Cartografía de coberturas del suelo mediante Sensores Remotos, de la ciudad de Resistencia, Argentina (2013)
; Insaurralde, Juan Ariel
; Cardozo, Osvaldo Daniel 
Otros responsables
Fecha
01/2014
Editorial
Universidad Nacional del Nordeste. Facultad de Humanidades. Instituto de Geografía
Revista
Geográfica Digital
Título del libro
Nombre del evento
Fecha del evento
Institución organizadora
ISSN
1668-5180
eISSN
ISBN
Idioma
Español
Tipo del evento
Clasificación temática
Palabras clave
Resumen
Los Sensores Remotos se han convertido en los últimos tiempos en una herramienta importante para la planificación territorial, debido a la originalidad metodológica con la cual operan y el amplio espectro de observación de la superficie terrestre. El número de algoritmos disponibles en la literatura científica, permiten tratar individualmente muchas de las coberturas terrestres y caracterizarlas en base a su comportamiento espectral. En el presente trabajo se han aplicado dos métodos de clasificación de imágenes: árboles de decisión o reglas y clasificación no supervisada. También se han calculado índices de vegetación y de agua (NDVI y MNDWI), para identificar las coberturas y generar cartografía temática que sirva como base para posteriores estudios. Además, se ha elaborado la matriz de confusión con el fin de determinar la validez de la clasificación obtenida. Como resultado, se han obtenidos coberturas de vegetación, urbano, cursos y cuerpos de agua, clasificados a partir de técnicas oportunamente mencionadas, que brindan una fiabilidad global del orden del 61%, y si se consideran de manera individual (urbano y cuerpos y cursos de agua), las mismas alcanzan una precisión superior al 70%.
Remote Sensing has become in recent years an important tool for land use planning, due to the methodological originality with which it operates and the wide view of the Earth surface. The range of image processing algorithms available today in scientific literature allows that many coverages can be treated individually and can be characterized by their spectral behavior. In this paper we have applied different image classification methods such as decision tree or rules, unsupervised classification, have also been calculated Normalized Difference Vegetation Index (NDVI) and Modified Normalized Difference Water Index (MNDWI) to identify coverages and generate thematic mapper, as the basis for further studies. Furthermore, has been applied the confusion matrix in order to determine validity of the classification obtained. Have been obtained coverages vegetation, urban, courses and water bodies, that they have classified by mentioned techniques the previous paragraph, which provide a global reliability of the order of 61%, and if considered individually (urban and bodies and waterways) , they reach an accuracy above 70%.
Remote Sensing has become in recent years an important tool for land use planning, due to the methodological originality with which it operates and the wide view of the Earth surface. The range of image processing algorithms available today in scientific literature allows that many coverages can be treated individually and can be characterized by their spectral behavior. In this paper we have applied different image classification methods such as decision tree or rules, unsupervised classification, have also been calculated Normalized Difference Vegetation Index (NDVI) and Modified Normalized Difference Water Index (MNDWI) to identify coverages and generate thematic mapper, as the basis for further studies. Furthermore, has been applied the confusion matrix in order to determine validity of the classification obtained. Have been obtained coverages vegetation, urban, courses and water bodies, that they have classified by mentioned techniques the previous paragraph, which provide a global reliability of the order of 61%, and if considered individually (urban and bodies and waterways) , they reach an accuracy above 70%.
Aprobación
Revisión
Complementado por
Referenciado por
Licencia Creative Commons
Excepto donde se indique lo contrario, la licencia de este ítem se describe como by-nc-sa
