Detección de cambios en la cobertura vegetal mediante interpretación de imágenes Landsat por redes neuronales artificiales (RNA). Caso de estudio: Región Amazónica Ecuatoriana

Abstract

The interpretation of classes and change detection in the vegetation cover of large areas are activities that are made possible by the use of technologies and methods associated to Remote Sensing. Satellite images of medium and high spatial and spectral resolution are fundamental tools for the execution of projects with objectives of classification of vegetal cover and detection of its temporal variations. To exploit the use of digital information of territory recovered by the satellite images, and in order to optimize the resources invested in the tasks of classification and interpretation, it is necessary to have tools and methods that allow the automation of the processes involved and prove to be the Artificial Neural Networks (ANNs) an adequate mechanism for the execution of these processes. The main objective of this work is to validate a methodology for the identification of changes in the vegetation cover of an area located in the Ecuadorian Amazon. The applied methodology seeks the change detection in the coverage of native forests prevailing in the study region.