorithm. While other techniques rely on an auxiliary
electrocardiogram signal and complex transformed domain features, this work
presents a computationally efficient time domain approach which does not
require any auxiliary signal and the extracted features are very descriptive
of the cardiac events thereby meaningful to the medical community. The classifier
accuracy is evaluated on two datasets: (1) six cycles each of six different
kinds of heart murmurs (2) eight cycles each of seven pathological heart
sounds and the normal heart sound. The classification accuracy is 100 and
84.38 percent for the first and second dataset respectively. |
|
CSPA889 |
Microstepping Drives Approaches to Improve Machine Carriage/Conveyor
Movement A. Irawan, M. F. Abas,
H. Hazha, N. H. Nordin Universiti Malaysia
Pahang, Pahang, MALAYSIA. Due to the microstepping
advantages, the research and development has been done to provide better
technique and signal processing for microstepping drives for each type of
stepper motor. This article will discuss on experiment on two techniques to
provide microstepping for bipolar stepper motor. The study is focused on
controlling microstepping using two methods; “ON”, “OFF” delay control signal
(ODC) and “ON”, “OFF” with pulse width modulation control (OPWMC) signal. The
purpose of the study is to improve the movement of the conveyor or carriage
part of a machine which is driven by a bipolar stepper motor. The experiment
is done on the carriage movement control with two type of driving style;
direct winding control (DWC) and conductive control (CC). The performance is
referred from the distance movement (axis) of the carriage machine
(left-right/right-left). The results are show that the DWC with OPWMC
technique give more precise movement from CC with ODC and DWC without OPWMC
technique to correct the distance movement error and reducing the resonance
occurred during stepper motor drives. |
|
CSPA890 |
On-Device Implementation of an Automatic Filipino Speech Recognition
System Federico M. Ang and Rowena Cristina L. Guevara DSP Laboratory, Department of Electrical and Electronics Engineering University of the This paper describes
the preliminary results in an effort to develop a speaker-independent
Filipino Speech-to-Text (STT) application on a smartphone. The system is composed of a front-end
Symbian C++ application that was built to run on Symbian OS S60 3rd Edition
smartphones. The application covers
significant portions of a standard speech recognition system, namely, capture
of speech input, feature extraction using PLP-RASTA directly converted to
cepstral values, and posterior probabilities estimation using a multi-layer
perceptron. Acoustic and language
models were produced using different LM tools. To obtain the accuracy of the system, it
was tested with a Linux-based decoder (NOWAY) from SPRACHcore, a full source
code release of speech recognition tools from ICSI, which was ported to run
on a Windows-based desktop computer.
Tests were done in a moderately noisy environment. System performance had average recognition
rates of 53.94% and 6.15% for word- and sentence-level recognition,
respectively. |
|
CSPA899 |
Evaluating and Adapting Non-Native Speech
in Isolated Speech Recognizer Noraini Seman, Siti Salwa Salleh and Naimah Mohd Hussin Computer
Science Department, Faculty of Information Technology and Quantitative
Sciences, Universiti
Technology MARA, 40450 Shah Alam, This paper present the pronunciation performance
between native and non-native speakers of Malay sounds in isolated speech
recognizer (ISR). Speaker adaptation methods are combined to solve the
performance decrease that recognizers are faced with native and non-native
speech of speaker-independent (SI) models. Native speaker is defining as a
speaker whose first language is any dialect of Malay language, and who is
speaking in that dialect in conversations is considered to be a native
speaker. Meanwhile, talkers whose first language is not Malay language are
considered as a non-native speaker. Often, speech recognition performance
degrades drastically if the recognizer which has been trained with native
speech is exposed to non-native speech eventhough speech recognition systems
have reached a certain level of maturity. We performed two experiments to
show the recognition accuracy of the baseline models trained with native
dataset was drastically low for the non-native speakers from non-Malay group
than for the native ones. Acoustic deviation has been discovered as one of
important factors affecting the performance of the ISR. In this experiment,
we implement a number of acoustic techniques to compare the performance on
native and non-native speech. We explore how acoustic models can be adapted
to better recognize the non-native speech. The experiments show that there
are many problems arise such as adaptation methods and the non-native
pronunciation pattern that remains to be investigated. In future, it will be
necessary to improve speaker adaptation methods by incorporating more
extensive knowledge of speaker variation at both the acoustic and the
pronunciation level. |
|
CSPA900 |
‘AIRSAR PACRIM’ Program: Its Application
for Environmental and Land Cover Mapping Abd. Manan
Samad1, Sharifah Norashikin Bohari1, Ismail Maarof2
and Khairil Afendy Hashim2 1 Department
of Surveying Science and Geomatics, Faculty of Architecture, Planning and
Surveying, Universiti
Teknologi MARA (UiTM) 2 Universiti
Teknologi MARA (UiTM) Perlis Kampus
Arau, Peti Surat 41, 02600 Arau, PACRIM II is a NASA sponsored science mission to advance
the development of polarimetric and interferometric radar in Pacific Rim
countries including South Pacific island nations, New Zealand, Australia,
Papua New Guinea, Indonesia, Malaysia, Cambodia, Vietnam, Philippines,
Taiwan, South Korea, Japan and the United States. AIRSAR is airborne
synthetic apertures radar which can be operated in three modes; polarimetric
SAR (POLSAR) provides high quality polarimetric data in three frequencies;
cross-track interferometric SAR (XTI or TOPSAR) allows precision digital
elevation information of the earth's surface to be obtained; and along-track
interferometric SAR (ATI) which can be used to detect ocean current
movements. In addition data was collected by a secondary instrument known as
the ‘MASTER’ or MODIS/ASTER airborne simulator. This is a multispectral
imaging scanner with 50 channels, 25 in the visible shortwave infrared
(0.4-2.5 um), 15 channels in the mid-infrared (3-5 um) and 10 channels in the
thermal infrared (7-13 um). The purpose of this study is twofold: first, to
demonstrate the advantages of using AIRSAR data for land cover mapping. And
finally, to improve knowledge about the AIRSAR PACRIM program in |
|
CSPA903 |
An Optimal Boundary Decomposition for
Shape Based Image Retrieval N. D.
Salih, Rosli Besar, F. S. Abas Centre for
Image Processing and Telemedicine (CIPTEM) Faculty of
Engineering and Technology, This paper presents and demonstrates a robust and
simple method for shape representation and description which is capable of
extracting reliable information of the object outline in a concise manner.
The major benefit of this method is that it operates directly on the boundary
image without the need for tracing the boundary pixels and allows for a more
flexible trade-off between the approximation accuracy and the execution time.
Our technique is translation, scale, and rotation invariant. It works well on
different types of shapes. |
|
CSPA906 |
An Effective Method for License Plate
Characters Segmentation Marwan D.
