Graduate

Browse

Now showing 1 - 20 of 231

Depth-aware Hand Gesture Recognition for Human-Computer Interaction
(Department of Computer Science and Engineering(CSE), Islamic University of Technology(IUT), Board Bazar, Gazipur-1704, Bangladesh, 1-06-22) Mahmud, Hasan
Hand gestures can be dened as the movement of the hands and ngers in particular orientations to convey some meaningful information. Recently, inexpensive depth cameras have opened ample research opportunities to work with depth-based features in parallel to image-based features. Existing computer vision-based approaches have limitations in capturing depth variations present in the fine-grained gestures and also in the coarse grained. Hence, we got a scope to exploit depth information and use them in the machine learning models to distinguish those hand gestures correctly. In this thesis, we propose a unique depth quantization technique that can effectively distinguish different hand gestures. Using the technique first, we generate contrast varying depth images that can help to extract salient features from gestural images of static gestures. Second, we use depth values to capture hand nger movement information in the Z-direction to discriminate on-air writing tasks of English Capital Alphabets (ECAs). We have used depth-based features, like raw depth values, quantized depth values, and non-depth features like finger joint points in 2D, ngertip coordinates, other derived features from them, then merge these features to generate a unique dataset for testing the signicance of depth features in terms of recognition accuracy. Experiments on both static and dynamic hand gestures showed that the proposed approach gives higher recognition accuracies. Third, to test our proposed method in deep learning settings, we design a depth-aware CNN-LSTM-based deep-learning model to recognize 14 and 28 dynamic hand gestures. The model takes gray-scale varying depth images and 2D hand skeleton joint points as multimodal input. We achieve better recognition accuracies by performing feature-level and score-level fusion techniques in the benchmark dataset.
Development of an Adaptive Neuro-Fuzzy Controller for Speed Control of Induction Motor Drives
(Department of Electrical and Electronic Engineering, Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2012-08-30) Azim, Md. Riyasat
Robust and precise speed control is of critical importance in high performance drive applications. Control schemes for AC machines are complicated as compared to DC machines since they exhibit nonlinear relationships between process variables e.g. speed and manipulated variables e.g. current, torque etc. Direct torque control (DTC) scheme offers faster and simpler control of AC machines with high dynamic performance but without extensively using coordinate transformations and hence with lesser computational burden on the processor. Artificial intelligent controllers (AIC) are capable of overcoming the limitations of the mathematical model dependent conventional fixed gain and existing adaptive controllers. In this thesis, an adaptive neuro-fuzzy inference system (ANFIS) based controller is proposed to improve the dynamic behavior of DTC based induction motor drive which offers the combined advantages of the flexibility of fuzzy logic and adaptability of neural networks. Hence the developed adaptive neuro-fuzzy controller can be utilized to minimize the effects of unavoidable system disturbances such as, system parameter variations, sudden impact of load changes etc. Since the use of intelligent controllers are very limited in the drives industry due to the relevant computational burden on the microprocessor, therefore, in order to reduce the computational burden, linear linguistic variables with an optimum number of membership functions have been selected for the adaptive neuro-fuzzy controller developed in this thesis. The effectiveness of the proposed NFC based DTC scheme of the IM drive is consolidated through the development of a simulation model using MATLAB/Simulink. The results obtained from the simulation of the proposed system are compared with the results from simulation of the same system using conventional proportional integral (PI) controller. The results have shown promising improvements both in the transient and steady state responses of the system. Further, as an integral part of the thesis, a fuzzy logic controlled dynamic voltage restorer (DVR) system has been developed in order to protect the induction motor drive system from the power quality problems (like- voltage sag, swell etc.). The DVR system was developed and simulated in MATLAB/Simulink environment to verify its functionality. The simulation results have confirmed the ability of the DVR system to perform at an expected level when used in conjunction with the induction motor drive. The DVR system kept the induction motor drive performance intact under both symmetrical and asymmetrical system fault conditions.
