Izmir Institute of Technology

Permanent URI for this communityhttp://65.108.157.135:4000/handle/123456789/1

Browse

Now showing 1 - 20 of 117

Accelerometer based handwritten character recognition using dynamic time warping
(Izmir Institute of Technology, 2016-07) Tunçer, Esra; Ünlü, Mehmet Zübeyir
Character and gesture recognition are one of the most studied topics in recent years. Character recognition studies are generally based on image processing. Only a few studies can be found about character recognition as gesture recognition. Gesture recognition is making the computers understand human body movements by using different kind of knowledge of the environment. This knowledge can be obtained by image or sensor-based efforts. Accelerometer is the most used sensor in gesture recognition, so in this study a 3-axis accelerometer is used. In this thesis, English alphabet‟s lowercase characters are used. A ring-like device which contains accelerometer in it is used. After obtaining the acceleration data of each character with 20 repetitions, we apply filtering, segmentation and normalization preprocessing steps for each signal. Since there are different accelerations and decelerations between each repetitions, Dynamic Time Warping (DTW) algorithm has been chosen to determine the similarities between signals. DTW is an algorithm that uses the amplitude values of the signals, so it is weak to amplitudes that shift in time domain. To overcome this shortcoming, the method called Derivative Dynamic Time Warping (DDTW) has been applied to the acceleration signals. DTW and DDTW methods have been compared and we have found that even we remove the normalization step; DDTW gives better results than DTW. By comparison of linear alignment and DTW, the results show that DTW gives better recognition rates for signals with different accelerations and decelerations. DTW also gives better result for the different length signals.
Air pollution effects on the façade of the botter apartment in İstanbul
(Izmir Institute of Technology, 2010) Parlak, Birsen; Böke, Hasan
Air pollution showed a large increase with the rapid development of industry in the middle of the 18th century. Air pollution has been affecting cultural heritage along with human health. Limestone is one of the most affected building materials from the air pollution. Carbon, sulphure and nitrogen gases are the main components which are affective in the deterioration of limestone and their rate increased with air pollution. Sulphure dioxide (SO2) reacts with the calcite crystals (CaCO3) that are the main structure of limestone and leads to the formation of gypsum (CaSO4.2H2O). In this study, the formation of gypsum on limestone was investigated on the façade of the Botter Apartment within the restoration and conservation studies. For this purpose physical, mineralogical and chemical compositions of weathered limestone were determined by XRD, SEM-EDS, FT-IR and TGA analyses. The weathering forms caused by air pollution were documented with drawing (mapping) and photographs of front façade of Botter Apartment.The results of the study indicated that gypsum formation on limestone surfaces is mainly originated from wet and dry deposition process of sulphur dioxide. Condensation may also play an important role in gypsum formation on such sheltered surfaces in addition to dry deposition. On sheltered surfaces of limestone, due to the dry deposition of gypsum formation proceeds as black crust formation. The gypsum formation has not restricted on the surface of limestone. Deeper penetration and absorption of sulphure dioxide are observed in limestone because of their more porous structure. In Istanbul average daily temperature is low and average relative humidity and sulphure dioxide concentrations are high in winter time. The results of this study indicated that İstanbul atmosphere, with its coinciding high relative humidity and high sulphure dioxide concentrations in winter have led to gypsum formation on limestone. This situation has been encountered in all calcareous stones used in the construction of the buildings.
An alternative approach to free space optical communication link
(Izmir Institute of Technology, 2004) Karatay, Okan; Dinleyici, Mehmet Salih
In recent years commercial and military interest in free-space optical (FSO) communication has been growing due to advantages of high bandwidths, portability and high security. Despite of the numerous advantages, atmospheric events such as attenuation and scintillation severely affects the link quality. Novel methods of mitigating atmospheric events have been applied. Usage of large aperture lenses and high transmitted power decreases the degradation effects, but increases the overall system cost. Fresnel lenses, as an alternative to the classical optic system for mitigating scintillation effects would be a low-cost solution. As a result of the technological improvements on output power and divergence angles of VCSEL (Vertical Cavity Surface Emitting Laser) high-speed cost effective FSO communication system design became possible. In this thesis commercially available FSO systems are reviewed and an alternative low-cost system relying on Fresnel lens optical unit is theoretically and experimentally investigated. The system has been tested and compared for two-day periods with various configurations under atmospheric effects. The results indicated that the Fresnel lenses perform comparably well in contrast to bulk optic lenses. Thus, the large aperture Fresnel optic unit can be deployed for better performance at the scintillation dominating situation without a penalty at the cost.