Saleh1, Hakim Mellah1, Ahmed Mueen2, Nbhan
D. Salih1 1. Faculty
of Information Science and Technology (FIST)
Multimedia University (Melaka Campus) Jalan Ayer keroh lama ,75450 Melaka, MALAYSIA. 2. Faculty
of Information Technology (FIT) Jalan
Multimedia, 63100 Cyberjaya, License Plate Recognition (LPR) is a machine vision
technology, used to support an effective traffic monitoring and vehicle
identification without any human intervention. This paper focuses on the
plate characters segmentation, which is a very important module in the LPR
system. The main objective of this module is to separate and extract
characters bounding boxes for easier identification. Our method is designed
for this purpose, depending on the morphological transformation and connected
component analysis. In order to measure the efficiency, our method has been
tested over a large number of images taken under various conditions, and
achieved promising and encouraging results; based on those results, we urge
that the proposed method is a robust, reliable, and fast enough to be used in
real-time systems due to its high-speed and accurate results. |
|
CSPA918 |
Modelling and
Control of a Quadrotor Unmanned Air Vehicle Atheer
L. Salih, Haider A. F. Mohamed, M. Moghavvemi, and S. S. Yang Centre
for Research in Applied Electronics (CRAE) This paper presents the modelling of
a four rotor vertical take-off and landing (VTOL) unmanned air vehicle known
as the quadrotor aircraft. The paper describes the controller architecture
for the quadrotor as well. The dynamic model of the quad-rotor, which is an
under actuated aircraft with fixed four pitch angle rotors, will be
described. The Modeling of a quadrotor vehicle is not an easy task because of
its complex structure. The aim is to develop a model of the vehicle as
realistic as possible. The model is used to design a stable and accurate
controller. This paper presents the development of a PD control method to
obtain stability in flying the Quad-rotor flying object. The model has four
input forces which are basically the thrust provided by each propeller
connected to each rotor with fixed angle. Forward (backward) motion is
maintained by increasing (decreasing) speed of front (rear) rotor speed while
decreasing (increasing) rear (front) rotor speed simultaneously which means
changing the pitch angle. Left and right motion is accomplished by changing
roll angle by the same way. The front and rear motors rotate
counter-clockwise while other motors rotate clockwise so that the yaw command
is derived by increasing (decreasing) counter-clockwise motors speed while
decreasing (increasing) clockwise motor speeds. |
|
CSPA922 |
Development of Low Cost RFID Door Access
System Rajparthiban
Kumar, Aravind CV, Jeyaraj Savarinayagam Faculty of
Engineering, Architecture and Built Environment, UCSI Development of a Low Cost Multiple Access Door System
using RFID is presented. The key objective is to develop and design a system
for attendance, to keep track and securely identify people in private rooms.
A microcontroller is incorporated into the system that integrates the tag and
the reader. The hardware is developed and interfaced with the software. The
testing is carried under real time and a GUI is developed for more
interaction. |
|
CSPA923 |
Development of a Rescue Robot Moey LK, Aravind CV, Rajparthiban R and Loh
Kin Yan Faculty of
Engineering Architecture & Built Environment In this paper the development of a 2D rescue robot
is presented. The rescue robot is capable of performing the surveillance
operation and victim detection in hazardous zones. The rescue robot is
designed to perform 2D movement with sufficient mobility to explore and
access rough, inclined and unstructured environment. The implemented control
uses a microcontroller and a charge coupled device (CCD) is employed for surveillance.
The prototype is developed under a lab environment and the functionality of
the vehicle is verified by experimental testing. |
|
CSPA924 |
Using An Enhanced Antenna Positioning
System for Performance Analysis of MANET J. A. Guma, N. M. Saad Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750
Tronoh, The most important feature of |
|
CSPA927 |
Road Surface Roughness Diagnostic Based
on Vehicle Acceleration by Using Fuzzy Neural Network A.
Soleamani Shahrood When a vehicle traveling on roads with different
roughness and speeds the vehicle vertical
acceleration decrease ride comfort and damage the vehicle. There for
finds this roughness is important. This paper focuses on a problem of
vibration-based condition and rough diagnosis of road. Vibration signals were
acquired and register from a MEMS acceleration sensor that attached to
appropriate place in the vehicle with road different roughness and speeds.
The amplitude of those signals is noisy and dependent in many factors such as
type vehicle, type tire, type road roughness, vehicle speed. In the first
step the acquired signals preprocessing in time domain and classified based
on type roughness. The classified signals separated in learning signals and
testing signals. The main diagnostic features are extracted from the
vibration signals and vehicle speed. Fuzzy logic principle is used as a
roughness diagnostic technique to describe the uncertain and ambiguous
relationship between different roughness and the dynamic vehicle system. A
fuzzy neural network (FNN) was trained by learning signals and used as a
diagnostic classifier. The test signals applying to FNN and observed the
results. The results demonstrate that the developed diagnostic method can
reliably separate different roughness conditions under the presence of speed
variations. |
|
CSPA929 |
Performance Study of Biomedical Signal
Transmission with Wireless Discontinuous Transmission Mode Md.
Shahidul Islam, Rosli bin Besar and Mohammad Tariqul Islam Faculty of
Engineering and Technology Wireless Discontinuous Transmission Mode (WDTX) can
be utilized effectively to minimize noise in the wireless telemedicine system,
especially for ECG signal as a significant portion between the peaks of the
signal is less informative. In DTX, the transmitter is turned off during the
silence period of the transmitted signal and in this paper we present the
performance improvement of ECG signal if such technique is employed. The
performance study is performed through the simulation experiment. Random data
transmission is controlled by the application of switching function based on
the characteristics of ECG signal. The average duration between the peaks of
ECG signal varies from age to age as well as population to population. Few
age groups are considered for our investigation to observe the performance
for each age group. In addition, population groups from deferent geographical
locations including |
|
CSPA930 |
Operational Space Dynamic Control of the
Redundant Manipulator Using Power Module Jeong-Soo
Kim, Sang-Wook Jeon, Jae-Ho Oh and Doo-Sung Ahn The
Department of Intelligent Mechanical Engineering, The redundant manipulators provide the execution of
complex tasks with flexibility. The redundancy of such manipulators can be
effectively used to avoid obstacle, avoid singularities while performing the
desired end-effector task. The operational space formulation analyzes the
manipulator dynamics as seen from the operational space. However, a dynamic
behavior using the operational coordinates in redundant manipulator is
insufficient to completely specify the dynamic behavior of the entire
redundant system. This paper presents dynamic control law in operational
space that guarantees the tracking of a given end-effector trajectory,
avoiding the obstacle in operational space. The simulation using the
SimMechanics in Matlab toolbox has been performed to evaluate the dynamic
control law in operational space using Power Module. |
|
CSPA933 |
Design and Modeling for Revolute Control
of USBM Simplified Model Mohd
Syakirin Ramli, Reza Ezuan Samin, Mohd Razali Daud and Mohd Ashraf Ahmad Faculty of
Electrical & Electronics Engineering Universiti Universal Stretch and Bending Machine (USBM) is a
combination of Stretch Machine and Bending Machine which are used in car door
sash production. The main purpose of combining these two machines is to
reduce the number of machines, space utilization and increase productivity.
This paper basically focuses on the design and modeling for revolute control
of USBM simplified model. The best controller between modern control
(Integral Control State Feedback Controller) and intelligent control (Fuzzy
Logic Controller) systems that suites the USBM simplified model are
evaluated. The evaluation is done by comparing the performances in terms of
maximum percentage overshoot and steady-state error. Prior to that,
mathematical model of the system is first derived and verified by
SIMULINK/MATLAB. Based on the simulation result, the Fuzzy Logic Controller
is better than the Integral Control State Feedback Controller in terms of
maximum percentage overshoot and steady state error. |
|
CSPA934 |
Unmanned Aerial Vehicle Using Micro
Controllers P.Lakshmi
Swetha And T.Pallavi Jb
Institute of Engineering and Technology Jawaharlal
Nehru Technological University, The use of Unmanned Aerial Vehicles (UAVs) has rapidly
increased over the last years. This has been possible mainly due to the
increased computing power of micro controllers and computers. An UAV can be
used in both civilian and military areas, for example surveillance and
intelligence. With the use of UAVs, the need for a positioning and navigation
system arises. Inertial sensors can often give good position estimation;
however, they need continuous calibration due to error build-up and drift in
gyros. An external reference is needed to correct for this drift and other
errors. The positioning system investigated in this paper is supposed to work
in an area defined by an inverted cone with the height of 25m and a diameter
of 10m. A comparison of different techniques suitable for position sensing
has been performed. These techniques include the following: a radio method
based on the Instrument Landing System (ILS), an optical method using a
Position Sensing Detector (PSD), an optical method using the Indoor GPS
system, a distance measurement method with ultrasound and also a discussion
of the Global Positioning System (GPS). An evaluation system has been built
using the PSD sensor and tests have been performed to evaluate its
possibilities for positioning. Accuracy in the order of a few millimeters has
been achieved in position estimation with the evaluation system. |
|
CSPA935 |
Wavelet Based Real Time Ventricular
Arrhythmia Detection and Classification System Komal Moazzam Dar1,
Yasir Mubeen Dar2, Fahad Moazzam Dar3 1Department of Computer Engineering, National 2Department of Electrical and Computer Engineering, National 3Department of Mechatronics Engineering, National The issues of “research required in the field of
bio-medical engineering” and “the severity of arrhythmias” are attracting
attention of regulatory bodies and the common people in various parts of the
globe. Appropriately flexible, efficient, sensitive and reliable software
architecture, allowing the concurrent processing of different ECG signals
coming in real time as an input to the system is the requirement of the day.