Development of Improved Cancer Classification Method by Integrating Metadata in Microarray Data Analysis
(Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2012-09-30) Habib, Md. Ahsan
Microarray experiments provide a high throughput to measure expressions of thousands of genes simultaneously. A systematic and computational analysis of this vast amount of data provides understanding and insight into many aspects of biological processes like Single Nucleotide Polymorphism (SNP), genetic disorders, cancer identification. Expression analysis of DNA (Deoxyribonucleic Acid) of microarray experiments is far from straightforward from statistical point of view. Dimensionality problem of microarray data, identifying significant and informative genes or DNA sequences are the prime challenges for cancer classification. Therefore, the aim of the thesis in cancer classification is to integrate metadata for optimal subset of genes that are useful for expert and embedded system design for different types of cancer classification. In this thesis, different filtering and classification algorithms will be compared on different set of data to conclude which will be efficient and effective method for cancer classification from microarray data. For the improvement of cancer classification performance and biological validation of optimal subset of genes, metadata ranking will be evaluated and integrated for cancer classification. The method will achieve better performance based on gene-independent covariance, trustworthy gene feature ranking and metadata ranking factors. The performance of the proposed method will be evaluated by different filtering to overcome the dimensionality problem, minimum number of genes used in classification techniques on publicly available benchmark dataset of ALL, brain, breast, kidney, lung, prostate for intracancer and inter-cancer classification.
Effective Facial Feature Representation Based on Directional Micro-Patterns
(Department of Computer Science and Engineering, Islamic University of Technology, Board Bazar, Gazipur, Bangladesh, 2012-09-30) Ahmed, Faisal
Automatic recognition and analysis of human face allow many interesting appli- cations in biometrics, human-computer interaction, and security industry, such as person identiﬁcation, age estimation, gender classiﬁcation, and facial expres- sion analysis. Deriving an eﬃcient and eﬀective feature representation is the fundamental component for any successful facial recognition system. However, the inherent variability of facial images caused by diﬀerent factors like variations in illumination, pose, facial expression, alignment, and occlusions make classiﬁ- cation a challenging task. Therefore, the aim of the ongoing research in facial recognition is to increase the robustness of the underlying feature representation against these factors. In this thesis, we work toward developing a simple, yet eﬀective appearance- based feature descriptor for representing human facial image. A new local texture pattern, namely the directional ternary pattern (DTP) has been introduced that can eﬀectively capture the texture properties of a local neighborhood and exhibits robustness against illumination variations and random noise. The proposed DTP operator encodes the texture information of a local neighborhood by labeling the edge response values in all eight directions around the center point using three diﬀerent levels. Our encoding scheme employs a threshold in order to diﬀerentiate between uniform and high-textured face regions, and thus, ensures the generation of ternary micro-patterns consistent with the local texture property. The location and occurrence information of the DTP micro-patterns within the facial image is then represented with a spatial histogram, which functions as the facial feature descriptor. i Abstract ii The performance of the proposed method has been evaluated in three diﬀer- ent applications, which are i) facial expression analysis, ii) face recognition, and iii) gender classiﬁcation from facial image. Diﬀerent publicly-available bench- mark image datasets were used for the experiments. We have also compared the performance of the proposed method with some widely-used local pattern-based feature descriptors. Experimental results demonstrate that, the DTP descriptor is more robust in extracting facial information and provides higher classiﬁcation rate compared to some existing feature representation techniques.
Assessment of competence of TVET trainee teachers in the field of ICT
(Department of Technical and Vocational Education(TVE), 2012-11-15) Usama, Kasagga
The importance of ICT in empowering teachers and learners, and enhancing teaching and students’ achievement has been highlighted in several studies. ICT integration in the developing nations needs to be addressed so as to ensure total integration of ICT in the school curriculum. This study examines the level of basic ICT competence of TVET trainee-teachers’ from different OIC countries. The differences in basic ICT competence of the trainee teachers having different educational background were also examined. The data collected through a test were analysed using percentages, means, and t-test statistics. It has been found that the majority of the TVET trainee teachers there were competent in the field of ICT. However, significant difference in performing their ICT competence among the TVET trainee teachers having educational background in Computer Science, Electrical and Mechanical Engineering was found.