Analysis and implementation of long period fiber grating and fresnel reflection-based sensors for refractive index measurement of liquids
(Izmir Institute of Technology, 2017-06) İde, Cansu; Yüksel Aldoğan, Kıvılcım
Refractive index (RI) is one of the physical properties of material, such as boiling/ melting point, density, and elasticity. As one of the distinctive features of the material, measurement of the RI is nowadays used in many areas including the food, oil and petro-leum industry, biomedical applications, and environmental analysis. In recent years, fiber optic based refractometers become highly popular thanks to the advantages they provide compared to other methods used in the above-mentioned fields. In this context, two intrinsic fiber optic refractive index sensors were investigated in this thesis: long-period fiber grating (LPFG) based- and Fresnel reflection based- re-fractometers. The former examines the changes in the resonant wavelength interrogated by an Optical Spectrum Analyzer (OSA) whereas the latter uses SMF tip as sensing point interrogated by an Optical Time-Domain Reflectometer (OTDR) from a distant location. In the framework of the thesis, we first of all, provided a detailed field-specific literature survey giving an overview of the fiber optic-based refractive index sensors. Then, the operation principles of LPFG were studied including the modelling aspects of this sensing element. For this purpose, transmission spectra of LPFGs to external refrac-tive index changes have been simulated employing the two-layer fiber geometry. Next, the principles of Fresnel reflection-based sensor were investigated. Finally, experimental work was realized on different concentrations of glucose-water, glycerol-water solutions, and various chemicals. Our experimental results show an excellent agreement with the theory which demonstrated the capability of measuring RI of liquids for both methods investigated in the scope of thesis.
Analysis and implementation of optical fiber sensors for process monitoring of composite materials
(Izmir Institute of Technology, 2016-12) Yılmaz, Anıl; Yüksel Aldoğan, Kıvılcım
The use of composite materials in many industrial applications and structures under high stress (airplane wings, unmanned air vehicles, wind turbines, etc.) has been exponentially growing thanks to their lightweight, superior strength, durability, and corrosion resistance. However, the unique mechanical properties of composite materials cannot be fully exploited without monitoring them when material is placed under constraints. It is essential to monitor their behavior not only in the field but also during manufacturing process to ensure the high quality of manufactured materials. For this purpose, the use of optical fiber sensing, particularly the embedding of fiber Bragg grating (FBG) sensors into composite materials has been gaining growing popularity thanks to various advantages of FBGs. In this context, the main purpose of this thesis is to demonstrate the feasibility of using optical fiber sensors for process monitoring (cure and resin flow) of fiber reinforced thermoset composite materials. The first sensor type studied in the thesis was based on Fresnel reflection. The capability of monitoring the curing profile of liquid matrix (mixture of resin and hardener) has been demonstrated by the way of Fresnel reflection sensor interrogated by Optical Time Domain Reflectometry (OTDR). Then, FBG sensors were embedded into composite plates to measure the temperature change during resin injection and cure cycle, as well as determining the residual strain inside the material. In spite of the popularity of using FBG sensors, most recent researches clearly prove that there is still a lack of technical maturity in real life applications in interrogating embedded FBGs for strain measurements. There is still room for new sensing approaches. In the last part of the thesis, we proposed a novel method to interrogate fiber cavity ring-down (CRD) loop by using OTDR which makes this sensing approach more practical and cheaper than the conventional techniques. Thanks to these advantages, our proposed interrogation method can be implemented for strain measurements inside the composite materials as an alternative to the FBG sensors.
Analysis and modeling of an actuation system to be used in light-weight collaborative robots
(Izmir Institute of Technology, 2019-01) Yılmaz, Mert; Dede, Mehmet İsmet Can
Collaborative robots are a relatively new concept in robotics and industrial automation. Traditional industrial robots are relatively expensive, heavy, and dangerous devices for humans since they are enclosed in working cells with fences. Nowadays, in small and medium enterprises (SMEs), instead of traditional robots lightweight and versatile robots are required which can work alongside human co-workers. In such applications, human safety is a critical factor. ISO has produced new standards for these robots to regulate their collaborative work with human co-workers and named these robots as collaborative robots. The subject of the thesis is related to the actuation systems of light-weight collaborative robots. For these robots, a variety of actuation systems are designed in recent years which are usually modular and compact actuation systems. The main aim of the thesis is to analyze and model of a commonly used actuation system in collaborative robots and to verify its model.
