Análisis multitemporal para la detección de cambios en el Municipio de Socha Boyacá y Popayán Cauca utilizando imágenes de ultra alta resolución espacial

Abstract

Remote Sensing has several applications, by processing satellite images it is possible to obtain high-level spatial information, thus in Google Earth Engine -GEE it is possible to deploy Machine Learning -ML classifiers to identify Land Use Land Cover- LULC data, these classifications are inputs in different studies, in this case for implementing Multipurpose Cadastre in Colombia, it requires targeting efforts in specific territories. With this work, we pretend to identify the spatial changes that could take place in the case study Socha Boyaca and in the validation case Popayan Cauca between 2015 and 2020, using Vexcel Ultracam D very high spatial resolution images provide by IGAC, the accuracy of different ML classifiers available in GEE was assessed, the results indicated Random Forest classifier showed higher accuracy than others.