Ali Akbar Shams Baboli
Improving the Face Recognition Accuracy with the new method Multilinear Discriminant Analysis(MDA) |
Abstract With the improvement of communication and identification and determination of people identity problems for achieving the information, transfer the money and controlling the import and exporting the people of country and different places become important topics to invest in. Biometrics is methods to automatically verify or identify individuals using their physiological or behavioral characteristics. The necessity for personal identification in the fields of private and secure systems made face recognition one of the main fields among other biometric technologies. The importance of face recognition rises from the fact that a face recognition system does not require the cooperation of the individual while the other systems need such cooperation. Feature extraction methods try to reduce the feature dimensions Used in the classification step. There are especially two methods used in Pattern recognition to reduce the feature dimensions; Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA). These algorithms transform the input image data into a 1-D vector, which ignores the underlying data structure so these methods suffer from curse of dimensionality and often leads us to the small sample size problem. For solving these problems we proposed a new algorithm MDA that transform the image into a tensor with its own order. MDA with optimizing the new criterion, DTC, achieves multiple subspaces that the number of these subspaces determined with the order of the tensor. Performance of this algorithm is evaluated with 3 standard databases. With the respect to these results, our proposed algorithms improve the face recognition accuracy and the time of finding these optimum results and also avoiding the curse of dimensionality and the SSS problem. |
Student : Ali Akbar Shams Baboli Superviser: Dr. Rezai rad |
Defense date : Sat. 14/12/1389 Time :12:30 Place: class 303 |
Seyyed Mohammad Sadegh Moosavi
| M.Sc. Thesis Defense | 269 |
Simulation of Iris Recognition Systems by Using Active Contours in Segmentation Stage |
abstract A biometric system provides automatic recognition of an individual based on some sort of unique feature or characteristic possessed by the individual. Biometric systems have been developed based on fingerprints, facial features, voice, hand geometry, handwriting, the retina, and the one presented in this thesis, the iris. Biometrics based on iris has been widely used in recent years because of the useful features of iris. The iris recognition system is to be composed of a number of sub-systems, which correspond to each stage of iris recognition. These stages are segmentation – locating the iris region in an eye image, normalisation – creating a dimensionally consistent representation of the iris region, and feature encoding – creating a template containing only the most discriminating features of the iris. The input to the system will be an eye image, and the output will be an iris template, which will provide a mathematical representation of the iris region. The segmentation stage is critical to the success of an iris recognition system, since data that is falsely represented as iris pattern data will corrupt the biometric templates generated, resulting in poor recognition rates. The unacceptable performance of traditional segmentation methods like Hough Transform in boundary detection leaded to developing Active Contours. Hence, current research is directed toward resolving the drawbacks of some active contours such as Balloon and Greedy in order to segment iris images. In fact, three different Iris Recognition systems have been developed based on the mentioned active contours as well as Hough transform. The simulation results show that the proposed active contours yield better results than Hough transform from both view points of Accuracy and time. Besides, the employed data-base is CASIA Iris Database. |
Student: Seyyed Mohammad Sadegh Moosavi Supervisor: Dr. Ahmad Ayatollahi Referees: Dr. G. Rezaei-Rad , Dr. S. B. Shokoohi , Dr. M. H. Ghasemian-Yazdi |
Defense Date: Saturday, March, 5, 2011 11 AM Location: Class No. 303, Department of Electrical Engineering |
Seyed Mohammad Seyedzadeh
Master of Science Thesis Defence |
Using AES Algorithm to Improve Fault Tolerance in Encryption of Satellite Images |
