In this thesis, the effect of registration process is evaluated as well as several methods proposed for 3D face recognition. Input faces are in point cloud form and have noises due to the nature of scanner technologies. These inputs are noise filtered and smoothed before registration step. In order to register the faces an average face model is obtained from all the images in the database. All the faces are registered to the average model and stored to the database. Registration is performed by using a rigid registration technique called ICP (Iterative Closest Point), probably the most popular technique for registering two 3D shapes. Furthermore some variants of ICP are implemented and they are evaluated in terms of accuracy, time and number of iterations needed for convergence. At the recognition step, several recognition methods, namely Eigenface, Fisherface, NMF (Nonnegative Matrix Factorization) and ICA (Independent Component Analysis) are tested on registered and non-registered faces and the performances are evaluated.

This thesis addresses the improvement of a physically based modeling simulator Naval Surface Tactical Maneuvering Simulation System (NSTMSS), that combines different simulators in a distributed environment by the help of High Level Architecture (HLA), to be used in naval tactical training systems. The objective is to upgrade a computer simulation program in which physical models are improved in order to achieve a more realistic movement of a ship in a virtual environment. The simulator will also be able to model the ocean waves and ship wakes for a more realistic view. The new naval model includes a 4 degrees of freedom (DOF) maneuvering model, and a wave model. The numerical results from real life are used for modeling purposes to increase the realism level of the simulator. Since the product at the end of the thesis work is needed to be a running computer code that can be integrated into the NSTMSS system, the code implementation and algorithm details are also covered. The comparisons between the wave models and physical models are evaluated for a better real time performance. The result of this thesis shows that the integration of a 4-DOF realistic ship model to the system improved the capability of NSTMSS to give more data to the student officers while making maneuvers. The result also indicates that the use of waves and ship wakes had taken the simulator to a next level in the environment perception.

Ray tracing is a computationally complex global illumination algorithm that is used for producing realistic images. In addition to parallel implementations on commodity PC clusters, recently, Graphics Processing Units (GPU) have also been used to accelerate ray tracing. In this thesis, ray tracing is accelerated on a GPU cluster where the viewing plane is divided into unit tiles. Slave processes work on these tiles in a task parallel manner which are dynamically assigned to them. To decrease the number of ray-triangle intersection tests, Bounding Volume Hierarchies (BVH) are used. It is shown that almost linear speedup can be achieved. On the other hand, it is observed that API and network overheads are obstacles for scalability.

Fusion of multiple stereo algorithms is performed in order to obtain accurate disparity segmentation. Reliable disparity map of real-time stereo images is estimated and disparity segmentation is performed for object detection purpose. First, stereo algorithms which have high performance in real-time applications are chosen among the algorithms in the literature and three of them are implemented. Then, the results of these algorithms are fused to gain better performance in disparity estimation. In fusion process, if a pixel has the same disparity value in all algorithms, that disparity value is assigned to the pixel. Other pixels are labelled as unknown disparity. Then, unknown disparity values are estimated by a refinement procedure where neighbourhood disparity information is used. Finally, the resultant disparity map is segmented by using mean shift segmentation. The proposed method is tested in three different stereo data sets and several real stereo pairs. The experimental results indicate an improvement for the stereo analysis performance by the usage of fusion process and refinement procedure. Furthermore, disparity segmentation is realized successfully by using mean shift segmentation for detecting objects at different depth levels.

Rendering large terrains on large screens at interactive frame rates is a challenging area of computer graphics. In the last decade, real-time terrain rendering on large screens played a significant role in various simulations and virtual reality systems. To fulfill the demand of these systems, two software tools are developed. The first tool is a Terrain Editor that creates and manipulates large terrains. The second is a Multi-Display Viewer that displays the created terrains on multiple screens. Since the typical large terrains consist of many polygons, graphics boards might have difficulties in rendering the terrain at interactive frame rates. The common solution to this problem is to use terrain simplification without losing image quality. To this purpose, in this study, a paged level of detail mechanism that works with multiple threads is developed and integrated on multiple screen display systems to increase the performance of the high resolution systems.

Software development is highly dependent on the use of tools. These tools support and automate activities performed in different sub-domains of software development. However, they don&lsquo / t adequately provide or support integration facilities, and act as &amp / #8213 / islands of automation&amp / #8214 / . This restricts their benefits to only specific parts of the process. To reap the benefits of integration, this thesis provides a process modeling based method named PLETIN to identify and implement software development tool integration-tuples. The method aims to present solutions for issues observed in tool integration for software development organizations by delivering an integrated tool set. The proposed solution approach is based on the idea that if there were no integrations between tools at all, users would perform the necessary actions to cooperate different tools. PLETIN is a method for the identification of the candidate integration situations (integration-tuples) from the interactions of users with the tools. These tuples constitute the requirements used to develop integration facilities. The software development process definitions are used as inputs to create process models and provide actual implementations. The research is supported with case-study work to identify the significance of the problems and the applicability of the method as a solution to issues in tool integration.