A study to identify the barriers of implementing the proposed competency based training programme in TVET institutions in Bangladesh
(Department of Technical and Vocational Education(TVE), 2012-11-15) Rahman, Mohammad Mahmudur
The purpose of the study was to identify the Barriers of Implementing the Proposed Competency Based Training Programme in TVET Institutions in Bangladesh. The descriptive survey was used for the study. Eleven research questions guided the study. The simple random and purposive sampling technique was used. The sample size was 40 from the TVET teachers in Bangladesh. Data were collected from the selected 8 TVET institute in Dhaka district in Bangladesh. Data were collected through the use of a questionnaire which was based on a five point Likert scale. Data were analyzed by using Chi squire test at 0.05 significant levels. The study revealed that poor administrative support , lack of CBT trained up teachers, CBT Technical Teachers Institute, inadequacy of modern laboratory facilities, weak linkage between technical institutes and industry, lack of public awareness and being used to with traditional curriculum emerged as barriers of implementing the proposed competency based training programme (CBT) in TVET institutions in Bangladesh
A study of Technical/Technology Education in The Gambia Upper Basic Schools
(Department of Technical and Vocational Education (TVE), Islamic University of Technology (IUT), 2012-11-15) Ndure, Bai
Historically, technical/technology education has always been perceived as "inferior education". It is time for this perception to change. With the high unemployment rate in the country, and the huge skills shortage, it is imperative for society to take a critical look at the role of technical /technology education in our current curriculum and that the present government looked more closely at the impact that the proper provision of technical/technology education would do to the country as a whole. This study focuses on the current status of technical/technology education provision in the country and the fostering of technical/technology education at the Upper Basic Schools in general. It examines the perceptions of principals, educators and learners towards the following The status of technical/technology education provision in The Gambia. Gender issues with regard to technical subjects. Curriculum appropriateness towards the employability and familiarity with the new FET curriculum Value of technical education for Gambian society Pedagogics related to the technical/technology education curriculum Several important findings that emerged from the data suggested that: Technical education is viewed as very positive by the youth of The Gambia towards addressing unemployment, skills shortages and for the future growth of the country There are gender imbalances with regard to the human resource within the realm of technical/technology education. Lack of suitably qualified educators in the new technical field. Lack of further training of technical education educators for the future Lack of equipment, resources and workshops. Preparedness of educators in meeting the requirements to teach the current FET syllabus. It is imperative for mindset of the general public to change with regard to the perception of the status of technical/technology education in this country, as the youth of today value and respect what technical/technology education has to offer, and for the possibility of what technical /technology education can do for the eradication of the skills shortage dilemma, for addressing the unemployment issue and for the future growth and development of The Gambia.
Numerical and Experimental Investigation of Shock wave Propagation in the Open Atmosphere
(Department of Mechanical and Production Engineering (MPE),Islamic University of Technology(IUT), Board Bazar, Gazipur, Bangladesh, 2012-11-15) Ehsan, Mohammad Monjurul
Shock wave propagation in the open atmosphere has been investigated both numerically and experimentally. Three-dimensional numerical code has been used to solve the Euler equations for the upstream and downstream parameters during propagation in the open Atmosphere. It is observed that due to decay of the shock strength, all parameters (velocity, pressure and density) are decreased with increasing the travel time of the wave in the open Atmosphere. It is found from the numerical images that the propagation of the shock front in the open Atmosphere is spherical in shape. An experiment has been conducted by a horizontal shock tube to generate subsonic wave as well as supersonic shock wave by rupturing different types of diaphragm. The material properties of diaphragm are the major factor to generate shock wave in a shock tube. It is observed that the material properties like toughness and brittleness are the best properties in the present case. In this experiment, transparent Raksin sheet and used X-ray film earlier are used where X-ray film gives the better performance as compare to Raksin for better rupture. For measuring the wave speed in the open Atmosphere, a time measuring technique has been developed to measure the wave travel time in open Atmosphere and it is simple in use and accurate in measurement as compare to other complicated technique like wave visualization by High Speed Video camera. During wave propagation, the wave speed decreases with increasing the travel time in open Atmosphere. So decreasing the speed indicates that the decay of the wave strength is normal phenomena during propagation in the open Atmosphere. In the present study, both experimental and numerical techniques are applied successfully to determine the decay of the wave strength.
