Evaluación y combinación de datos espectroscópicos de los granates como análogos minerales para el estudio de Marte

Abstract

Mars exploration is gaining more relevance as new advances in technology have allowed us to study its composition through rovers capable of landing on its surface. Due to the large amount of mineralogical data they collect, it is necessary to implement machine learning techniques to classify these samples. Thanks to the SuperCam tool of the Perseverance rover, it is possible to perform Raman and FTIR spectrometries in situ on the red planet. In this work, several classification algorithms for garnets trained using band parameters, PCA, and fusion of Raman and FTIR data are proposed with the aim of determining which one presents higher precision.