Loading...
Please wait, while we are loading the content...
Similar Documents
Detecção de zonas de baixa cobertura vegetal em imagens de satélites através do uso de operadores da Morfologia Matemática
| Content Provider | Semantic Scholar |
|---|---|
| Author | Guerra, Carlos Eduardo Matemática, Colegiado De |
| Copyright Year | 2007 |
| Abstract | The goal of this work is to detect some majors patterns related to forest absence. This detection is performed by means of morphological operators; using MATLAB Imaging processing toolbox (Gonzales, 2003). First at all, a color common ".tif" TM image is read and, each channel is separated and transformed into a binary one. Since the third channel is more sensitive to urban areas and deforested zones, the result binary image will be fulfilled by continues high pixels (value 1). Then, erosion and dilatation operators are applied to the third channel binary image. Finally, a boundary tracing operator is applied to the eroded image and the boundaries are moved to the ".tif" image in order to show the located zones. The methodology woks very well but there are some patterns which are not recognized correctly, like clouds; for instance. Another Critical point is the type and size of structuring element. In this case, small deforested areas simply can be taken out if we set up a structuring element with to many neighborhoods. In others cases, some structuring elements tested performed erosions on interested areas. In general the methodology is good for a quick deforest evaluation, since the procedures applied are completely automatic and they are vary fast. Palavras-chave : Morfologia Matematica, imagem TM, cobertura vegetal. 1. Introducao Em trabalhos e estudos ambientais e muito comum o uso de imagens de sensoriamento remoto, particularmente a fotogrametria e as imagens de satelites (Florenzano, 2002). Na Amazonia em particular, estes estudos tem se direcionado na caracterizacao da ocupacao humana, no uso do solo e nas mudancas de cobertura vegetal. Para um pesquisador interessado em avaliar a extensao de zonas com baixa ou nenhuma cobertura vegetal (cidades, areas desmatadas, savanas,...) as imagens de satelites sao fontes valiosas de informacoes, pois possibilitam uma visao sinoptica do ambiente considerado, alem de permitir o calculo aproximado de extensoes de areas de interesse. Para facilitar o estudo das imagens existe uma colecao de programas e pacotes de software, desde simples aplicativos de processamento de imagens ate os complexos sistemas de geoprocessamento e os SIG's (sistema de informacoes geograficas). No caso do processamento de imagens digitais, a ideia principal e construir uma serie de procedimentos (computacionais) para analise de dados obtidos por sensores orbitais, agilizando as tarefas manuais realizadas durante o processo de interpretacao. Tais processos incluem a delimitacao de areas, confeccao de mapas, dentre outros (Gonzalez,1987 e Silva, 1999 ). Uma das tecnicas de processamento de imagens que podem ser aplicadas trata-se dos operadores morfologicos obtidos da Morfologia Matematica. A origem da |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://marte.dpi.inpe.br/col/dpi.inpe.br/sbsr@80/2006/11.16.07.46/doc/5799-5806.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
