A tendência de utilização de dados voluntários e colaborativos no contexto de fenômenos naturais é crescente. Esse fato, aliado aos alagamentos e inundações que ocorrem na cidade de São Paulo, traz a possibilidade de exploração sobre o modo voluntário e colaborativo de geração e transmissão do dado geográfico de forma dinâmica. Isso é proporcionado por tecnologias acessíveis à população, como a Internet, o GPS (Global Positioning System) e demais sistemas de localização embarcados em celulares. A presente pesquisa tem como objetivo a proposta de um esquema conceitual para um sistema dinâmico e colaborativo de mapeamento dos pontos alagados, cuja fonte de dados advém das pessoas equipadas com aparelhos celulares que permitem a sua localização. Os resultados apresentados correspondem aos esquemas conceituais do sistema, bem como ao protótipo Pontos de Alagamento - mapa disponibilizado via web com os pontos de alagamento da cidade, fornecidos no momento da ocorrência do evento por pessoas comuns. O protótipo foi desenvolvido por meio da plataforma livre e de código aberto Crowdmap/Ushahidi. O sistema foi avaliado através de um questionário respondido por usuários, os quais opinaram sobre a viabilidade do mesmo, bem como os ajustes que devem ser realizados para o uso efetivo da população. Constatou-se a complexidade e as particularidades da aplicação para alagamentos e inundações, em especial com relação à questão temporal. / The trend of using volunteered and crowd data in natural phenomenon contexts is growing. This fact coupled with flooding that occurred in the city of São Paulo, brings the possibility of exploration about the voluntary and collaborative approach to the generation and transmission of the geographic data dynamically. And these are provided by technologies accessible to the population, such as internet, GPS (Global Positioning System) and other positioning systems embedded in cell phones. This research aims to propose a conceptual scheme for a dynamic and collaborative mapping system of flooding, whose source of data corresponds to those equipped with mobile devices that enable location. The results correspond to the conceptual schemes of the system as well as the prototype Points of Flooding available map on the web with the points of flooding, provided at the time of the event by people. The prototype was developed through the free and open source platform Crowdmap/Ushahidi The system was evaluated by a questionnaire answered by users, who opined about the viability of this as well as the adjustments that must be done for the effective use by population. There has been a thoughtful analysis of the complexity and particularities of the application to flooding, particularly related to temporal issue.

The term “home range” refers to the area in which an animal spends most of its time during everyday activities. This study examined the effects of four different home range estimation techniques on the proportions of habitats located therein. The study utilized a point dataset collected for twenty individual Florida Black Bears (Ursus americanus floridanus), occurring in five different areas throughout the state of Florida. Each dataset was used to create home ranges using the following techniques: (1) Minimum Convex Polygon, (2) Kernel Density Estimation, (3) Characteristic Hull Polygon, and (4) Time-Geographic Density Estimation, a new home range estimator which has not been thoroughly tested prior to this study. A dataset of land cover types was clipped with each home range and the areas of habitats were recalculated. The proportion of each land cover type was evaluated and the results compared first within each dataset, then between all datasets used. Repeated measures analysis of variance (ANOVA) was used to determine whether the four home range estimation techniques produced significantly different areas and proportions of each habitat type. These results were then evaluated to determine whether the method of home range estimation has an effect on which land cover types are most utilized by a species and, therefore, which habitats are considered preferable. While the choice of home range estimation did not have an effect on which habitats were determined to be most frequently visited, it did affect the amount of each habitat found within each home range. Furthermore, there was a statistically significant change in the amount of developed areas, specifically between the characteristic hull polygon and kernel density estimation methods. These results suggest the choice of home range estimator affects habitat analysis and that researchers should use the method best suited for the dataset.

The South African coastline has a high potential for mariculture development, particularly in land-based systems, due to excellent water quality, good infrastructure, and relatively cheap land in certain areas. Development of mariculture however, has been slow primarily due to the absence of a national sector development plan to coordinate contribution to development by government, industry and academia. Recent mariculture development plans however, supported by government commitment to stimulate coastal development, offers new opportunity for the sector's growth. These new sector development plans require a multi-disciplinary intensive information base with a strong regional and national spatial component. Geographic Information System (GIS) is a spatial analytical tool, which is capable of handling such large coastal databases and analysing them. Nevertheless, many mariculture GIS planning applications, often developed in isolation and with limited practical use for decision makers, highlighted the need to develop GIS in relation to the mariculture sector development plan. Hence, the aim of this study was to develop and evaluate a GIS for the national and regional mariculture sector planning process in South Africa, in order to identify potentially suitable areas along the South African coastline for mariculture operations. The GIS development in this study was based on the mariculture planning approach developed by PAP/RAe (1996), which recognizes the variations of scales in mariculture planning. Three case studies, representing the diversity of the South African sector plan, were selected to develop the GIS. They included: a national suitability analysis for all land-based culture systems, a specific national analysis on abalone ranching, and a detailed regional analysis of land-based culture and abalone ranching along the Namaqualand coastline. The GIS was developed separately for each case study based on a strategy consisting of six main phases, including: identification of project requirements, developing analytical framework, selection and location of data sources, organization and manipulation of data, analysing data and verifying and evaluation of the outputs. Biophysical, coastal use, and infrastructure criteria were collected, sorted and analysed to identify development constraints. Based on a set of conditions, and Boolean logic and arithmetic operations, unsuitable areas were identified and eliminated. Main constraints for national land-based development included competition over space along KwaZuluNatal Province coastline and restricted access to the coast along the south region of the Northern Cape Province, Wild Coast along the Eastern Cape Province, and Maputaland along KwaZulu-Natal Province. Ten areas along the country coastline were therefore identified as potentially suitable for land-based mariculture. Furthermore, South African abalone ranching potential was found to be limited mostly along the Western and Northern Cape Province's coastline due to the high risk of Paralytic Shellfish Poison (PSP) occurrences, alternative resource use and activities such as commercial fishery, poaching, and conservation. Hence, a total of nine areas along the Northern Cape, Western Cape, and Eastern Cape coastlines were identified as premier areas for abalone ranching. The regional study along the Namaqualand coastline revealed potential conflict between mining activity and mariculture development due to security issues and restricted access to the coast. Land-based mariculture development was confined to the four main coastal urban areas. The highest potential for land-based mariculture was along Port Nolloth and Kleinsee coastlines, whereas marine-based (i.e. abalone ranching) potential was poor along the north part of the coast due to intense marine mining activities, restricted access to the coast, and low kelp bed density. High potential abalone ranching areas were identified south to Kleinsee, and around Port Nolloth. Potential conflict with marine mining activity was minimal since it was localised and not related to kelp bed locations. It was concluded that GIS is a relevant and compatible tool for South African mariculture sector planning. However, future development of GIS as integrated planning tool in mariculture and coastal planning, requires updated spatial data (e.g. recreational activity), and continued interaction among project planners, mariculture specialists and GIS analysts.