Analysis of observed chaotic data
(Izmir Institute of Technology, 2004) Çek, Mehmet Emre; Savaci, Ferit Acar
In this thesis, analysis of observed chaotic data has been investigated. The purpose of analyzing time series is to make a classification between the signals observed from dynamical systems. The classifiers are the invariants related to the dynamics. The correlation dimension has been used as classifier which has been obtained after phase space reconstruction. Therefore, necessary methods to find the phase space parameters which are time delay and the embedding dimension have been offered. Since observed time series practically are contaminated by noise, the invariants of dynamical system can not be reached without noise reduction. The noise reduction has been performed by the new proposed singular value decomposition based rank estimation method.Another classification has been realized by analyzing time-frequency characteristics of the signals. The time-frequency distribution has been investigated by wavelet transform since it supplies flexible time-frequency window. Classification in wavelet domain has been performed by wavelet entropy which is expressed by the sum of relative wavelet energies specified in certain frequency bands. Another wavelet based classification has been done by using the wavelet ridges where the energy is relatively maximum in time-frequency domain. These new proposed analysis methods have been applied to electrical signals taken from healthy human brains and the results have been compared with other studies.
Analysis of olfactory evoked potentials
(Izmir Institute of Technology, 2014) Olcay, Bilal Orkan; Savaci, Ferit Acar
With the growing opportunities of laboratories and measurement techniques, cognitive science attracts many researchers interest from other branches of science. In the literature, lack of studies related to the brain's responsiveness against the olfactory stimuli has been the main source of motivation for our work on this issue. In this thesis, it is examined by means of time-dependent wavelet entropy of Electroencephalographic (EEG) signals which is collected from individuals that how olfactory and trigeminal effective odor stimuli affects responsiveness of the brain. Significance and meaningfulness of the results are shown with statistical tests of average entropy in the discrete time windows. Due to its nature of small amplitude in comparison with ongoing EEG activity, it’s hard to observe the components of olfactory evoked potentials and trigeminal evoked potentials. In order to separate these components from ongoing EEG, different signal processing techniques have been employed in this thesis. And, findings from these techniques have been conveyed to statistical tests to determine the most suitable technique for that purpose. Additionally, a novel smell performance identification metric have been offered for clinical studies that is not affected by basal activity of brain and subjective review, for objective assessment of smell performance. Statistical test result have shown that, results of this technique which is performed on 19 participants, and their TDI scores obtained from Sniffin’ Stick test battery, are in a strong correlation.
Analysis of the electromagnetic scattering from flat plates by using different sinc-type basis fuctions in method of moments
(Izmir Institute of Technology, 2012) Özbakış, Başak; Kuştepeli, Alp
Electromagnetic scattering from three dimensional arbitrary flat surfaces illuminated by a plane wave is investigated by using sinc-type basis functions in the method of moments (MoM) in this study. Sinc-sinc based Galerkin MoM formulation is obtained and its fortran program is developed firstly. The number of integrals to be computed is decreased by using analytical properties of the sinc function in the formulation. Therefore, the CPU time for obtaining the coefficient matrix is decreased appreciably. The relative error during the generation of the main matrix elements is obtained between 0.058% and 0.095% for considerably large matrices. Rooftop based MoM formulation is developed and it is coded with a similar approach used in sinc based program. The accuracy and CPU time of the sinc based bistatic scattering results are compared with those of rooftop. The MoM formulation of the sinc-pulse (sinctop) basis functions with non-Galerkin case is developed to reduce the overflow problem at the edges. Induced surface currents and far-field results of the sinc-sinc and sinc-pulse based formulations are compared with those of the rooftop basis functions and with the ones obtained from SuperNEC. Both sinc-sinc and sinc-pulse based simulation results are in good agreement with the results of rooftop based and SuperNEC simulation results. The main purpose in this study is to obtain accurate bistatic scattering results by using different sinc-type basis functions in MoM procedure while filling the main matrix in less CPU time when compared with the well-known methods.
