Uma abordagem para mineração de dados e visualização de resultados em imagens batimétricas

• Ano de defesa: 2012
• Autor(a) principal: Gonzalez, Luis Fernando Planella
• Orientador(a): Ruiz, Duncan Dubugras Alcoba
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação
• Departamento: Faculdade de Informáca
• País: BR
• Palavras-chave em Português: INFORMÁTICA; MINERAÇÃO DE DADOS (INFORMÁTICA); PROCESSAMENTO DE IMAGENS
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://tede2.pucrs.br/tede2/handle/tede/5178

Bathymetry is the measurement of the depth at various places in a body of water, as well as information derived from such measurements. It has several important applications, and has been attracting increasing interest over the last years. Bathymetry maps may cover the entire extent of the Earth globe. However, the analysis of such maps by visual inspection solely is difficult, due to subtle variations on the seafloor conformation. Thus, it would be interesting to have available computational tools capable of assisting a domain expert in problems related to bathymetry images, by analyzing them automatically or semi-automatically. The contribution of this work is an approach to use data mining for such analysis, and an iconography for results visualization, as well as map characteristics. We propose techniques to process input images, in order to extract records and their features, which can be processed by classic data mining algorithms. We also propose an iconography for visualization of knowledge discovery process results, as well as characteristics of areas in the processed map. Finally, the proposed approach is tested by applying it on a real database, under a domain expert supervision.