Performance analysis of model predictive controller in non linear applications
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Ridwan, Abdur Raquib
This thesis is concerned with depicting the performance of MPC as a controller in nonlinear systems covering various fields of engineering. The Adaptive PID controller is used as the generic means of controller comparison. Main Emphasis and theme of this thesis is to visualize the performance of MPC as a controller for the flight control of the Micro Aviation Vehicle Ornithopter. The velocity, altitude and angular position are taken as the flight outputs to be controlled. A two dimensional model of the ornithopter is selected because it accommodates the inseparability of thrust and lift in its system which is instrumental for describing the flight operation of the ornithopter. Both the acceleration phase and steady state cruise motion control of the ornithopter is considered. The entire nonlinear mathematical model describing the equations of motion of the ornithopter is simulated in the Simulink Environment of the Matlab Software. The performance of the Model Predictive Controller for the flight control of MAV ornithopter is evaluated by comparing the responses with that of an Adaptive PID controller. Although a very improved and significant flight control of the ornithopter is possible with the Model Predictive Controller, a disturbance prone scenario, changing wind direction and speed is needed to completely visualize the impact of Model Predictive Controller in the flight control of an ornithopter. Moreover a better system model comprising a three dimensional mathematical model is required to successfully proceed with further altitude and velocity control. Other directional outputs such as the angular velocity, acceleration can also be investigated. Hence this thesis can be considered as the basis of properly understanding the control mechanism of the Model Predictive Controller and finally would serve as the platform from which the flight control of ornithopter can be investigated with Model Predictive Controller under more realistic and practical instances in which the ornithopter needs to be controlled.
The effects of Social Networking Sites to the lifestyle of Teachers and Students in Higher Education Institutions
(Department of Technical and Vocational Education (TVE), Islamic University of Technology (IUT), 2012-11-15) Asad, Seguya
With the creation of Social Networking Sites, colleges and universities across the globe have been playing catch-up with students. This new technology carries much weight as a new medium for students to build social connections and grow as members of their institutions (Boogart, 2006). With the aim of exploring the effect of social networking sites to the lifestyles of teachers and students in higher educational institutions, this research was conducted at four private universities in Bangladesh and one university in Uganda. A sample of one hundred teachers and one hundred students was respectively selected from four private universities in Bangladesh and one university in Uganda. Both qualitative and quantitative methods were applied during data analysis. The results indicated that social networking sites had several impacts on the lifestyle of students and teachers with the positive effects outweighing the negative effects. The findings confirmed that students can formulate group discussions so as to exchange their ideas, communicate to their teachers and appeal to their friends about assignments. Teachers share course related materials with their students, enables research work, creates student groups to collaborate on projects and communicate with their fellow teachers from other universities. So the principal conclusion was that the use of social networking sites should be incorporated into teachers‟ and students‟ lifestyles as its beneficial nature takes over its adverse side.
Designing Spam Mail Filtering Using Data Mining by Analyzing User and Email Behavior
(Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2012-11-15) Islam, Abdullah Ibn Nurul
Electronic Mail is the “killer network application”. It is ubiquitous and pervasive. In a relatively short timeframe, the Internet has become irrevocably and deeply entrenched in our modern society primarily due to the power of its communication substrate linking people and organizations around the globe. Much work on email technology has focused on making email easy to use, permitting a wide variety of information and information types to be conveniently, reliably, sent throughout the Internet. However, the analysis of the vast storehouse of email content accumulated or produced by individual users has received relatively little attention other than for specific tasks such as spam and virus filtering. Users in the email continuously receive spam and they get into trouble wasting their time and also harmful emails can cause harm to the computers. This thesis presents an implemented framework for data mining behavior models from email data. The EMT is a data mining tool kit designed to analyze email corpora, including the entire set of email sent and received by an individual user, revealing much information about individual users as well as the behavior of groups of users in an organization. A number of machine learning and anomaly detection algorithms are embedded in the system to model the user’s email behavior in order to classify email for a variety of tasks. There are different methods for detection of spam through email. The main goal is to develop a method that outperforms the existing methods in terms of detection of spam, ham and wrongly classified spam, i.e. need is to improve the accuracy of the proposed method compared to the other existing methods. The other goal is to implement the proposed algorithm for reducing the time. So, to recapitulate, this thesis also deals the accuracy and process timing based on prioritization of detecting email messages. The proposed method uses prioritization of process criterion which is unavailable in the earlier existing methods. It also uses the post-filtering concept which contributes for the enhancement of accuracy of the proposed method. Thus the proposed method, which we name as MAN is responsible for spam detection and outperforms Abstract xii the existing methods. This method also provides user convenient spam detection process. So, by using the concepts of post-filtering, process prioritization and different criterion in order to detect spam, the optimum accuracy for detecting spam will be possible.