An approach to summarize video data in compressed domain
(Izmir Institute of Technology, 2007) Şimşek, Gökhan; Gümüştekin, Şevket
The requirements to represent digital video and images efficiently and feasibly have collected great efforts on research, development and standardization over past 20 years. These efforts targeted a vast area of applications such as video on demand, digital TV/HDTV broadcasting, multimedia video databases, surveillance applications etc. Moreover, the applications demand more efficient collections of algorithms to enable lower bit rate levels, with acceptable quality depending on application requirements. In our time, most of the video content either stored, transmitted is in compressed form. The increase in the amount of video data that is being shared attracted interest of researchers on the interrelated problems of video summarization, indexing and abstraction. In this study, the scene cut detection in emerging ISO/ITU H264/AVC coded bit stream is realized by extracting spatio-temporal prediction information directly in the compressed domain. The syntax and semantics, parsing and decoding processes of ISO/ITU H264/AVC bit-stream is analyzed to detect scene information. Various video test data is constructed using Joint Video Team.s test model JM encoder, and implementations are made on JM decoder. The output of the study is the scene information to address video summarization, skimming, indexing applications that use the new generation ISO/ITU H264/AVC video.
Automatic identification of abnormal regiones in digitized histology cross-sections of colonic tissues and adenocarcinomas using quasi-supervised learning
(Izmir Institute of Technology, 2012) Önder, Devrim; Karaçalı, Bilge
In this thesis, a framework for quasi-supervised histopathology image texture identi- cation is presented. The process begins with extraction of texture features followed by a quasi-supervised analysis. Throughout this study, light microscopic images of the hematoxylin and eosin stained colorectal histopathology sections containing adenocarcinoma were quantitatively analysed. The quasi-supervised learning algorithm operates on two datasets, one containing samples of normal tissues labelled only indirectly and in bulk, and the other containing an unlabelled collection of samples of both normal and cancer tissues. As such, the algorithm eliminates the need for manually labelled samples of normal and cancer tissues commonly used for conventional supervised learning and signicantly reduces the expert intervention. Several texture feature vector datasets corresponding to various feature calculation parameters were tested within the proposed framework. The resulting labelling and recognition performances were compared to that of a conventional powerful supervised classier using manually labelled ground-truth data that was withheld from the quasi-supervised learning algorithm. That supervised classier represented an idealized but undesired method due to extensive expert labelling. Several vector dimensionality reduction techniques were evaluated an improvement in the performance. Among the alternatives, the Independent Component Analysis procedure increased the performance of the proposed framework. Experimental results on colorectal histopathology slides showed that the regions containing cancer tissue can be identied accurately without using manually labelled ground-truth datasets in a quasi-supervised strategy.
Automatic transcription of traditional Turkish art music recordings: A computational ethnomusicology appraoach
(Izmir Institute of Technology, 2012) Gedik, Ali Cenk; Savacı, Ferit Acar
Music Information Retrieval (MIR) is a recent research field, as an outcome of the revolutionary change in the distribution of, and access to the music recordings. Although MIR research already covers a wide range of applications, MIR methods are primarily developed for western music. Since the most important dimensions of music are fundamentally different in western and non-western musics, developing MIR methods for non-western musics is a challenging task. On the other hand, the discipline of ethnomusicology supplies some useful insights for the computational studies on nonwestern musics. Therefore, this thesis overcomes this challenging task within the framework of computational ethnomusicology, a new emerging interdisciplinary research domain. As a result, the main contribution of this study is the development of an automatic transcription system for traditional Turkish art music (Turkish music) for the first time in the literature. In order to develop such system for Turkish music, several subjects are also studied for the first time in the literature which constitute other contributions of the thesis: Automatic music transcription problem is considered from the perspective of ethnomusicology, an automatic makam recognition system is developed and the scale theory of Turkish music is evaluated computationally for nine makamlar in order to understand whether it can be used for makam detection. Furthermore, there is a wide geographical region such as Middle-East, North Africa and Asia sharing similarities with Turkish music. Therefore our study would also provide more relevant techniques and methods than the MIR literature for the study of these non-western musics.
Beam selection techniques in millimeter wave communications
(Izmir Institute of Technology, 2019-12) Cumalı, İrem; Özbek, Berna
Millimeter wave (mmWave) communication is an advantageous technology which is capable of meeting the needs of future mobile networks. On the other hand, the propagation characteristics and system requirements are the restrictive factors for utilization of mmWave communication. Hybrid and digital beamforming architectures can be evaluated as worthy candidates to utilize mmWave communication. In the hybrid architecture, selection of a few number of beams by exploiting the sparse structure of the beamspace channel provides high spectral efficiency with low complexity. In this thesis, the multi-user mmWave communication in sparse and dense environments are investigated. Beam selection algorithms presented in the literature are performed for the sparse environment. While the number of users is equal to the number of radio frequency (RF) chains in a sparse environment, the number of RF chains is less than the number of users in a dense environment. Therefore, an algorithm which performs beam and user selection for the dense environment is proposed. The user selection in the proposed beam and user selection algorithm is performed based on the correlation among users’ channels. Since the users’ channels are highly correlated in mmWave communication, the proposed beam and user selection algorithm improves the spectral efficiency considerably. Furthermore, a non-uniform rectangular array (NURA) antenna configuration for mmWave communication is investigated when the digital beamforming architecture is employed. Then, a user selection algorithm is proposed under the case of lower number of antennas. The simulation results demonstrate the improvement in sum data rate through the proposed user selection algorithm in mmWave communication with NURA configuration.