As already used systems cannot fulfill the challenges of fast real time
signal processing and precise and accurate detection, the design and
implementation of a ventricular arrhythmia detection algorithm, processing
the real time ECG signals as an input is presented in this paper. The
algorithm processes these signals using signal processing techniques and
detects ventricular arrhythmias in such ECG signals. This will help the
cardiologists and Electro Physiologists in general and the people suffering
from the heart diseases in particular. In this study we have developed a
wavelet based algorithm for detecting and classifying Ventricular arrhythmia
in real time ECG signals. Wavelet transform has emerged as a powerful time
frequency analysis tool, favored for the interrogation of non stationary
signals. As we are processing real time signals so the algorithm required
high processing rate and we have achieved this goal by reducing the amount of
data to be processed by using wavelet decomposition techniques. We have used
Daubechies wavelet of length four and have accurately detected ventricular
flutter, ventricular fibrillation, ventricular tachycardia and
supraventricular tachycardia. Ventricular Flutter (VFL) and Ventricular
Fibrillation (VFIB) have quite similar characteristics so the algorithm first
of all checks for these two. For this the relative energies of D4 and D6 are
compared. Then processing for the detection of Ventricular Tachycardia (VT)
and SupraVentricular Tachycardia (SVT) is done. For this purpose the relative
energies of D6 and D7 are compared. |
|
CSPA940 |
Fast
Construction of Disposable Prefix-Free Codes Danny Dub´e
and Vincent Beaudoin Universit´e
Some data compression
techniques use large numbers of prefix-free codes. The following two techniques
do so: adaptive Huffman encoding and bit recycling. Adaptive Huffman encoding
allows successive symbols to be encoded where each one is encoded according
to the statistics of the symbols seen so far. Bit recycling, on the other
hand, is a technique that is designed to improve the efficiency of a certain
class of compression techniques (that is, the ones that allow for the
existence of multiple encodings of the same data) and that repetitively has
to build prefix-free codes that are used to encode or decode only one symbol.
In the case of adaptive Huffman encoding, the simple but inefficient solution
consists in building a prefix-free code from scratch according to the current
statistics (using, say, Huffman’s algorithm) before encoding each symbol.
However, there exist efficient algorithms for adaptive encoding that take
advantage of the fact that the statistics evolve only progressively (e.g.,
Vitter’s algorithm). Bit recycling, on the other hand, is unlikely to reuse
the same, or even a similar, prefix-free code. Consequently, a lot of
prefix-free codes need to be constructed from scratch. What we propose is to
use a fast technique to construct prefix-free codes. The technique trades
speed in exchange of the optimality of the prefix-free codes it builds. We
measured that the technique is 3 to 4 times faster than Huffman’s algorithm,
while the encodings of the symbols are only 4% or 1.4% longer on average,
depending on whether the technique is used in a general context or in a
bit-recycling one, respectively. |
|
CSPA942 |
Object
Recognition Using Correlation Technique Sangeeta Kakarwal, Shubhangi Sapkal, Pradeep Ahire and Dr. D. S.
Bormane While it may be obvious that people are capable of recognizing
objects under many variations in conditions. This paper addresses a number of
fundamental issues relating to object recognition. Object recognition is the
sub field of Computer Vision whose goal is to recognize objects from image
data. Applications are many and include industrial machine vision, medical
image analysis, and content-based image retrieval. Extracting the features
using correlation performs object classification. In this paper we have done
object recognition using correlation. The correlation problem is to find all
places in the image that match a given sub image and is easy to implement.
The images are compressed and decompressed that is they are transformed for
more accuracy for recognition of images for better result. For the compression
purpose we have used wavelets which provide good compression and fast
operation for the transformation. Genuine as well as imposter images are used
for testing. |
|
CSPA944 |
HOS-Based
Non-Minimum Phase MA Parameter Estimation using Genetic Algorithm M.Lankarany1,
M.H. Savoji2 1 Department
of Electrical Engineering, Shahrood 2 Department
of Electrical Engineering, In this paper we present a novel method for
identification of linear time invariant, non-minimum phase (NMP), FIR
systems, when only output data are available and the order of filter is
higher than four. We generally model a non-minimum phase system as an MA
model of known order. To estimate the parameters of our model, we exploit 1-D
diagonal slice of third order cumulant of output, which may be contaminated
by additive, zero mean, Gaussian white noise of unknown variance. This method
is based on a new point of view on the third order cumulant equation and
using recursive optimization method in order to identify un-known parameters
by minimizing the sum of squared differences between the observed cumulant
(diagonal slice) and the cumulant of proposed model. We propose both
analytical and optimization-based solutions for identifying the filter
coefficients and using analytical-based solution as a view point of applying
our recursive optimization algorithm which causes to obtain parameters recursively.
Also Genetic Algorithm optimization is considered as optimization method in
this paper. Moreover Experimental results indicate the robustness and
accuracy of proposed algorithm for high order systems. |
|
CSPA945 |
Deconvolution
of non-Minimum Phase FIR Systems Using Recursive Genetic Algorithm M.Lankarany1,
M.H. Savoji2 1 Department
of Electrical Engineering, Shahrood 2 Department
of Electrical Engineering, The problem of estimating the input sequence of a
known, non-minimum phase system is considered in this paper. The proposed
method is based on minimizing the sum of squared differences between the
original and the estimated output. The estimated output is obtained by
exciting the system with un-known input signal which begins with initial
values and is updated step by step in order to minimize the mentioned error.
New viewpoint of the convolution equation allows to: 1) identify the un-known
parameters of the input sequence recursively and 2) apply any optimization
algorithm in the deconvolution problem. Genetic Algorithm optimization is
considered in this paper because of its power in searching the entire
solution space with more probability of finding the global optimum. This
approach covers the deconvolution of the both FIR and IIR non-minimum phase
filters. Also simulation results show the accuracy and simplicity of our
proposed algorithm in the deconvolution of the non-minimum phase, high order
FIR filters common in seismic and speech signal processing. |
|
CSPA947 |
Intelligent Active Force Control of a
Manipulator Actuated by McKibben Artificial Muscles Using Fuzzy Logic Hossein Jahanabadi, Musa
Mailah, Mohd Zarhamdy Md Zain Department of Applied
Mechanics, Faculty of Mechanical Engineering Universiti Teknologi The use of pneumatic
artificial muscle (PAM) actuator has achieved increased popularity since it
can provide inherent safety and mobility assistance to humans performing
tasks due to its high strength and power/weight ratio, low cost, compactness,
ease of maintenance, cleanliness, readily available and cheap power source.
