This thesis focuses on implementation of an algorithm for retinal vessel segmentation in high resolution retinal images.A neural network with two hidden layers was used as the method. A total of 7 features were obtained from matched filtering based on vessel thickness, texture analysis and individual pixels brightness. Within the thesis, the whole database was manually annotated for the implementation of the algorithm and the results. The achieved mean sensitivity reached 80%, specificity 70% and Dice coefficient is 59%.

This masters thesis is focused on filtering the ECG signal for suppression of spurious frequency components of the network. The theoretical part is talking about electrocardiography, ECG signal interference and about principle different types of filtration. In practical part of this thesis are described linear filtering methods and wavelet transform methods with discrete time. The main topic of this work is recommended the best type of filtration.

Goal of this master thesis is to introduce and bring up basics and principles of NetCOPE framework in many details using "first approach" method for exploration of its internal structures - mainly focusing on application core using VHDL for focus itself. Furthermore, this knowledge is used for design and implementation of filtration system for network traffic with more details within phase of design in VHDL language.

Passive radar detects and ranges targets by receiving signals which are reflected off targets. Communication transmissions are generally used, however, theoretically any signal with a suitable ambiguity function may be used. The exploitation of an existing transmitter and the removal of emissions allow passive radars to act as a complementary sensor which is useful in environments where conventional active radar is not well suited. Such environments are in covert operations and in situations where a low cost or spectrally efficient solution is required. Most developed passive radars employ intensive signal processing and use application specific equipment to achieve detection. The high-end processors and receiver equipment, however, detract from some of the inherent advantages in the passive radar architecture. These include the lower cost and power requirements achieved by removing transmitter hardware. This study investigates the challenges faced when removing application-specific and high end components from the system and replacing them with low-cost alternatives. Solutions to these challenges are presented and validated by designing and evaluating a radar using these principles. It was found that the major limitation in passive radar is the dynamic range of the receiver. While processing the signals was, and is, a significant challenge, be implemented on a low-cost, low-power embedded processor. This was achieved by asserting a few limitations to the configuration, exploiting the subsequently generated redundancy, and taking advantage of the parallelism by using general purpose graphics processing.. Even on this processor, the system was able to run in real time and able to detect targets up to 91 km (bistatic range of 195 km) from the radar. / Dissertation (MEng)--University of Pretoria, 2016. / Electrical, Electronic and Computer Engineering / MEng / Unrestricted

The first experimental demonstration of feedback suppression of rotating external kink modes near the ideal wall limit in a tokamak using Kalman filtering to discriminate the n = 1 kink mode from background noise is reported. In order to achieve the highest plasma pressure limits in tokamak fusion experiments, feedback stabilization of long-wavelength, external instabilities will be required, and feedback algorithms will need to distinguish the unstable mode from noise due to other magnetohydrodynamic activity. When noise is present in measurements of a system, a Kalman filter can be used to compare the measurements with an internal model, producing a realtime, optimal estimate for the system's state. For the work described here, the Kalman filter contains an internal model that captures the dynamics of a rotating, growing instability and produces an estimate for the instability's amplitude and spatial phase. On the High Beta Tokamak-Extended Pulse (HBT-EP) experiment, the Kalman filter algorithm is implemented using a set of digital, field-programmable gate array controllers with 10 microsecond latencies. The feedback system with the Kalman filter is able to suppress the external kink mode over a broad range of spatial phase angles between the sensed mode and applied control field, and performance is robust at noise levels that render feedback with a classical, proportional gain algorithm ineffective. Scans of filter parameters show good agreement between simulation and experiment, and feedback suppression and excitation of the kink mode are enhanced in experiments when a filter made using optimal parameters from the experimental scans is used.