Design and analysis of a wideband microstrip antenna for high speed WLAN
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Al-Arif, S. M. Masudur Rahman
The wired local area network is becoming wireless day by day. The allocation of frequency spectrum for this Wireless LAN is different in different countries. This presents a myriad of exciting opportunities and challenges for design in the communications arena, including antenna design. The high speed WLAN has many standards and most antennas available do not cover all the standards. The objective in thesis is to design a suitable antenna that can be used for high speed WLAN application. The primary goal is to design antenna with smallest possible size and better polarization that covers all the high speed WLAN standards ranging from from 4.90 GHz to 5.82 GHz. Many works are going on throughout the world in field of antenna designing. Recent works on this field is studied in this thesis to find out most suitable antenna shape for the specific application. Zeland’s IE3D simulation software has been used to design and simulate antenna for WLAN frequency band. A parametric study has been presented in process to the designed antenna for an effective bandwidth of 4.9- 5.825 GHz. Return loss -10 dB or lower is taken as acceptable limit. Finally single E shape microstrip patch antenna having a dimension of 24×19 mm2 is demonstrated with a frequency band of 4.89 GHz to 5.83 GHz. The antenna presents a good current distribution and radiation pattern for all four standards of WLAN that falls inside this frequency range.
A Study to Compare the Results of Technical and Pedagogical Subjects in the Undergraduate Programme in the Department of Technical and Vocational Education at the Islamic University of Technology (IUT), Dhaka.
(Department of Technical and Vocational Education (TVE), Islamic University of Technology (IUT), 2012-11-15) Moluh, Nsangou
The objective of the study was to compare the results of the undergraduate students between the technical and pedagogical subjects. There was a feeling that the evaluation in the TVE department was not in line with other departments. The pedagogical subjects were taught by the teachers of TVE department, whereas technical courses were taught by teachers of engineering department including Mechanical, Electrical and Computer science. From this study, the administration of IUT may get a clear picture of grades, given by its four departments, hence, it will be easier for them to identify, if there is any deviation from the expected standard. It will also help the particular department specially (TVE) in maintaining a standard of grading system, in line with other departments. The study followed descriptive research methodology. The population was the graduated students of the last five years (2006-2010), in the one year B.Sc.T.E. programme. As the population was of small size, all the students over the last five years were considered as sample in this study. The tool of research was documents of the results in term of achievement of the undergraduate students over the last five academic years. The data was collected from the IUT administration by the researcher himself. One sample paired t-test was done to compare the results of technical and pedagogical subjects at 0.05 level of significance. The study was delimited to only the last five years students who completed the Bachelor of Science in Technical education in one year programme. The null hypothesis was tested, meaning that there is no difference of results between pedagogical and technical subjects in the undergraduate program. The data was collected from the IUT Registrar Office by the researcher himself. For analyzing the data, one sample paired t-test was done to compare the results of technical and pedagogical subjects at 0.05 level of significance. The study found that during the academic year 2006 there was no significant difference in results between pedagogical and technical subjects in B.Sc.T.E one year programme. Results for the academic year 2007 to 2010 and combined results from 2006 to 2010, showed significant difference in results between pedagogical and technical subjects in B.Sc.T.E 1
Development of an adaptive neuro-fuzzy controller for speed control of induction motor drives
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Azim, Md. Riyasat
Robust and precise speed control is of critical importance in high performance drive applications. Control schemes for AC machines are complicated as compared to DC machines since they exhibit nonlinear relationships between process variables e.g. speed and manipulated variables e.g. current, torque etc. Direct torque control (DTC) scheme offers faster and simpler control of AC machines with high dynamic performance but without extensively using coordinate transformations and hence with lesser computational burden on the processor. Artificial intelligent controllers (AIC) are capable of overcoming the limitations of the mathematical model dependent conventional fixed gain and existing adaptive controllers. In this thesis, an adaptive neuro-fuzzy inference system (ANFIS) based controller is proposed to improve the dynamic behavior of DTC based induction motor drive which offers the combined advantages of the flexibility of fuzzy logic and adaptability of neural networks. Hence the developed adaptive neuro-fuzzy controller can be utilized to minimize the effects of unavoidable system disturbances such as, system parameter variations, sudden impact of load changes etc. Since the use of intelligent controllers are very limited in the drives industry due to the relevant computational burden on the microprocessor, therefore, in order to reduce the computational burden, linear linguistic variables with an optimum number of membership functions have been selected for the adaptive neuro-fuzzy controller developed in this thesis. The effectiveness of the proposed NFC based DTC scheme of the IM drive is consolidated through the development of a simulation model using MATLAB/Simulink. The results obtained from the simulation of the proposed system are compared with the results from simulation of the same system using conventional proportional integral (PI) controller. The results have shown promising improvements both in the transient and steady state responses of the system. Further, as an integral part of the thesis, a fuzzy logic controlled dynamic voltage restorer (DVR) system has been developed in order to protect the induction motor drive system from the power quality problems (like- voltage sag, swell etc.). The DVR system was developed and simulated in MATLAB/Simulink environment to verify its functionality. The simulation results have confirmed the ability of the DVR system to perform at an expected level when used in conjunction with the induction motor drive. The DVR system kept the induction motor drive performance intact under both symmetrical and asymmetrical system fault conditions.