However, the complex nonlinear dynamics of PAM manipulator make it a
challenging and interesting system for modelling and control design. The
problems with high nonlinearity and hysteresis of PAM have made it
complicated to comprehend accurate position control. In this paper, a configuration
of a PAM actuator comprising a pair of McKibben muscles that can be operated
to imitate a bicep-triceps system is described and experimented by means of a
simulation study. The muscles are modelled to actuate a robotic arm
incorporating a feedback control method using an active force control (AFC)
strategy with a fuzzy logic (FL) mechanism embedded in its control loop. It
is applied as an efficient method to robustly and precisely control the
motion of the link. This is on top of proportional-integral-derivative
(PID) controller that is already integrated in the outermost control loop.
The results of the simulation study demonstrate the intense robustness and
effectiveness of the proposed control scheme in countering the loading and
operating conditions compared to the PID control scheme. |
|
CSPA948 |
Wildfire
Risk Assessment Using Fuzzy Artificial Neural Networks Estimation M. Reza Mashinchi, M. Hadi Mashinchi, Ali Selamat Faculty of Computer Science and Information System, Universiti Teknologi Wildfire, which is referred to uncontrolled fire, is
a major issue in natural resource management. The facts of costs exist behind
and decades of fuel accumulation from fire suppression efforts, lead more researches
on this issue. In this regard, one of the research objectives is to make
fewer costs via suppression efforts by analysis of the wild fire risk. In
order to that, factors of wildfire are used to define where and when it would
likely occur. Fuzzy values are used for linguistic variables, which are among
the factors used for risk analysis. Therefore, regarding abilities of fuzzy
artificial neural networks (FANNs) on dealing with linguistic variables, it
is used to construct the wildfire risk estimator support system. |
|
CSPA949 |
Hand
Gesture Recognition using Rotation Invariant Ratika Pradhan1, Prateem Chakraborty2,
Prashant Sarawgi2, Ankit Mehrotra2, Gaurav Agarwal2 1Reader,
2Final Year B.Tech Department of Computer Science and
Engg., SMIT, Hand
Gesture Recognition (HGR) system are very useful for deaf and dumb people to
interact or communicate with. Moreover it has great potential in wide number
of application that includes: Human Computer Interaction, Medical Imaging,
Remote device handing, etc. In this
paper we have presented the work to recognize sign language for learning
numerical digits -0(one) to 9 (nine) correctly and efficiently. The system
takes gesture input from the real world and compares the gesture representing
any one of the sign language stored in the database. And then retrieves the
closest match form the database. In this work we have used rotation invariant
method to classify and recognize correctly the input gesture. In this paper,
we propose a theoretically and computationally simple approach which is
robust in terms of grayscale variations and which is shown to discriminate a
large range of rotated textures/pattern efficiently. We present a gray-scale
and rotation invariant texture operator based on local binary patterns.
Rotation invariance is achieved by recognizing that the gray-scale invariant
operator incorporates a fixed set of rotation invariant patterns. The method
used to recognize the gesture gives satisfactory results for almost all the
sign gesture used for numerical digits. |
|
CSPA951 |
Comparison
of Brain Wave Signals between Electrical Engineering Students and Sport
Science Students of Universiti Teknologi MARA Using EEG Zunairah Haji Murat, Mohd Nasir Taib, Zodie
Mohamed Hanafiah, Ros Shilawani S. Abdul Kadir and Husna Abdul Rahman Faculty of Electrical Engineering, Universiti Teknologi The
purpose of this research is to compare the brainwave patterns of electrical
engineering students (EE) and sport science students (SS) of Universiti
Teknologi MARA using Electroencephalogram (EEG) for four frequency bands:
Delta, Theta, Alpha, and Beta. EEG signals were captured using two-channel
bipolar connection from 42 participants (21 EE and 21 SS) in a controlled
environment. The signals were filtered and classified into the four frequency
bands and graphs were plotted. Furthermore, paired T-test was used to show
the correlation between the left and right hemisphere for verification of
brainwave balancing. It was observed that in general, the EE had higher
brainwave levels for all frequency bands. In addition, for the brainwave
balancing, EE showed higher correlation coefficient implying that the
brainwave is more balanced. |
|
CSPA952 |
FPGA
Design of Boyer-Moore Algorithm for Spyware Detection Mohd.
Yamani Idna Idris1, Emran Mohd. Tamil1, Madihah Mohd Saudi2,
Wong Lim Chien1, 1Faculty of
Computer Science & Information Technology, 2Faculty of Science and Technology, Islamic Spyware
is becoming increasingly destructive and sophisticated in threatening the
privacy and efficient computing of the internet users. This new paradigm has
opened the path for vendors to stop or substantially reduce the spreading of
this uninvited program. Software and hardware approach has been identified to
be implemented in order to lessen the threat. However, software approach is
known to take many CPU cycles making it less efficient technique in the age
of high speed gigabit network. Hence, the author proposed the implementation
of a hardware implementation of spyware detection using configurable hardware
called Field Programmable Gate Array (FPGA). The FPGA apply Boyer-Moore
algorithm which is a pattern matching algorithm in order to detect pernicious
spyware. This paper also provides the detailed evaluations of time complexity
for processing spyware. |
|
CSPA953 |
Review
of Low Power Bandgap Voltage Reference for CMOS Mixed Signal Process Asmah Truky, Norhayati Soin, Sairan Sakrani Department of Electrical Engineering,
Engineering Faculty, Bandgap
voltage reference is an extremely important building block in analog and
mixed signal IC. It has been widely used to support many high-performance
electronic circuitries found in many applications such as cellular
telephones, laptops and other consumer electronics products. The bandgap
voltage reference is required to exhibit high power supply rejection ratio
value and low temperature coefficient and is probably the most popular
voltage reference used nowadays. As IC
design is now dominated by low power, CMOS has become the technology of
choice. Since the bandgap reference is bipolar in nature, solutions are
required to create the reference without having to pay higher cost for more
exotic fabrication process such as Bi-CMOS.
This paper overview the design methods and approaches of low power
bandgap voltage reference. The details of the methods will be described in
this paper. |
|
CSPA954 |
Review
of Lithium-ion battery charger system and Electrostatic Discharge (ESD)
safety devices Khairi bin Omar, Norhayati Soin1,
Wan Nor Liza Mahadi1 and Sairan Sakrani2 1 Department
of Electrical Engineering, Faculty of Engineering, 2 ICmic UniKL
Academy, In
this paper, the review of the Lithium-ion (Li-ion) battery charger, charge
system and their electrostatic discharge (ESD) safety devices are presented.
The Li-ion battery charger IC usually encompasses voltage reference, thermal
regulator, charge system, pass element and protection circuit. Earlier work
focused on the usage of the external power pass element such as MOSFET and
Bipolar Junction Transistor (BJT) to pass a very high current. This power
pass element is controlled by the charging system using switch loops
topology. This topology offer smooth transition between the current control
loop and the voltage control loop during the various phases of the charging
process. This Li-ion charging system, which is protected with ESD protection
circuit, protects the power pass element from the electrostatic discharge or
electric over stress. The review focuses on the current charging system, the
main charging components and the ESD protection circuits which are typically
found in the lithium-ion battery charging system. The best power pass element
that offers low power consumption is identified and the different switching
topologies are explained and compared. As the ESD issue is one of the major
concerns in the battery charger IC, the suitable component for the protection
circuit is identified. |
|
CSPA956 |
Sound
Generated During the Compaction of Alumina Reinforced Aluminum Powders Al
Emran Ismail Faculty of Mechanical Engineering
& Manufacturing Engineering, Universiti
Tun Hussein Onn Malaysia, Parit Raja, Batu Pahat, 86400 Johor, MALAYSIA. Many
non-destructive techniques (NDT) are available to evaluate the internal
defects of any materials. During the processing stages these techniques are
very crucial task to be considered because earlier rejection of the compacted
materials capable to save the manufacturing costs before they are experienced
further processing levels. In this work, sound generated during the powder
compaction of metallic and non-metallic powders are studied. Different
reinforcing percentages are used to examine their effect on the sound
generations. Sound detectors are attached on the surface of the mould and
identical sound filtering software package is used to filter out the unwanted
sound. Prior to compaction, surface roughness of the internal mould is
determined and it is followed the industrial value of the surface roughness
and another mould is highly polished. It is found that higher fraction of the
reinforcing elements and the influence of surface roughness affected the rate
of sound generation during the quasi-static powder compactions. |
|
CSPA957 |
Conditioning
of fourth order cross-cumulant slices in blind deconvolution to reduce noise
effect of the severely distorted semsic data Muhammad Shahzad Younis, Ahmed Fadzil
Hani, M.Ahsan Department of Electrical & Electronics
Engineering, Universiti In
this paper; modifications in fourth order cumulant slice have been proposed
to improve the blind deconvolution algorithm.