Brushless DC motor control under varying load conditions
(Izmir Institute of Technology, 2017-12) Kara, Salih Murat; Özdemirel, Barbaros
Brushless direct current motors (BLDCMs) are popular in various applications due to their high-power density, wide speed range and control flexibility. They work without brushes and require electronic commutation, where stator's currunt state is synchronized with rotor position. Various methods are used to determine the rotor position. Sensorless method of position detection is preferable instead of external sensors, because external sensors are not cost effective and they create wiring complexity. This thesis presents control methods and driver algorithms for BLDCM under variable load conditions from start-up to continuous run with speed regulation. The inductive sense method, which is one of the sensorless position measurement methods, is used when the motor is stationary and at low speed. The rotor position was accurately calculated by using this method based on inductance differences. Afterwards, back electromotive force (BEMF) sense method was used at high speed.The drive parameters, mainly the step time and excitation step, are adjusted according to the rotor position to drive the motor efficiently. The dynamic control of the drive parameters is done by estimating the load with the rotor position.
A case study on logging visual activities: Chess game
(Izmir Institute of Technology, 2005) Ozan, Şükrü; Gümüştekin, Şevket
Automatically recognizing and analyzing visual activities in complex nenvironments is a challenging and open-ended problem. In this thesis this problem domain is visited in a chess game scenario where the rules, actions and the environment are well defined. The purpose here is to detect and observe a FIDE (Federation International des Echecs) compatible chess board and to generate a log file of the moves made by human players. A series of basic image processing operations have been applied to perform the desired task. The first step of automatically detecting a chess board is followed by locating the positions of the pieces. After the initial setup is established every move made by a player is automatically detected and verified. A PC-CAM connected to a PC is used to acquire images and implement the corresponding software. For convenience, .Intel R Open Source Computer Vision Library (OpenCV). is used in the current software implementation.
Cell selection algorithms for terrestrial trunked narrow band radio systems
(Izmir Institute of Technology, 2017-07) Karataş, Azad; Özbek, Berna
Since the interest in mobile communication sector is increasing day by day, it makes traffic volume problem more important. There are different works focused on developing more secure and qualified service for professional users and companies. The Professional Mobile Radio (PMR) system, specially developed for professional users in the communication sector, can be offered to the service of professional users and companies with the desired specifications. With the Tetra system, which is one of the PMR systems, users can get more advanced technological services than the conventional PMR systems. Cell selection algorithms have a great importance for these systems which are needed for more reliable, private and seamless communication. In this thesis, we present two novel cell selection algorithms that can be applied to the Tetra based PMR systems. In these algorithms, both the received power of users and the fair distribution of the overall system are considered. Performance evaluation of algorithms with different traffic characteristics is considered in different environments.
Cell selection and interference coordination techniques for heterogeneous wireless networks
(Izmir Institute of Technology, 2015-07) Mahmuda, Subaha; Özbek, Berna
The rapid growth of traffic demands during past years, has led to the immense deployment of heterogeneous wireless networks consisting large-scale macro cells overlaid with multiple tiers of small cells. This is conceived as the major capacity and performance enhancement coordinator by means of increasing the spectral efficiency per unit area. However, heterogeneous networks implementation comprises new technical challenges related to interference issues and throughput deterioration. Advanced interference coordination techniques are introduced to handle these challenges. The usage of range expansion allows captivating more users and hence attaining performance improvement, however causes extra downlink interference. This becomes exquisite for higher bias values; hence the benefits convert into significant deterioration. To overcome these issues, range expansion should be jointly designed with inter-cell interference coordination. The main objective of this thesis is to analyze the concept of heterogeneous network, the cell selection strategies including range expansion, interference coordination schemes and energy efficiency. The performance evaluations are obtained to different macro-pico base stations deployment scenarios for heterogeneous network by using various cell selection algorithms with and without interference coordination depending on frequency allocation schemes to figure out their impact on the system performance for different contours.