More often than not, we cannot directly measure many phenomena that are crucial to us. However, we usually have access to certain partial observations on the phenomena of interest as well as a mathematical model of them. The filtering problem seeks estimation of the phenomena given all the accumulated partial information. In this thesis, we study several topics concerning the numerical approximation of the filtering problem. First, we study the continuous-time filtering problem. Given high-frequency ob- servations in discrete-time, we perform double discretization of the non-linear filter to allow for filter estimation with particle filter. By using the multilevel strategy, given any ε > 0, our algorithm achieve an MSE level of O(ε2) with a cost of O(ε−3), while the particle filter requires a cost of O(ε−4). Second, we propose a de-bias scheme for the particle filter under the partially observed diffusion model. The novel scheme is free of innate particle filter bias and discretization bias, through a double randomization method of [14]. Our estimator is perfectly parallel and achieves a similar cost reduction to the multilevel particle filter. Third, we look at a high-dimensional linear Gaussian state-space model in con- tinuous time. We propose a novel multilevel estimator which requires a cost of O(ε−2 log(ε)2) compared to ensemble Kalman-Bucy filters (EnKBFs) which requiresO(ε−3) for an MSE target of O(ε2). Simulation results verify our theory for models of di- mension ∼ 106. Lastly, we consider the model estimation through learning an unknown parameter that characterizes the partially observed diffusions. We propose algorithms to provide unbiased estimates of the Hessian and the inverse Hessian, which allows second-order optimization parameter learning for the model.

The emergence of micropower-type applications such as self-powered sensors and miniaturized electronic systems has increased interest in on-chip electrochemical energy storage such as microsupercapacitors. Microsupercapacitors (MSCs) are high rate and high power yet miniaturized versions of macroscopic supercapacitors. MSCs with planar configuration have higher power density at potentially comparable energy density to thin-film batteries, while possessing essentially infinite cycle life. They could also offer compatible integration with smart electronic devices on an integrated chip (IC). In this dissertation, state-of-the-art microsupercapacitors based on Ti3C2Tx MXene and other pseudocapacitive electrode materials are proposed. The proposed strategies involve engineering both intrinsic properties of materials, fabrication methods and device architecture.

Full Waveform Inversion (FWI) is a non-linear optimization problem aimed to estimating subsurface parameters by minimizing the misfit between modeled and recorded seismic data using gradient descent methods, which are the only practical choice because of the size of the problem. Due to the high non-linearity of the problem, gradient methods will converge to a local minimum if the starting model is not close to the true one. The accuracy of the long-wavelength components of the initial model controls the level of non-linearity of the inversion. In order for FWI to converge to the global minimum, we have to obtain the long wavelength components of the model before inverting for the short wavelengths. Ultra-low temporal frequencies are sensitive to the smooth (long wavelength) part of the model, and can be utilized by waveform inversion to resolve that part. Unfortunately, frequencies in this range are normally missing in field data due to data acquisition limitations. The lack of low frequencies can be compensated for by utilizing wide-aperture data, as they include arrivals that are especially sensitive to the long wavelength components of the model. The higher the scattering angle of a 5 recorded event, the higher the model wavelength it can resolve. Based on this property, a scattering-angle filtering algorithm is proposed to start the inversion process with events corresponding to the highest scattering angle available in the data, and then include lower scattering angles progressively. The large scattering angles will resolve the smooth part of the model and reduce the non-linearity of the problem, then the lower ones will enhance the resolution of the model. Recorded data is first migrated using Pre-stack Exploding Reflector Migration (PERM), then the resulting pre-stack image is transformed into angle gathers to which an angle filtering process is applied to remove events below a certain cut-off angle. The filtered pre-stack image cube is then demigrated (forward modeled) to produce filtered surface data that can be used in waveform inversion. Numerical tests confirm the feasibility of the proposed filtering algorithm. However, the accuracy of the filtered section is limited by PERM’s singularity for horizontally-traveling waves, which in turn is dependent on the velocity model used for migration and demigration

A long-standing controversy concerning the pros and cons of Vll selective amplification for the sensorineural hypacusic has been and is now being waged. There exists clinical evidence to the effect that some cases with high -frequency sensorineural hearing loss can receive benefit through selective amplification. The purpose of this study was to examine several aspects of the speech signal that could be affecting intelligibility when speech is presented at high -intensity levels.

The use of small satellites to measure the native electric field of the earth has historically presented many problems as a result of the generally modest pointing capabilities of small satellites. In spite of this, the cost of small satellites makes them ideal for just such scientic missions. This thesis details many of the constraints of electric field measuring missions as well as the requirements on any spacecraft designed to accomplish such. The data from a small sounding rocket mission is then analyzed and its usefulness discussed. Possible other methods for use are also discussed.