Cross-Cumulants of fourth order statistics is used to extract deconvolution
operator for non-Gaussian seismic data. For the ideal non-Gaussian i.i.d
data, a cross-cumulant slice consists of a delta function at zero lag. At
lags other than zero, the value of cross-cumulant must be equal to zero. In
case where noise is dominant in the non-Gaussian i.i.d. signals, it is not
possible to get the ideal value of cumulant slice for further processing.
Seismogram is the output of a mixed phase source wavelet driven by the non
Gaussian input signal in presence of additive Gaussian, color Gaussian noise.
Existing HOS based techniques are good in processing of non-minimum phase
system but most of them fails when noise dominates the actual signal. In
regions like volcanic, anhydrite, complex geological areas, it is difficult
to acquire the seismic data with good SNR, and convolution noise is dominant.
Convolutional noise makes it difficult to identify the closely spaced
bedding. Proposed blind equalization technique with certain modifications in
cross-cumulant slices is incorporated to reduce the minimum means square,
maximum distortion and convolution noise effectively. |
|
CSPA958 |
The
use of discrete prolate spheroidal sequences for modeling short time noisy
seismic traces Muhammad Shahzad Younis, Ahmed Fadzil
Hani, M. Ahsan Department of Electrical &
Electronics Engineering, Universiti The
discrete prolate spheroidal sequences (DPSSs) are a set of time-limited,
band-limited, and mutually orthogonal sequences and can be used to overcome the
spectral leakages of the short time signals. Seismic trace is as a result of
convolution of an impulsive non minimum phase source with non Gaussian
reflectivity series. The character of an impulsive seismic source should be
narrow in time and rich in frequency, allowing to image finely layered
geology. Due the very character of earth, it only passes a limited band of
frequencies because of attenuation and absorption of frequency. Problem with
seismic traces is, due to the spreading and frequency absorption of with
depth, it results in blurring the seismic image of the geology. Deconvolution
operation is used to recover the amplitude and phase of the seismic signals.
Due to the requirement of large data length and its ability to enhance the
noise spectrum, its alone use is not effective for the short time
band-limited seismic trances. The use of DPSSs before deconvolution is
required to minimize the spectral leakages and to reduce the Gaussian white
noise. Thresholding of DPSSs is employed to reduce the noise before feeding
the data to the deconvolution operator. The proposed technique is used to
process the noisy seismic traces with quasi-stationary assumption and non
minimum phase character of source wavelet. By using this technique effects of
scattering, boundary value truncation and attenuation of the propagation wave
can be removed from the measured data. Low value of minimum mean square error
indicates the improvement of resolution of seismic signals. |
|
CSPA959 |
Electroencephalogram (EEG)
Brain-Wave Feature Extraction Using Short-Time Fourier Transform Emran
Mohd Tamil1, Haslina Mohd. Radzi1, Mohd Yamani Idna
Idris1, Zaidi Razak1, Azmi Mohd Tamil2 1Faculty of Computer Science and Information
Technology, 2Department of Community Health, Medical Faculty, Universiti
Kebangsaan Electroencephalogram (EEG) signal represent the
brain activity for a subject and give an objective mode of recording brain
stimulation. It has been suggested by several studies that EEG signal can be
used to detect and diagnose several diseases such as CJD, Alzheimer,
dementia, epilepsy and schizophrenia. In this paper, the EEG feature is
extracted using Short-Time Fourier Transform. Short-Time Fourier transform
(STFT) algorithm is an efficient algorithm to compute the discrete Fourier
transform (DFT) and it’s inverse. DFT
provides an efficient means for detecting directionality or periodicity in
the frequency domain and able to remove noise by deleting high frequency
coefficients. The extracted data are
then classified using adaptive artificial neural network. |
|
CSPA960 |
Heartbeat
Electrocardiogram (ECG) Signal Feature Extraction Using Discrete Wavelet
Transforms (DWT) Emran M. Tamil1, Nor
Hafeezah Kamarudin1, Rosli Salleh1, M. Yamani Idna
Idris1, Noorzaily M.Noor1, Azmi Mohd Tamil2. 1Faculty of
Computer Science and Information Technology, 2Department
of Community Health, Medical Faculty, Universiti Kebangsaan ECG
is a method used to measure the rate and regularity of heartbeats to detect
any irregularity to the heart. An ECG translates the heart electrical
activity into wave-line on paper or screen. In this paper, discrete wavelet
transform (DWT) will be used to extract the relevant information from the ECG
input data in order to perform the classification task. Previous study
suggests DWT-based feature extraction technique yields better data. The
Wavelet transform is a two-dimensional timescale processing method. DWT is
suitable for the non-stationary ECG signals as it has adequate scale values
and shifting in time. The data will then be analyzed and classified using
neuro-fuzzy which is a hybrid of artificial neural networks and fuzzy logic. |
|
CSPA961 |
Chip Design for Jawi Features
Extraction and Recognition Using Discrete Wavelet Transform and K-Mismatch
Algorithm Zaidi Razak1, Nurulhuda Abd
Ghani1, Emran Mohd Tamil1, Mohd Yamani Idna Idris1,
Noorzaily Mohamed Noor1, Mohd. Yakub
@ Zulkifli Bin Mohd Yusoff2 1Faculty of Computer Science and Information
Technology, 2Department of Al-Quran & Al-Hadith, Academy of
Islamic Studies, Jawi differs from English or Latin character
in several crucial aspects. Jawi is written from right to left and normally
the letters are connected on the base line This paper will present a feature
extraction and recognition using DWT for offline Jawi character. DWT is
widely used in many areas such as image analysis and digital signal
processing since it can transform the signal into bands with both time and
frequency information. The k-mismatch algorithm will be also being used to
recognize the Jawi character. K-mismatch objective is to find all occurrences
of a short pattern in a long text string with at most k-mismatches. The
system would be developed in hardware based environment using VHDL
programming language in Xilinx to be implemented in Field Programmable Gate
Array (FPGA). |
|
CSPA962 |
Quranic
Verse Recitation Feature Extraction Using Mel-Frequency Cepstral Coefficient
(MFCC) Zaidi Razak1, Noor Jamaliah
Ibrahim1, Emran Mohd Tamil1, Mohd Yamani Idna Idris1,
Mohd. Yakub @ Zulkifli Bin Mohd Yusoff2 1Faculty of Computer Science and Information
Technology, 2Department of Al-Quran & Al-Hadith, Academy of
Islamic Studies, Each
person’s voice is different. Thus, the Quran sound, which had been recited by
most of recitors will probably tend to differ a lot from one person to
another. Although those Quranic sentence were particularly taken from the
same verse, but the way of the sentence in Al-Quran been recited or delivered
may be different. It may produce the difference sounds for the different
recitors. Those same combinations of letters may be pronounced differently
due to the use of harakates. This paper explores the viability
of Mel-Frequency Cepstral Coefficient (MFCC) technique to extract features
from Quranic verse recitation. Features extraction is crucial to prepare data
for classification process. MFCC is one of the most popular feature
extraction techniques used in speech recognition, whereby it is based on the
frequency domain of Mel scale for human ear scale. MFCCs consist of
preprocessing, framing, windowing, DFT, Mel Filterbank, Logarithm and Inverse
DFT. |
|
CSPA964 |
Map
Route Extraction and Shortest Path Algorithm for Public Transport Information
System Mohd Yamani Idna Idris, Emran Mohd
Tamil, Noorzaily Mohamed Noor, Zaidi Razak, Faculty of Computer Science and
Information Technology, The
domination of private vehicle which emerged globally has caused range of setback
such as traffic congestion and pollution. For that reason, people are
encouraged to use public transport systems. In order to encourage such
responses, improved public transport information is proposed. The system is
able to suggest unfamiliar public transport users to choose a route based on
their preferences. Users can choose a shorter route with more frequent
station change or longer route with less station change. The system
implements a shortest path route algorithm into an extracted public transport
raster map to calculate the solution path from the starting station to the
intended station. The raster map has to undergo several image processing
techniques before the shortest path algorithm can be implemented. Based on
the solution path, the system will suggest a bus numbers or rail stations.