Characterization of selected Beylikler Period mortars and processing of new intervention mortars for restoration purposes
(Izmir Institute of Technology, 2005) Budak, Meral; Akkurt, Sedat
In this study, Saruhanlı Emirates period mortars collected from Manisa Çukur Hamam (Bath) and Manisa Hacet Mescidi (Mosque) in Western Turkey were characterized, and the results of characterization tests were used for production of new compatible mortars, which would then be used for conservation of the buildings. For this purpose, physical, mineralogical and chemical compositions of the mortars were characterized by XRD, SEM-EDS, TGA and chemical analysis. The results indicated that the mortars were hydraulic and were produced by the use of pozzolanic aggregates. The use of hydraulic lime mortars can be related to the structural requirements of these buildings. The binder aggregate ratio of mortar samples were in the range of 1:2-1:4. Aggregate particle sizes that are grater than the 1180 .m composed the largest sizes of the total aggregates and aggregates were composed of quartz, albite and anorthite. Lime aggregate ratios (L/Ag) of the produced intervention mortars were in the range of 1:6- 1:1. Standard sand was used as aggregate. The pozzolanic additive was commercial clay, which was heat treated between 25 and 1000oC. It was found that the addition of K-244 clay which was heat-treated above 600oC improved compressive strengths of the mortars.
Colloidal plexcitonic nanocrystals
(Izmir Institute of Technology, 2022-06) Sarısözen, Sema; Yalçın, Şerife Hanım
Noble metal nanocrystals, especially gold and silver, which have attracted a great deal of attention due to the supporting of surface plasmon polaritons (SPPs), have been extensively investigated and studied. With recent developments in colloid chemistry, synthesis of noble metal nanocrystals with tunable optical properties in the visible region of the electromagnetic spectrum has become easier. Until now, noble metal nanocrystals (NPs) synthesized by using various synthetic methods, have a variety of shapes, such as bipyramid, rod, disk, prism, and ring, etc. In the strong coupling regime, SPPs supported by metal nanocrystals interact strongly with excitons of organic dyes, semiconducting quantum dots (carbon or perovskite quantum dots) to generate a new hybrid optical mode called plexciton (plasmon-exciton). Plexcitonic nanocrystals have received interest owing to their ease of synthesis, scalability, and ability to provide sub-wavelength confinement of incident light and offer promising applications. Plasmon–exciton interaction at nanoscale dimension can be improved by generating new plexcitonic nanoparticles with tunable optical properties, which may be utilized in critical applications such as nanolasers, sensors, nano-optics, solar cells, and light emitting diodes. Therefore, there has been a tremendous amount of interest in the synthesis of new plexcitonic nanocrystals having excellent optical and chemical properties. The main goal of this thesis is to synthesize new plexcitonic nanoparticles with tunable optical properties in the visible spectrum: (i) synthesis of different shaped colloidal monometallic and bimetallic nanocrystals, (ii) synthesis of new colloidal plexcitonic nanocrystals, (iii) synthesis of carbon quantum dots (CDs), (iv) coupling of excitons of CDs and SPPs on the silver thin film.
A comparative analysis of coherence measures for electroencephalography
(Izmir Institute of Technology, 2018-12) Çağdaş, Serhat; Karaçalı, Bilge
Functional connectivity is often used in brain-computer interface studies as well as other neuroscience fields as a feature extraction method. In the functional connectivity using electroencephalography (EEG), connectivity patterns are extracted by a dependency matrix showing the coherence between electrode pairs. A variety of dependence measures can be used to calculate this matrix. In this study, a total of 15 coherence measures were analyzed comparatively in terms of computation time, accuracy and statistical significance in discriminating motor/motor imagery activities. As dependence measures, in addition to methods used in the literature for brain connectivity, five other methods used as contrast function in independent component analysis and two novel mutual information calculators proposed in this study were evaluated. Furthermore, a novel hierarchical clustering based statistical test procedure was also proposed for motor/motor imagery activity comparison, along with a similar statistical significance test applied on data from 103 subjects on four different activity types. In experiments on real data set, significance results of dependence measures differed according to the type of activity and time window duration of activity signals. Considering both computation time and accuracy performances on synthetic data, a number of methods with high statistical significance and different dependence characteristics were identified as feasible for a connectivity based brain-computer interface.

Browse

Browsing Izmir Institute of Technology by Department "Electrical and Electronics Engineering"