Abstract Security information in the satellite images is rapidly growing and has been the focus of many researches. Illegal access to satellite networks for the purpose of accessing satellite data has proved the vulnerability of most of these data to Hacking threats. To overcome these threats, the security of satellite data can be achieved by encryption. Currently, most of the Earth Observation (EO) Satellites are equipped with the onboard encryption system for protection of the data that are transmitted to the earth Stations. Regarding to the fact that satellites operate in the Harsh Radiation Environment, each electronic onboard system and typically encryption processors are prone to Single Event Upsets. To increase the Fault Tolerance of the AES algorithm in the encryption of satellite image, this thesis proposes architecture based on the Hamming code error detection and correction method. This architecture can detect and correct single bit fault through the encryption process. The architecture uses a module for the detection and correction of faults in each transformation in each round of the AES algorithm. This module detects and corrects the fault that occurs during the encryption. The fault propagation in the proposed architecture is simulated by a java program and then validated in 5 operation modes of the AES algorithm. Synthesizing of this architecture was done by the ISE software. The value of Efficiency, Overhead, Throughput, Maximum Frequency, and Power were 1.1 Mbps/slice, 16%, 24.2 Gbps, 188.718 MHz, and 1232 mW respectively. These results show that the proposed architecture has more satisfactory performance than the other architectures mentioned in the thesis. |
Student : Seyed Mohammad Seyedzadeh Supervisor: Dr. Sattar Mirzakuchaki Referees: Dr. Houshang Hasibi, Dr. Karim Mohammadi, Dr. Hadi Shahriyar Shahhoseini |
Date of Defense: March 5, 2011 Time:13 Place: Class 305 Faculty of electrical engineering |
Meisam Rezaeian
Department of Electrical Engineering
| Session Defense of MSc Thesis | 269 |
MicroGrid Online Management Using Multi-Objective Optimization |
Abstract: The MicroGrid (MG) concept assumes a cluster of loads and microsources operating as a single controllable system that provides both power and heat to its local area. This thesis presents scheme of optimal management of microgrid using multi-objective genetic algorithms (MOGA) aimed at reduction of cost and emission with satisfying constraints, electrical and heat loads. microgrid structure, system component models, objective functions and problem solving method is described. Proposed MG consists of a diesel generator, a fuel cell, a micro turbine, a wind turbine, a photovoltaic array and battery storage. Recovered heat by micro-turbine and heater can be used to supply heat demand. Also heat penalty, hourly revenue and outgo from sold and purchased power added to the cost function. The main goal of this thesis is to demonstrate impacts of using CHP system, heat penalty and also sale and purchase price changes on management and operation of microgrid. Therefore optimization problem solves as multi-objective for several scenarios to supply loads and satisfy constraints during twenty-four hour period. Simulation by MATLAB demonstrates the accuracy of results on proposed microgrid. |
Student: Meisam Rezaeian Advisor: Dr. Sh. Jadid Opponents: Dr. S. H. Hosseini, Dr. M. Kalantar, MSc. A. Kazemi |
Date: Wednesday, 1389/12/18 Time: 1-3 pm Location: Department of Electrical Engineering Building, Class 303. |
Abolfazl Mosaddegh
Electrical Engineering Department
Master of Science Defense Meeting
Determination of DG Capacity to Maintain the Coordination between Fuses and Reclosers in Order to Prevent the Reliability Degradation |
Abstract The Structure of distribution systems are conventionally radial and for this reason, the downstream feeders of the network are just supplied with the main available source and also the protection systems are designed with respect to radial structure of the networks. If the fault occurred in these networks, the fault current is solely fed by one source. Connecting Distributed Generations (DGs) to networks, its structure is not radial anymore and in the fault situations, the fault current is fed by all available sources i.e. DGs participate to supply the fault current. In this dissertation, two strategies are introduced to maintain the coordination between fuses and reclosers in distribution systems. In the first strategy, the DGs capacity in the network is determined whereas with these capacities, the coordination between protective components of fuse and recloser is maintained correctly. In the second strategy, the protective settings of micro-processor reclosers are changed in order to keep the coordination of protective devices with maximum DGs capacity. The micro-processor reclosers can be performed by multiple operation curves while the simple reclosers only operate with one operation curve. Choosing the test system RBTS bus 2, the simulations are done and the results are derived for mentioned test system. Finally, the reliability calculations are presented and compared in three cases of without any DGs in the network, the network in presence of DG with respect to strategies (coordination is maintained) and the network in presence of DG with maximum DG capacity (miscoordination may be occurred). The results show the effect of maintaining the coordination of protective devices on reliability of distribution systems. |