The fare of the suggested route will also be displayed for the users’
convenience. |
|
CSPA966 |
Parking Guidance System using RFID and
Image Processing Techniques in WSN Environment
Mohd Yamani Idna Idris1, Yong Yoke
Leng1, Emran Mohd. Tamil1, Nor Zaidi Haron2
1Faculty
of Computer Science and Information Technology,
|
|
CSPA967 |
Implementation
and Optimization of Digital Baseband Signal Processing
Functions on DSP Processor Roslina Mohamad1, Nuzli
Mohamad Anas, Wan Nor Syafizan1, Rosmalini1 and
Kaharudin Dimyati2 1
AdSIGNet Group, Faculty of Electrical Engineering, University Technology MARA, 40450 Shah Alam, 2
Faculty of Engineering, This
paper presents the implementation of digital baseband signal processing with
Output Shifted Coded Modulation (OSCM)[1] function on Digital Signal
Processor (DSP) Kit TMS320C6711. Single DSP chip is used to implement the
baseband processing function due to compactness, low-power consumption and
flexibility. Although FPGA is compact and high performance, it consumes much
more power whereas ASIC has the downside of limited flexibility [2]. C6711 is
a member of the high performance DSP family from Texas Instruments (TI) that
incorporated the real-time kernel known as DSP/BIOS. The design environment
is based on Code Composer Studio IDE (Integrated Development Environment) and
the DSP functions are simulated on 150MHz clock C6700 DSP simulator. The
research begins with designing the DBSP functions and then optimized the
source codes using lookup table method and DSP/BIOS modules. Finally, the optimized
codes are implemented on TMS320C6711 DSK and the codes performances are
collected through DSP/BIOS analysis modules. |
|
CSPA969 |
High Gain Microstrip Patch Antenna Using Multiple Slot Mohammad Tariqul Islam1,
2, Norbahiah Misran2, Md. Shahidul Islam1, Mohd Fairuz Affendy bin Abdul Rahim1, Mohammed Nazmus Shakib1 1Faculty of
Engineering & Technology, 2Institute
of Space Science (ANGKASA), University Kebangsaan In
this paper, a novel design technique for enhancing gain that improves the
performance of a conventional microstrip patch antenna is proposed. This
paper presents a novel wideband probe fed inverted multiple slot microstrip
patch antenna. The design adopts contemporary techniques; probe feeding,
inverted patch structure and stacked multiple slotted patch. The composite
effect of integrating these techniques and by introducing the novel multiple
shaped patch, offer a low profile, broadband, high gain, and compact antenna
element. The result showed satisfactory performance with maximum achievable
gain of about 13 dBi. The design is suitable for array applications
especially for base station. |
|
CSPA971 |
Robust speaker recognition using
missing feature theory for reconstruction of spectral peaks of
autocorrelation sequence Amir Hossein. Hadjahmadi1,
Mohammad Mehdi Homayounpour1, Seyed Mohammad Ahadi2 1Laboratory
for Intelligent Sound and Speech Processing, Department of Computer
Engineering and Information Technology, 2Department
of Electrical Engineering Amirkabir It
is well known that when there is an acoustic mismatch between the speech
obtained during training and testing, the accuracy of speaker recognition
systems drastically deteriorates. Extracting spectral peaks of
autocorrelation sequence of the speech signal is a new front-end diagram in
robust acoustic feature extraction that improves the robustness of |
|
CSPA972 |
Study of Harmonic Distortion in
Loudspeaker Fault Detection Paulraj
M P1, Sazali Yaacob1, R. Badlishah Ahmad2,
M. Hariharan1, Mohamad Radzi Saad1 and Mohd Hafiz Fazalul
Rahiman1 1School of Mechatronic
Engineering, Universiti 2School of Computer and
Communication Engineering, Universiti Traditionally
quality control has been done manually and inspection of loudspeaker faults
is time consuming and causes error in the quality evaluation. In order to
reduce the time consumption and errors in the quality evaluation, in this
work a study of harmonic distortion in loudspeaker fault detection has been
proposed. In this study, three main fault issues in loudspeaker manufacturing
such as Distortion, Resonant Frequency (f0) inconsistence and Sound Pressure
Level (SPL) inconsistent are considered. The faulty and normal loudspeakers
are tested with difference frequency of signal range from 20Hz to 20000 Hz
and 2nd and 3rd Harmonic Distortion curves are plotted.
From 2nd Harmonic curve, the first 3 maximum Sound Pressure Level
and its corresponding frequency are noted. Similarly, the same procedure is
applied on 3rd Harmonic Distortion. These parameters are used as
input features for training the Neural Network. A simple Neural Network model
is developed for the automatic detection of loudspeaker faults. The
experimental results indicate that the proposed method enables the detection
of loudspeaker faults as well as defect causes an improvement of loudspeaker
quality. |
|
CSPA975 |
Multiple Regression Approach for
Inputs Selection in Electricity Consumption Forecasting Dedy Purwanto and
Herman Agustiawan Department of Electrical
Engineering University Technology Multiple
regression approach is used to investigate the effects of the temperature,
humidity, chilled water production, and gas fuel consumption on electricity
energy consumed by UTP. Some coefficients related to the multiple regressions
have been used to represent the energy consumption factors. Based on the monthly energy consumption
taken from year of 2004 to 2006, it is found that the order from the highest
to the lowest factors are chilled water production, gas fuel consumption,
temperature, and humidity whose correlation values are 0.823, 0.565, 0.10,
and -0.001 each. |
|
CSPA976 |
Motion
Coordination of Multi Agent Robots using Feedback Control Salman
Ahmed and Mohd N. Karsit Electrical
and Electronic Engineering, Universiti Teknologi PETRONAS, Bandar Seri Iskandar,
31750 Tronoh, Perak, MALAYSIA. Multi agent robots can be modeled by a nonlinear underactuated
system which is not stable around a point by continuous time-invariant
feedback. Hence, linear control is ineffective, so innovative design
techniques such as feedback control are required. However, most of the
feedback control laws rely on the measurement of all states. In terms of
implementation, all of the states may not be available for measurement.