Improvement of machinability of steel during turning operation by magnetic cutting
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Mahmood, Mohammad Nauzef
This study presents the results of an investigation of machinability improvement in turning of mild steel (0.25% C) by magnetic cutting. Turning experiments were done by using a precision lathe both without application of magnetic field and with the application of magnetic field. The external magnetic field was implemented using two different magnetic setups: one created by permanent magnet another by electro magnet. Poor machinability results in surface defects and dimensional inaccuracy of the work piece as well as reduction of useful tool life and damage to machine. A recent and promising way to improve the machinability of steel is to reduce the vibration effect of machine-tool-work system by applying external damping effect using magnetic cutting technique. In this investigation the effect of external magnetic field (both permanent magnet and electromagnet) was evaluated on machinability parameters such as tool life, surface roughness, cutting temperature and cutting force. The characteristics of chip produced during turning were also observed to further study the magnetic effect on cutting mechanism. Improvements in tool life, reduction in average surface roughness, increase in cutting temperature and decrease in cutting force were observed during the experimentation by magnetic cutting. Certain changes of chip formation characteristics were also observed when the mild steel was being cut. It has been observed that the chip serration nature has been changed significantly in magnetic cutting. Experiments were carried out by changing cutting speed and depth of cut to determine the optimal magnetic field to be applied which would translate into greatest amount of benefits. Due to the limitation of feed variation within the existing facilities of lathe, the feed value was kept constant at 0.095 mm/rev. The observations of effect of magnetic field on machinability parameters, lead to the conclusion that magnetic cutting improves the machinability factors of mild steel significantly and further enhance the possibility of using magnetic effects to reduce machining costs.
A novel approach for diagnosis of breast cancer using ultrasound imaging
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Quader, Niamul
Breast cancer continues to be the most occurring cancer in women around the globe. With advancements in medical procedures, breast cancers are now virtually treatable in their early stage. Early detection, in turn, is vastly improved with community-wide organized sustainable programs. However, such schemes demand resources, and sizable amount of fund is being wasted for biopsy of benign lesions. Ultrasound imaging, because of its low ionization, effectiveness in diagnosing lesions of younger women and lesser resource intensiveness makes it a viable option for mass population programs. It has been shown to be effective in prevention of significant number of unnecessary biopsies. Thus, ultrasound imaging modality remains a prominent tool for diagnosis of breast cancer. Rigorous work is being done to improve overall imaging modality. In light of breast cancer, quantization of acoustic and morphometric features has proven to give good Receiver Operating Characteristic (ROC) area performance. However, ambiguity may still be there at concluding a lesion to be benign. The natural trend for such scenarios is the follow-up checkups recommended by experts. With thousands of follow-up checkups occurring each year, the need for a systematic study is imminent. We develop an effective algorithm that systematically processes over-time-data with the use of thin plate spline smoothing. This is followed by systematic categorization of different sets of physiological changes in light of benign and malignant lesions. Finally a versatile community wide scheme is outlined.
