A new approach is proposed which uses a combination of a Backprop paradigm neural network along with some perceptron processing elements performing logic operations to construct a numeric-to-symbolic converter. The design approach proposed herein is capable of implementing a decision region defined by a multi-dimensional, non-linear boundary surface. By defining a "two-valued" subspace of the boundary surface, a Backprop paradigm neural network is used to model the boundary surf ace. An input vector is tested by the neural network boundary model (along with perceptron logic gates) to determine whether the incoming vector point is within the decision region or not. Experiments with two qualitatively different kinds of nonlinear surface were carried out to test and demonstrate the design approach.

Wavelet transform provides an alternative to the classical Short-Time Fourier Transform (STFT). In contrast to the STFT, which uses a single analysis window, the Wavelet Transform uses shorter windows at higher frequencies and longer windows at lower frequencies. For some particular wavelet functions, the local maxima of the wavelet transform correspond to the sharp variation points of the signal. As an application, wavelet transform is introduced to the character recognition. Local maximum of wavelet transform is used as a local feature to describe character boundary. The wavelet method performs well in the presence of noise. The maximum of wavelet transform is also an important feature for analyzing the properties of brain wave. In our study, we found the maximum of wavelet transform was related to the P300 latency. It provides an easy and efficient way to measure P300 latency.

Auditory graphsdisplays that represent graphical, quantitative information with soundhave the potential to make graphical representations of data more accessible to blind students and researchers as well as sighted people. No research to date, however, has systematically addressed the attributes of data that contribute to the complexity (the ease or difficulty of comprehension) of auditory graphs. A pair of studies examined the role of both data density (i.e., the number of discrete data points presented per second) and the number of trend reversals for both point estimation and trend identification tasks with auditory graphs. For the point estimation task, results showed main effects of both variables, with a larger effect attributable to performance decrements for graphs with more trend reversals. For the trend identification task, a large main effect was again observed for trend reversals, but an interaction suggested that the effect of the number of trend reversals was different across lower data densities (i.e., as density increased from 1 to 2 data points per second). Results are discussed in terms of data sonification applications and rhythmic theories of auditory pattern perception.

Turing patterns have been studied for over 50 years as a pattern forming mechanism. To date the current focus has been on the reaction mechanism, with little to no emphasis on the diffusion terms. This work focuses on combining the simplest reaction mechanism possible and the use of nonlinear cross diffusion to form Turing patterns. We start by using two methods of bifurcation analysis to show that our model can form a Turing instability. A diffusion model (along with some variants) is then presented along with the results of numerical simulations. Various tests on both the numerical methods and the model are done to ensure the accuracy of the results. Finally an additional model that is closed to mass flow is introduced along with preliminary results. / vi, 55 leaves : ill. ; 29 cm.

Many studies of the Archaic Greek kore focus exclusively on stylistic considerations in an attempt to date these statues more and more accurately. Other studies propose various meanings for the kore. Each of these approaches can be extremely subjective, with the result that the large body of extant literature about the kore tends to be repetitive and argumentative in nature, and, with several exceptions, does not advance the understanding of the kore to any appreciable degree past where it had developed by the 1980s. I use a different, more empirical methodology to study a small group of korai, found in the 1880s near the Erechtheion on the Athenian Akropolis. Symmetry analysis of the patterns painted onto these korai at the time of their creation reveals both consistency of pattern use through the period of seventy years between c560 BCE and c490 BCE, as well as some anomalous patterns. I tabulate the various patterns, as well as their frequency of occurrence, and briefly speculate that there is a correlation between the pattern consistencies and anomalies and events in the known historical record, such as the mid-6th century rule by the Peisistratids and the democratic reforms of Kleisthenes. I also propose other directions in which the study of the kore could be taken using symmetry analysis. / Thesis (Master, Classics) -- Queen's University, 2008-01-31 13:25:56.567

No description available.

Digital images acquired at construction sites contain valuable information useful for various applications including As-built documentation of building elements, effective progress monitoring, structural damage assessment, and quality control of construction material. As a result there is an increasing need for effective methods to recognize different building materials in digital images and videos. Pattern recognition is a mature field within the area of image processing; however, its application in the area of civil engineering and building construction is only recent. In order to develop any robust image recognition method, it is necessary to choose the optimal machine learning algorithm. To generate a robust color model for building material detection in an outdoor construction environment, a comparative analysis of three generative and discriminative machine learning algorithms, namely, multilayer perceptron (MLP), radial basis function (RBF), and support vector machines (SVMs), is conducted. The main focus of this study is on three classes of building materials: concrete, plywood, and brick. For training purposes a large-size data set including hundreds of images is collected. The comparison study is conducted by implementing necessary algorithms in MATLAB and testing over hundreds of construction-site images. To evaluate the performance of each technique, the results are compared with a manual classification of building materials. In order to better assess the performance of each technique, experiments are conducted by taking pictures under various realistic jobsite conditions, e.g., different ranges of image resolutions, different distance of camera from object, and different types of cameras.