Hence, observer based feedback control laws are required. This paper presents
a comparative analysis between full state linearized and observer based
feedback laws for the leader robot. The follower robots derive their inputs
based on the control information transmitted by the leader robot. From the
simulation results, it is found that for some of the given trajectories, the
error using observer based feedback law converges to zero. Furthermore, the
full state linearized law globally stabilizes the system whereas the observer
based feedback law results in a K-exponentially stable system. For the follower robots, the input-output
feedback linearized laws minimize the error between the desired and actual
formation. The work presented here is an initial study concerning the design
of observer based feedback laws for multi agent robots. |
|
CSPA977 |
Feedforward
Multilayer Perceptron Supervised Learning for ISFET Sensor Arrays Wan Fazlida
Hanim Abdullah1, Masuri Othman2 1Universiti
Kebangsaan 43600 UKM
Bangi, 2MIMOS Berhad, Taman
Teknologi Malaysia, 57000 Ion-selective field transistors (ISFETs) are
electrochemical sensors that can detect ion activities albeit low selectivity
issues for mixed-ion environments. This paper presents the application of
neural network algorithm as post-processing stage at the outputs of the
sensor array. The role of the neural network is to classify ion activity of
the main ion of interest under the presence of an interfering ion. The
architecture is feedforward multilayer perceptron (MLP) and the supervised
learning algorithm is back-propagation. Generated data, with K+
and Na+ ions mixed by the Nikolsky equation, is incorporated into
the linear drain current MOSFET equation. The algorithm is written in MATLAB.
Results show that learning is achieved at approximately 90% correlation with
mean square error 0.178 for a single hidden layer from an array of potassium
sensors of the same membrane type with ±0.5% drive current fabrication
variation. |
|
CSPA978 |
Early
Tests of Low Cost and Simple Optical Tomography Based on a Non-Invasive
Detection Margi
Sasono and Hariyadi Soetedjo Center for
Integrated Research and Innovation, Optical tomography technique has been developing
rapidly for various applications particularly in medicals. This technique is
considerably interesting due to some safety and low cost source and
components, respectively such of use of commercialized an infrared Light
Emitting Diode (LEDs) and a photodetector. In this research work, the early
test has been carried out to the non-invasive objects of the phantom immersed
in the liquid filed in the cylinder. These phantoms were placed between a
light source of infrared LED and a detector in the home-made tomography
apparatus. From the measurement, the image reconstruction was carried out and
then followed by a filtered back projection method. From this, the 2D image
of phantoms was successfully observed as distinguishable objects. This result
was potentially to be used for medical applications which have advantages of
low cost and simple in operation. |
|
CSPA985 |
Novel Face Recognition Method Using
Bit-Wise Images in Neural Network D.B.L. Bong
and K.C. Ting Faculty of Engineering, Universiti Conventional face recognition usually used gray or
color face images and trained on a pattern recognition algorithm. This paper
presents a novel recognition method of using bit-wise images as input
database and trained using feedforward neural network. Analysis is performed
on single and combined bit-planes and compared to conventional gray image
recognition. Every single pixel in an 8-bits gray level digital image
consists of 8 bits. Among these eight bit-planes, bit-planes 4, 5, 6 and 7
provide better recognition rates than bit-planes 0, 1, 2 and 3. Feed-Forward
Neural Network is used in this paper for training and testing of the
bit-planes. The face database used for evaluation is Yale Face Database.
Testing result shows that bit-planes 4, 5, 6 and 7 achieve over 80% accuracy
after training. Combined bit-planes shows better result in recognition rate,
lower false acceptance (FAR), false rejection (FRR) and half total error (HTER)
rates than gray images. . |
|
CSPA986 |
Contingency
Monitoring and Voltage Collapse Estimation for Iraqi National Super Grid
System Lokman H.
Hassan, Haider A. F. Mohamed, M. Moghavvemi, and S. S. Yang Centre for Research in Applied Electronics (CRAE) This paper focuses on the contingency monitoring and
voltage collapse prediction of the Iraqi National Super Grid System. The
system consists of 22 buses and 35 400 kV overhead transmission lines. In the
last decade, this system’s electrical load went through a rapid growth that
some of the transmission lines came to their thermal limits. It is absolutely
difficult to construct a new overhead transmission lines due to the high
cost, political and environment constraints. In this paper, the most
effective way applied for online system status monitoring and thus voltage
collapse prediction is described. The work presents the stability index that
is based on the concept of maximum possible power transfer through a line
permissible loading condition. The proposed method is applied to calculate
the stability indicator that varies from 0 to 1 utilizing normal power flow
calculations. In addition, the stressed conditions of the lines are detected
and the weak areas prone to the voltage collapse are revealed. |
|
CSPA987 |
Small Signal Voltage Stability
Analysis for Iraqi National Super Grid System Haider A.
F. Mohamed, Lokman H. Hassan, M. Moghavvemi, and S. S. Yang Centre for Research in Applied Electronics (CRAE) Normally, to study the behavior of any power system
when subjected to small disturbances, such as a gradual load changes, the
Eigenvalue analysis method is used. This analysis is considered as a
complementary tool to the usual transient stability analysis. In this paper,
the eigenvalue analysis is used to determine the small signal stability of
the Iraqi National Super Grid System (400 Kv). In the last decade, this
system’s electrical load went through a rapid growth that some of the
transmission lines came to their thermal limits. Therefore, the weak areas of
the system prone to voltage collapse need to be identified and protected
against these collapses. The connection between the stability of the machine
and voltage collapse phenomena are also discussed in this paper. First, the
mathematical model of the power system is formed after establishing the
initial conditions. Then, the high order nonlinear differential equations are
linearized by using the perturbation method. In here, the machine speed
equations of the system are derived and the eigenvalues are obtained from the
state matrix. The contribution of a particular machine in a particular mode
is examined for the small signal stability studies. Based on the system’s condition,
the magnitude of the eigenvalues varied from negative infinity to a maximum
value of zero when the machine speed is said to be instable. To verify the
proposed method, the results are compared with an established stability index
method. |
|
CSPA988 |
Voice Operated Module for Car
Accessories Interfacing Via Speech Recognition Haider A. F. Mohamed, S. S. Yang, M. Moghavvemi, and
Ernie Che Mid Centre for Research in Applied Electronics (CRAE) This paper presents the design, development, and
implementation of a speech recognition system to control the functionality of
car accessories and also the movement of the car. The speech recognition
system is interfaced with one car processing unit that is developed to
communicate with other parts of the car system using Controller Area Network
(CAN). The paper describes the design of the main two modules; speech
recognition and interface modules. The speech recognition module receives
speech command as input. Upon successive recognition of the speech command,
it outputs a corresponding recognized code. On the other hand, the interface
module receives the recognized code, processes it, and sends the
corresponding identifier code to the CAN processing unit. The system is
successfully developed using PIC microcontrollers and the speech recognition
module is trained with multiple versions of the same spoken command to
improve and increase the accuracy of the system. Testing of the developed
system shows promising results that can help physically challenged people who
have lost their hand(s) and/or leg(s) drive a vehicle. |
|
CSPA989 |
Model
Based Sensor Fault Tolerant Control For Uncertain Temperature Control Systems S.S. Yang, Ernie
Che Mid, Haider A.F. Mohamed, M. Moghavvemi Centre for
Research in Applied Electronics (CRAE) This paper proposes a model based sensor fault
tolerant control system for uncertain temperature control. Sensor faults
affect the system’s performance in the closed loop when the faulty sensor
readings are used in a feedback control system. This fault-tolerant control
design consists of two parts: a nominal performance controller and a fault
detection element to provide fault compensating signals to the feedback loop.
