Identificação de padrões morfoestruturais utilizando Clustering Large Applications, um estudo de caso no Quadrilátero Ferrífero

Abstract

Among the stages of a mining project, mineral research stands out, with the objective of identifying, studying and evaluating mineral deposits. In this specific stage, the inferred mineral resources are transformed into indicated and finally measured, and if their exploitation is feasible, into probable and/or proven mineral reserves. The discovery of these reserves is an impacting milestone for the industrial, technological and economic development of a society. The main objective of this article is to present the use of a machine learning technique to identify structures of particular geological interest, from satellite images. The technique applied was the Clustering Large Applications (CLARA) which is an unsupervised algorithm for clustering data, with high performance in massive databases. The area used as a case study was the Quadrilátero Ferrífero, one of the largest mineral provinces on the planet, located in the state of Minas Gerais, Brazil. The results of the CLARA model allowed the delineation of all the features that form the Quadrilátero Ferrífero. In this context, it is believed that this can be a good tool for selecting exploratory targets, reducing uncertainty and risk to investors. This not only attracts new companies for mineral research, but also expands the reserves of Brazilian mineral resources.