Satellite Image Time Series Analysis for Big Earth Observation Data

The development of analytical software for big Earth observation data faces several challenges. Designers need to balance between conflicting factors. Solutions that are efficient specific hardware architectures can not be used in other environments. Packages work on generic and open standards will have the same performance as dedicated solutions. Software assumes its users computer programmers flexible but may difficult learn a wide audience. This paper describes sits, an open-source R package satellite image time series analysis using machine learning. To allow experts use imagery fullest extent, sits adopts time-first, space-later approach. It supports complete cycle land classification. Its API provides simple powerful set functions. works different cloud computing Satellite input learning classifiers, results post-processed spatial smoothing. Since methods accurate training data, includes quality assessment samples. also validation accuracy measurement. thus comprises production environment EO analysis. We show this approach produces high cover maps through case study Cerrado biome, one world's fast moving agricultural frontiers year 2018.