By: Abolfazl Mosaddegh Supervisor: Prof. Ahad Kazemi Referees: Prof. Seyyed Mohammad Shahrtash; Prof. Davood Arab Khaburi; Prof. Seyyed Hosein Hoseinian |
Date of Defense: Monday March 07, 2011 Time: 17-19 Place: Class 303 Electrical Engineering Department |
Arash Alizadeh Vandchali
Design and Simulation All Optical Switches Based on Nonlinear Photonic Crystal Device |
Abstract The main purpose from designing photonic devices is increasing the effective use of optic fibbers bandwidth. One of the devices to achieve this is using photonic switch. Photonic switches are variously used to increase switching Speed. Recently, photonic crystal switches are variously fascinated due to specific characteristics of photonic crystals in controlling and conduction of light. Here an all optical photonic crystal switch based on MZI has been proposed to achieve maximum switching speed and lowest consuming power. The amount of nonlinearity of each switch can be a good factor in determining its quality and performance. To increase the nonlinearity, here we put a polymeric material with high nonlinearity Kerr effect in one arm of the MZI waveguide Kerr nonlinear index of n2=3 × 10-15(m2/W). Necessary analysis has been done to obtain switch parameters such as length of the arm and required power for switching. Since having a bend with low dissipation is considered in switch design, rectangular structure has been used. With considering this fact that photonic crystal wave guide based on rode has great band width and great group speed velocity. Therefore general structure is selected as rectangular photonic crystal with dielectric rods in background of air. All the results have been verified through the mathematical equations and simulations with two numerical techniques: FDTD and PWE. In most cases identifying the band structure of one or some parameters are important for optimizing photonic crystal designs. Material used here is SiGe and has refractive index of n0=3.6. To find the optimized value for lattice, the centre of this band gap considered as designed frequency, here maximum band gap lattice constant is a=0.602?m then r=0.155 × a=0.0933?m. First we designed 1 × 1 all optical switch based on MZI 1By doing calculations, switching occurs at I= 7(W/mm). Direct coupler has been designed to eliminate ring mode in MZI output. Designed coupler has new structure with vertical bends. Next we designed one 1 × 2 that switching occurs at I= 11(W/mm). Finally we designed one 2 × 2 switch, which switching occurs when input intensity of light exceeds 4.7W/?m and essential phase difference for switching obtained when the intensity of input light reaches 6.5W/?m . These switches have capability to have good performance in integrated photonic crystal circuit and also applications in WDM lattice. |
STUDENT: Arash Alizadeh Vandchali Supervisor: Dr.Shahram Mohammad nejad |
Defence date:11/12/89 time:15:00 |
Mohammad Akbari
Electrical Engineering Department
| Master of Science Thesis Defense Session | 269 |
.jpg)
Performance Improvement of FHSS-Based Cognitive Radio Network in The Presence of Jamming, Using Secure Cooperative Spectrum Sensing Techniques |
Abstract A serious threat to cognitive radio networks which sense the spectrum in a cooperative manner is the transmission of false spectrum sensing data by malicious secondary nodes. This enforces the system to acquire authentication. This issue is beyond regular authentication of data origin which assurances that messages are indeed from who they claim they are from. Actually, the decision maker needs to determine the reliability or trustworthiness of the shared data. This thesis utilizes the potential benefits of cooperative spectrum sensing method in a way that it can become applicable when the CR nodes encounter wireless channel uncertainty and some CR nodes that behave maliciously. To this end, after estimating the channel and the PU signal statistics accurately, they are employed to determine the likelihood of CR sensing reports in a MAP hypothesis test scheme for decision about channel occupancy. Also, a computational trust evaluation algorithm is proposed to determine the trustworthiness of each CR user's data and eliminate the effect of SSDF attackers. Defense against Primary User Emulation Attacker (PUEA) is the second issue that is considered in this thesis. To this end, an adaptive transmission technique for frequency-hop packet transmission is proposed and its performance is evaluated for certain channels with unknown attacker’s signal parameters. The power in the transmitted signal and the rate of the error-control code are adapted in a way that provides an energy-efficient communications. The performance of the adaptive transmission system is evaluated for channels with some PUE attackers that occupy some fraction of the spectrum band with time-varying partial-band interference and propagation loss. |