Electrical conductivity of biological tissues is a distinctive property which differs among tissues. It also varies according to the physiological and pathological state of tissues. Furthermore, in order to solve the bioelectric field problems accurately, electrical conductivity information is essential. Magnetic Resonance Electrical Impedance Tomography (MREIT) technique is proposed to image this information with high spatial resolution. However, almost all MREIT algorithms proposed to date assumes isotropic conductivity in order to simplify the underlying mathematics. But it is known that most of the tissues in human body have anisotropic conductivity values. The aim of this study is to reconstruct anisotropic conductivity images with MREIT. In the study, five novel anisotropic conductivity reconstruction algorithms are developed and implemented. Proposed algorithms are grouped into two: current density based reconstruction algorithms (Type-I) and magnetic flux density based algorithms (Type-II). Performances of the algorithms are evaluated in several aspects and compared with each other. The technique is experimentally realized using 0.15T METU &ndash / EE MRI System and anisotropic conductivity images of test phantoms are reconstructed using all proposed algorithms.

Electrochemical biosensors are cost effective, fast and portable devices, which can determine the existence and amounts of chemicals in a specific medium. These devices have many potential applications in many fields such as determination of diseases, process and product control, environmental monitoring, and drug research. To realize these potentials of the devices, many studies are being carried out to increase their sensitivity, selectivity and long term stabilities. Surface modification studies with various types of particles (metal nano particles, carbon nano tubes etc.) can be count among these studies. Although zeolites and zeo-type materials are investigated for many years, they still hold interest on them due to their capabilities. By means of their chemical resistances, large surface areas, tailorable surface properties, and porous structures they can be applied in many applicational fields. In some recent studies, these properties are intended to be used in the field of biosensors. The purpose of the current study was to investigate the effect of zeolite nanoparticles on electrochemical biosensor performances. Firstly, several different procedures were investigated in order to find the best and optimum methodology to attach previously synthesized zeolites on Si wafer substrates for the first time. For this purpose, the ultrasonication, spin coating and direct attachment methods were used and their efficiencies were compared. Perfectly oriented, fully covering, zeolite monolayers are produced by direct attachment method. Successively produced zeolite thin films were then patterned with the help of Electron Beam Lithography technique to show the compatibility of coating methods to the CMOS technology. Combination of Direct Attachment and EBL techniques resulted well controlled zeolite monolayer patterns. Then zeolite modified electrochemical biosensors were tested for their performances. With these experiments it was intended to improve the selectivity, sensitivity and storage stabilities of standard electrochemical biosensors. Experiments, conducted with different types of zeolites, showed that zeolites have various effects on the performances of electrochemical biosensors. Amperometric biosensor response magnitudes have been doubled with the addition of Silicalites. Faster conductometric electrode responses were achieved with enzyme immobilization on zeolite film technique. Also it is seen that Beta type zeolites modified through different ion exchange procedures, resulted different responses in IS-FET measurements.

This thesis investigates the effect of emotional facial expressions of a virtual character on people&rsquo / s performance for interactive digital tasks. The basic and universal emotions are used in the study. Facial expressions of these emotions are created according to the Facial Action Coding System (FACS), which is a system that describes facial movements in the face. The patterns of cooccurences of Action Units (descriptions of facial movements defined in FACS) for basic emotions are also implemented into emotional facial expressions with regard to findings of the studies in the literature. A study was conducted to validate the recognition of emotion specific facial expressions that are built by Poser software. To investigate the effect of emotional facial expressions on people&rsquo / s performance for digital interactive tasks in a virtual environment, a digital interactive application created by Unity software was used in the final study of the thesis.

The aim of this thesis is to explore design directions to minimize a food wastage problem in western domestic kitchens. Central to the thesis is an understanding of people&rsquo / s behavior towards the food waste phenomenon. Three interconnected studies and one design project are included. In Study I, 18 participants were interviewed to explore their perceptions and attitudes towards food waste, revealing the most wasted food types and reasons for food wastage. The findings of Study I are clustered under four phases of food handling: acquisition, preparation, consumption and storage. Study II comprised a generative session with three users and two designers, devised to explore latent and tacit knowledge regarding food wastage. Study II resulted in user-generated ideas for minimizing food waste, which were analyzed so as to reveal possible design directions. From these results, a set of criteria for a &lsquo / perfect&rsquo / kitchen appliance, which could minimize food waste, was drawn-up. The design project took the research findings of Study I and II and devised a collection of design concepts as possible ways to help reduce domestic food waste. Two concepts &ndash / Philips Dispense and Canvas - are taken further because they relate to the most wasted food types: &lsquo / bread&rsquo / and &lsquo / vegetables and fruits&rsquo / . In Study III, Philips Dispense and Canvas were evaluated with a questionnaire. According to the results, in households containing busy couples without children, Philips Dispense is valued highest (it takes the food waste responsibility away from users) whereas Philips Canvas was valued lower (it gives feedback on current stocks and persuades homeowners not to waste food).