This thesis examines the feasibility of using chemiluminescence (CL) for detection in miniaturised analytical systems. The aim of this project was to design a miniaturised device that could potentially be used for the remote sensing of metal ions. The development of miniaturised analytical devices for use with two well known chemiluminescent reactions; namely the tris(2,2'-bipyridyl)ruthenium (II) reaction and the luminol reaction, for the detection and quantification of codeine and cobalt (II) respectively are discussed. Chapter 1 introduces the concept, the manufacture, operating principles and applications of miniaturised analytical systems, while the use of chemiluminescence, its requirements and applications as a sensitive, selective yet simple method of detection are reviewed in chapter 2. Chapter 3 describes the manufacture and development of the robust and practical miniaturised analytical devices used for the analyses described in chapters 4 and 5. Several novel developments are described in this chapter. These included the use of thicker top plates that enable the reservoirs to be contained within the single unit structure. This design was intended to prolong the lifetime of the chip system and increase the available reagent volumes. A microwave furnace for thermal bonding of the two glass plates was also used and the detection of the chemiluminesence produced in the chip system was carried out from underneath the chip base. Chapter 4 details the application of the tris(2,2'-bipyridyl)ruthenium (II) (TBR) for analysis in a miniaturised analytical system. This reaction was selected as a model chemiluminescence reaction for the optimisation of the detection system in order to measure the very low levels of light produced. The incorporation of non-ionic surfactants into the analysis and their effect on the enhancement of the chemiluminescence emission intensity and the modification of electroosmotic flow is discussed. A quantitative analysis of codeine was then successfully performed using this set-up. The points for the codeine concentrations of 5x10⁻⁷ to 1x10⁻⁴ mol 1⁻¹ were plotted to give a linear calibration plot. The equation of the line was y = 6.0136x + 0.0949, R2 = 0.9999, where x was the codeine concentration in mol 1⁻¹ and y was the mean CL emission intensity in mV. A limit of detection for codeine was determined at the 95% confidence limits to be 8.3x10-7 mol 1⁻¹ codeine, with an RSD of 8% (n=5) at the 5x10⁻⁵ mol 1⁻¹ level. The sample throughput time including removal of products and water wash was found to be an average of 2 minutes. The work described in chapter 5 builds on the findings of chapter 4 and examines the use of the luminol reaction in a miniaturised analysis system for the quantification of cobalt (II) ions. A multivariate experimental design programme was carried out as part of the work described in this chapter to simultaneously optimise most of the reagent variables. The application of cationic surfactants to this reaction in the miniaturised analysis system is also discussed, with particular emphasis on the observed enhancement of chemiluminescence emission intensity and lifetime, and the modification of the electroosmotic flow characteristics. A quantitative determination of cobalt nitrate was successfully carried out with a calibration over six orders of magnitude. The equation of the linear portion of the graph (10-10- 10-8 mol 1⁻¹) was found to be y = 64.625x + 735.71 with R2 = 0.999, where x (n=3) was the concentration in mol 1⁻¹ and y was the mean CL emission in mV. The limit of detection for cobalt nitrate at the 95% confidence limits was determined as -4x10⁻¹¹ mol 1⁻¹ which equates to 0.01 ng ml⁻¹ cobalt nitrate. An RSD of 6.9% (n=3) was obtained for the 1x10-8 mol 1⁻¹ standard. The sample run time was approximately 12 minutes, which resulted in an average overall throughput time of 15 minutes. The conclusions and ideas for future work are detailed in chapter 6.

No description available.

The main trend of computer graphics is the creation of photorealistic images however, there is increasing interest in the simulation of artistic and illustrative techniques. This thesis investigates a profile based technique for automatically extracting artistic sketches from regular grid digital elevation models. The results resemble those drawn by skilled cartographers and artists.The use of cartographic line simplification algorithms, which are usually applied to complex two-dimensional lines such as coastlines, allow a set of most important points on the terrain surface to be identified, these form the basis for sketching.This thesis also contains a wide ranging review of terrain representation techniques and suggests a new taxonomy.