Extracción de Mapas Temáticos a partir de la Clasificación en Imágenes Satelitales
A method for classifying multi-spectral satellite images based on some knowledge, called Multi-Model Classification Scheme (MMCS), is presented in this work. The MMCS is divided into two parts: Descriptive and Contextual. The descriptive part refers to the texture, geometrical shape and spectral fea...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Estado: | Versión aceptada para publicación |
| Fecha de publicación: | 2007 |
| País: | México |
| Institución: | Instituto Nacional de Astrofísica, Óptica y Electrónica |
| Repositorio: | Repositorio Institucional del INAOE |
| Idioma: | español |
| OAI Identifier: | oai:inaoe.repositorioinstitucional.mx:1009/2243 |
| Acceso en línea: | http://inaoe.repositorioinstitucional.mx/jspui/handle/1009/2243 |
| Access Level: | acceso abierto |
| Palabra clave: | info:eu-repo/classification/Inspec/Classifying multi-spectral satellite info:eu-repo/classification/Inspec/Geometrical shape info:eu-repo/classification/Inspec/Spectral features info:eu-repo/classification/cti/7 info:eu-repo/classification/cti/33 info:eu-repo/classification/cti/3304 info:eu-repo/classification/cti/120312 |
| Sumario: | A method for classifying multi-spectral satellite images based on some knowledge, called Multi-Model Classification Scheme (MMCS), is presented in this work. The MMCS is divided into two parts: Descriptive and Contextual. The descriptive part refers to the texture, geometrical shape and spectral features of a region. The contextual part refers to the topological relationships among the image regions. Data mining techniques are used to discover both types of knowledge. The extracted knowledge from the regions of interest, coming from their descriptive features (texture, shape and spectral) and contextual features (topological relationships), is organized in a knowledge representation scheme based on semantic networks. The MMCS reclassifies the previously classified regions by the parametric algorithms (Minimum Mean Distance, Parallelepiped, Maximum Similarity and Mahalanobis Distance) in order to improve the algorithm classification results that only take into account the spectral features of the satellite image. The Halcon Image-Analysis System (version. 7,1) was used to process satellite images. Matlab was used to preclassify a satellite image using parametric algorithm. C++ was used to classify the regions in the MMCS. Experiments were done over 30 segments of multi-spectral satellite images (SPOT-5) from two areas in Mexico, specifically the coastal zones of the Port of Veracruz and Campeche. 20 segments were used during the training processes and the remaining 10 at the classifiers testing processes. |
|---|