The nominal controller can have any given structure that satisfies the
performance specification, such as a PID controller. The detection element
will operate in parallel with the system until a fault is detected. At this
time, the compensation enters the feedback loop; when a sensor failure is
detected, the controller structure is augmented by signals from plant model
to compensate for the fault. This paper describes such a control structure
designed for a temperature control system. The proposed design ensures that
the performance with the nominal controller may be maintained even in the
case of sensor failure. A simulated example is provided to illustrate the
efficiency of the proposed design method. It shows the advantages of the
fault tolerant control system used in temperature control system even under
uncertainty. |
|
CSPA991 |
Initial
Investigation of Brainwave Synchronization after Horizontal Rotation
Intervention Using EEG Zunairah
Haji Murat, Mohd Nasir Taib, Zodie Hanafiah, Ros Shilawani S. Abdul Kadir,
Husna Abdul Rahman Faculty of
Electrical Engineering Universiti
Teknologi This research investigates the effect of one session
of Horizontal Rotation (HR) on brainwave synchronization using EEG. EEG
signals were captured from 42 participants before and after they underwent HR
using two-channel bipolar connection in a controlled environment. The signals
were filtered and classified into the four frequency bands: Delta Theta,
Alpha and Beta. Graphs were plotted and paired T-test analysis was used to
show the correlation between the left and right brainwave before and after HR
to verify the brainwave synchronization. It was observed that, after HR,
brainwaves were more synchronized for Delta, Theta and Beta, whereas less
synchronized for Alpha frequency bands.
In general, there was evidence that HR could synchronize brainwave. |
|
CSPA992 |
An
Automatic Rice Dispenser System Using Microcontroller A. Zabidi, W. Mansor, A. Mohamed Faculty of Electrical Engineering, Universiti Teknologi
MARA, 40000 Shah Alam, Selangor, MALAYSIA. A conventional rice
dispenser system offers manual way of measuring the amount of rice and often
provides incorrect measurements which need to be improved to satisfy the
customer requirements. Furthermore, the rice measurement is based on a cup
size which varies according to the country where the system is manufactured.
Thus, the system cannot be used globally. A solution to these problems is to
employ an automatic system that provides correct measurements and flexibility
in setting the desired amount of rice. This paper describes an automatic rice
dispenser system that allows the user to select the desired measurement, to
view the current selections and to have the exact amount of rice. The system
uses a microcontroller to read the user selection, measure the amount of rice
and sense the removal of rice. A conveyor attached to a motor is used to fill
in the drawer with rice from the rice storage. The measurement process was performed
successfully by the software controlling the data transfer between the
microcontroller and the interfacing circuits. |
|
CSPA994 |
Freeman Chain Code Generation For
Shape Boundary Representation Farah Yasmin Abdul
Rahman1, Aini Hussain2, Nooritawati Md Tahir1 1Faculty
of Electrical Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor
2Department
of Electrical, Electronics & Systems Engineering, Faculty of Engineering,
Universiti Kebangsaan It is well known that the Freeman Chain code is a
simple and a popular technique for representing shape boundary. It encodes
the shape boundary as a sequence of connected line based on its length and
direction. In this paper various techniques to generate the Freeman Chain
Code is discussed. Two approaches namely the Star-Center and Fixed-Window
method are used. Initial results showed that the Star-Center method is suitable
for simple images while the Fixed-Window is apposite for complex images. |
|
CSPA995 |
Hybrid
Color Image Segmentation Algorithm Azilah Saparon and Sasni Ismail 1 Faculty
of Electrical Engineering University Technology MARA, 40450 Shah Alam, The objective of color
image segmentation is to separate spatial regions of an image on the basis of
similarity within each region and distinction between different regions
without a priori knowledge about specific surfaces. It is the first step in
image analysis and pattern recognition.
Several techniques can be used for color image segmentation such as
empirical evaluation of various color spaces, clustering in feature space,
and physics-based modeling. This paper presents some analyses on several
color image segmentation techniques and amongst them are edge detection,
histogram threshold and region-based approaches. This is done by testing some edge
detector algorithms and comparing their limits and their contribution to
image segmentation. The same process
is done on region based algorithms and histogram threshold approaches. Based on the findings of each technique,
this work is aimed to identify which of the approaches works best at
achieving a system that can reliably split up objects under active contour or
boundary and variable color conditions so that we can have accurate
segmentation results. A procedure to integrate the techniques has been
identified which is the main contours of the image are used to
identify the different regions present in the image and then the segmentation
stage is performed based on the active region model which allows us to take
region and boundary information into account in order to segment the whole
image. This is known as Hybrid approach and from the results; Sobel edge
detector and |
|
CSPA996 |
An
Observer Design Of Nonlinear Quarter Car Model For Active Suspension System
Using Sliding Mode Control (SMC) Adizul Ahmad 1, Siti Alawiyah Ismail1,
Norlela Ishak1, Abdul Hadi Abdul Razak1, and 1 Faculty
of Electrical Engineering University Technology MARA, 40450 Shah Alam, 2 Faculty
of Electrical Engineering University Technology The purposes of this
project are to present a new modeling an active suspension system for a
nonlinear quarter car model with observer design in state space form. The
strategy used the Sliding Mode Control (SMC) scheme and the road profiles are
estimated by using an observer design. A quarter car model is applied in the
study and the performance of the controller is compared to the Linear
Quadratic Regulator (LQR). There are three parameters to be obtained in this
study specifically, wheel deflection, body acceleration and the suspension
travel. A simulation works by using computer is performed to prove the
performance of the controller. |
|
CSPA997 |
Performance Analysis on Decoding
Algorithms of Turbo Codes Noni Nadiah binti Md Yusoff1; Muhammad @
Yusoff bin Ibrahim2; Azilah Saparon2 1 UTWINTech 2Faculty of Electrical
Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor,
MALAYSIA. Turbo
codes has been invented for many years back. Many coding theorist and
researchers further improved the performance of these codes. Recently, it’s
performance has been reported to achieve closer to the |
|
CSPA998 |
A
Study on Near Infrared Characteristics in Optical Tomography Application Mazidah Tajjudin1 and Wong Jenn Woei2
1Faculty
of Electrical Engineering, Universiti Teknologi MARA, 40450 Shah 2
Faculty of Engineering, Universiti Industri Selangor, 70000 This paper describes the development and
experimental evaluation of an optical tomography instrumentation system that
has been developed for the measurement of concentration profile of solid
particles in a pipeline. The system employs fan-beams array to create the
optical sensing field interrogating the entire cross-section in the testing
pipe. The report will present
experimental results that explain the characteristics of infrared beam
specifically in a near infrared region. Several aspects were investigated
including the penetration limit and the distance of radiation and also the
effect of beam reflection. |
|
CSPA999 |
Online
Solid Flow Monitoring Using PIC16F873 Microcontroller Mazidah Tajjudin, Nazmy Mahadzir, and Mohd. Hezri
Fazalul Rahiman Faculty of Electrical Engineering, Universiti
Teknologi MARA, 40450 Shah Tomography is one of the methods to identify the flow pattern of solid
particles in a pipeline. In this paper, the development of an optical process
tomography applying a microcontroller (PIC16F873) as a data acquisition
module will be presented. The feasibility of the system will be discussed.
The system under study was developed using four sets of infrared transceiver
to acquire the data of the solid flow. The analog signal measured from
infrared receivers will be converted into digital signal by using built-in
ADC in PIC16F873 and transmitted to 8 bits serial communication (UART). The
signal will be process through computer algorithm to produce the output
display. Some sample output will be captured for analysis purposes. |
ACKNOWLEDGEMENT
VISION
To be the leading survey service provider
in the region
To fore front the industry by providing
competent and
professional offshore geotechnical and
geophysical survey
services while maintaining highest safety
and quality standards.