By: Mohammad Akbari Supervisor: Dr. Abolfazl Falahati Reference Committee: Dr. M. Soleimani; Dr. B. Abolhasani; Dr. A.F. Ahmadi |
Date of Defense: Saturday, February 26, 2011 Time: 3.0 pm Place: Class 303, Electrical Engineering Department |
Milad Ghazal
Iran University of Science and Technology School of Electrical Engineering | 269 |
Robust Observer Design for Stator Winding Fault Detection in Induction Motors |
Abstract In this thesis, two methods are proposed to detect stator winding fault in induction motors. The first method is based on a model based approach and a robust observer is presented to detect stator winding faults despite the load variations. Since the proposed observer should be robust, Unknown Input Observer (UIO) is used as robust observer. In the UIO design procedure, for flexible design, the stator currents and speed of induction motor are assumed as measurement signals. After designing a nonlinear UIO, an Extended Kalman Filter (EKF) is used for speed estimation of the induction motor. Therefore, speed measurement is relieved. Simulation results of stator winding fault with various short-inter-turns are shown to evaluate ability of the proposed approach. Finally, an experimental study on a 5-hp induction motor is performed under simulated short-turn stator winding. The results verify that faults can be reliably detected, and hence the proposed method can be used to improve safety and efficiency of the motors in industrial plants. In the second method, a Wiener Filter method has been proposed for detecting stator winding fault in induction motors. This filter is used for estimating and removing useless components from stator currents in order to detect stator winding fault with high reliability. Since the Park’s vector modulus is sensitive to abnormality in three-phase stator currents, it will be used for detection purposes. Finally, the validation of proposed method will be verified by experimental data from a 5-horespower induction motor. |
Student: Milad Ghazal Supervisor: Dr. Poshtan Referee teachers: Dr. Shahri , Dr. Momeni |
Date of defense:27 Feb 2011 Location: 303 room |
AbbasAli kazemian
Iran University of Science and Technology
Electrical Engineering Department
Defense for Degree of Master of Science in Electronic
| 269 |
Using Partial Differential Equations (PDEs) FOR Image Denoising |
Abstract: In this article, a new method is presented for the elimination noise from images, which makes use of the anisotropic diffusion equations and the wavelet transformation. In the existing methods for image noise elimination that are based on the wavelet threshold marking, the existing approximate coefficients in the wavelet transformation, which contain the overall information of the image, are not affected by threshold demarcation. Therefore, the final image (obtained after the noise elimination) will have a limited improvement. Since noise affects both the detail and approximate coefficients of the image, we apply the noise elimination technique based on the anisotropic diffusion equations, on the image's approximate coefficients in order to alleviate the problem associated with the wavelet-based noise elimination methods. By using the anisotropic diffusion equations, the existing noise data in this sub-band will also be preserved. |
STUDENT: AbbasAli kazemian Supervisor: Dr.Ahmad Ayatollahi |
Defence date:27/11/89 time:16:00 |
Alireza Ghorbani
Iran University of Science and Technology
Electronics Department
Degree of Master of Science in electronics Defense session
Design a current operational amplifier with high CMRR and low power |
Current-Mode processing has attracted great attention in recent decades. Current –Mode circuits are distinguished from voltage-mod ones in two main criteria; 1) Instead of voltage, they process current signals as the main variable and data. 2) Ideally there should be no high impedance nodes in their signal paths. In this thesis the current operational amplifier are introduced and their important specifications are noted. Moreover an overall review is performed on the researches that dealt with improvement in specification of this block. In this thesis various methods are proposed to improve the specification of current operational amplifier which any one is introduced to improve some of the current operational amplifier specifications. Designing of the current output stage with ultra high output impedance (13GΩ) , ultra low power consomptions (86µW) and a current follower with very low input resistance (38mΩ) , low power consomptions (145µW) and high CMRR (110dB) are some of the results of this thesis. HSPICE simulator is used to perform simulation and the simulation results are given to prove the validity of works. |
By: Alireza Ghorbani Supervisor: Dr. Seyed Javad Azhari Jury : Dr. Shamsi; Dr. Ayatolahi; Dr. Abrishamifar |
Sunday 6th February 2011,Time:13:00 Seminar room, first floor, electrical and electronic engineering faculty, Iran university of science and technology (IUST) |