A Novel Approach for Diagnosis of Breast Cancer using Ultrasound Imaging
(Department of Electrical and Electronic Engineering, Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2012-11-15) Quader, Niamul
Breast cancer continues to be the most occurring cancer in women around the globe. With advancements in medical procedures, breast cancers are now virtually treatable in their early stage. Early detection, in turn, is vastly improved with community-wide organized sustainable programs. However, such schemes demand resources, and sizable amount of fund is being wasted for biopsy of benign lesions. Ultrasound imaging, because of its low ionization, effectiveness in diagnosing lesions of younger women and lesser resource intensiveness makes it a viable option for mass population programs. It has been shown to be effective in prevention of significant number of unnecessary biopsies. Thus, ultrasound imaging modality remains a prominent tool for diagnosis of breast cancer. Rigorous work is being done to improve overall imaging modality. In light of breast cancer, quantization of acoustic and morphometric features has proven to give good Receiver Operating Characteristic (ROC) area performance. However, ambiguity may still be there at concluding a lesion to be benign. The natural trend for such scenarios is the follow-up checkups recommended by experts. With thousands of follow-up checkups occurring each year, the need for a systematic study is imminent. We develop an effective algorithm that systematically processes over-time-data with the use of thin plate spline smoothing. This is followed by systematic categorization of different sets of physiological changes in light of benign and malignant lesions. Finally a versatile community wide scheme is outlined
Design and analysis of a model predictive unified power flow controller (MPUPFC) for improving power system stability
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Shahriar, Md. Shoaib
This thesis addresses model predictive controller (MPC) as an effective solution for improving the oscillations in a single machine infinite bus (SMIB) power system connected with a FACTS device named unified power flow controller (UPFC). UPFC is mainly used in the transmission systems which can control the power flow by controlling the voltage magnitude, phase angle and impedance. Linearized model of UPFC connected with SMIB system is modeled by five state equations. System oscillations of the plant are attempted to be controlled with power system stabilizer (PSS) and proportional integral (PI) controller. MPC was introduced to damp out the oscillations to improve the performances those of obtained using PSS and PI controller. As a controller, MPC not only provides the optimal control inputs, but also predicts the system model outputs to reach the desired goal. So, model predictive unified power flow controller (MPUPFC), a combination of UPFC and MPC along with proper system model parameters can provide a satisfactory performance in damping out the system oscillations in order to obtain a stable system. Simulation is done in Matlab simulation software. Responses are shown for four different states controlling four different control signals of UPFC
Development of a smart controller for a switched reluctance motor drives
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Hasan, Md Rezaul
Nowadays, switched reluctance motors (SRMs) attract more and more attention. Switched reluctance machines have emerged as an important technology in industrial automation. They represent a real alternative to conventional variable speed drives in many applications. It not only features a salient pole stator with concentrated coils, which allows earlier winding and shorter end turns than other types of motors, but also features a salient pole rotor, which has no conductors or magnets and is thus the simplest of all electric machine rotors. Simplicity makes the SRM inexpensive and reliable, and together with its high speed capacity and high torque to inertia ratio, makes it a superior choice in different applications. However, the control of the SRM is not an easy task. The motor flux linkage appears to be a nonlinear function of stator currents as well as rotor position, as does the generated electric torque. Apart from the complexity of the model, the SRM should be operated in a continuous phase-to-phase switching mode for proper motor control. This makes the control of SRM a tough challenging. This thesis attempts to first create a MATLAB model of multiphase SRM using the equations governing the dynamic behavior of linear inductance profile SRM. Based on this model, an example case study of single phase SRM operation has been. Small signal analysis of linearized single phase SRM was simulated. Performance analysis of the speed control loop, current control loop and overall SRM drive using PI controller was simulated in MATLAB environment and later all those were simulated again after PI controller was tuned by genetic algorithm. Performance improvement of genetically tuned PI controller is proved in this thesis.
Design and analysis of therapeutic ultrasound systems to develop a diathermy device
(Department of Electrical and Electronic Engineering, Islamic University of Technology, 2012-11-15) Dhali, Maruf Ahmed
Ultrasound continues to be one of the major tissue heating modalities used for the diathermy treatment of a number of medical conditions. Therefore there is innovation that is continually increasing the quality of the ultrasound diathermy systems currently available. However focus is just beginning to shift into the field of low-cost systems to facilitate the mass population. In this work a high power ultrasound system is developed that can deliver therapeutic acoustical energy waves to the ultrasound transducer probe at user defined frequency of 1MHz and 3MHz. Different output power and customized RCA connector facilitate the use of multiapplicator or probe. Output parameter of tone-burst mode is introduced along with continuous and pulsed mode. BJT is used for switching instead of FET, which has added some extra consumption of power but has greatly reduced the price. Still the low cost system is slightly held back with the high cost burden of piezoelectric probe, but enhanced with the new parameter of tone-burst mode. Experimental results are compared with simulation and practical tests with existing products. The total system is implemented using local hardware resources and hence the system is cost effective and beneficial for the third world countries.