This study employed geographic information systems (GIS) technology to evaluate the vulnerability of groundwater to pesticide pollution. The study area included three provinces (namely, Kanchana Buri, Ratcha Buri, and Suphan Buri) located in the western part of central Thailand. Factors used for this purpose were soil texture, percent slope, primary land use, well depth, and monthly variance of rainfall. These factors were reclassified to a common scale showing potential to cause groundwater contamination by pesticides. This scale ranged from 5 to 1 which means high to low pollution potential. Also, each factor was assigned a weight indicating its influence on the movement of pesticides to groundwater. Well depth, the most important factor in this study, had the highest weight of 0.60 while each of the remaining factors had an equal weight of 0.10. These factors were superimposed by a method called “arithmetic overlay” to yield a composite vulnerability map of the study area. Maps showing relative vulnerability of groundwater to contamination by pesticides were produced. Each of them represented the degree of susceptibility of groundwater to be polluted by the following pesticides: 2,4-D, atrazine, carbofuran, dicofol, endosulfan, dieldrin & aldrin, endrin, heptachlor & heptachlor epoxide, total BHC, and total DDT. These maps were compared to groundwater quality data derived from actual observations. However, only the vulnerability maps of atrazine, endosulfan, total BHC, and heptachlor & heptachlor epoxide showed the best approximation to actual data. It was found that about 7 to 8%, 83 to 88% and 4.9 to 8.7% of the study area were highly, moderately, and lowly susceptible to pesticide pollution in groundwater, respectively. In this study a vulnerability model was developed, which is expressed as follow: V = 0.60CW + 0.10CS + 0.10CR + 0.10CL + 0.10CSL. Its function is to calculate a vulnerability score for a certain area. The factor “V” in the model represents the vulnerability score of a certain area, whereas CW, CS, CR, CL, and CSL represent the values or classes assigned to well depth, soil texture, monthly variance of rainfall, primary land use, and percent slope in that area.

No description available.

This thesis investigates the application of GIST features [13] to the problem of object detection in images. Object detection refers to locating instances of a given object category in an image. It is contrasted with object recognition, which simply decides whether an image contains an object, regardless of the object's location in the image. In much of computer vision literature, object detection uses a "sliding window" approach to finding objects in an image. This requires moving various sizes of windows across an image and running a trained classifier on the visual features of each window. This brute force method can be time consuming. I investigate whether global, easily computed GIST features can be used to classify the size and location of objects in the image to help reduce the number of windows searched before the object is found. Using K–means clustering and Support Vector Machines to classify GIST feature vectors, I find that object size and vertical location can be classified with 73–80% accuracy. These classifications can be used to constrain the search location and window sizes explored by object detection methods.

No description available.

No description available.

<p>Blood in Canada is donated by a volunteer base that is increasingly challenged, through a combination of demographic aging and immigration, to meet the needs of the health sector. Canadian Blood Services, the agency with the mandate to manage blood products in Canada with the exception of Quebec, is therefore actively involved in the development of programs to help increase the number of donors, to improve the retention of existing donors, and to increase the frequency of donation of repeat donors. An important factor that influences blood donation is the accessibility to clinics. Accessibility to clinics is determined by the location of clinics, the resources allocated to each clinic in terms of number of beds and hours of operation, and the distribution of the population in the areas serviced by the clinics. The objective of this research is to investigate, given a set of fixed sites for clinic locations and population characteristics, the potential for increasing the donor yield as a function of accessibility. A case study is presented of the Hamilton Census Metropolitan Area, in Canada. Using donor and clinic data provided by Canadian Blood Services, and census information, an objective function is derived by estimating a generalized linear model of donations. The objective function is maximized globally using Genetic Algorithm techniques, subject to total resources available for clinic operations. The results suggest that an optimized allocation of resources to clinic sites has the potential to increase the donor yield by approximately 50% of the current donor base.</p> / Master of Arts (MA)

The main scope of this project is to rebuild and deploy web applications that will help share historical artifacts related to the Santa Ana watershed. This project is designed to give a consistent user interface and add extra tools to enhance the functionality of existing applications at the Water Resources Institute (WRI) at California State University San Bernardino (CSUSB). The purpose of this project is to migrate all applications to a single server and update the applications using ArcGIS Server 9.2. Also to give a consistent look to the applications and to make them user friendly.