Aplicación de técnicas de clasificación basadas en un sistema de aprendizaje para la generación de un mapa de usos de suelo

Abstract

The use of remote sensing techniques and geographic systems (GIS), now are unavoidable for the study of territory, giving considerable help in the generation of cartography. For a best proficiency of the information provided for the satelital images, it is well known the use of different techniques of classification; in this work were applyed differents methods of classification support on algorithms of knowledge, as decision trees and rules generatives of a land use map. The work initiated with the georeferencing of the images, then with the help of other satelital images were delimited areas concern to the distinct kinds differentiated on the zone of study, to be used a part as training set and other for validation. The training set was put to a lapse of time of knowledge support on the algorithms used, then of the wich was proceeded to the classification of the totally of images. From the analysis and validation of the obtained results with the application of the indicated algorithms for the classification of images, it is concluded the viability of method for the generation of a land use map.