Integração e análise de dados aerogeofísicos por meio da aplicação de técnicas de processamento digital de imagens e classificação não supervisionada: o exemplo do Greenstone Belt Rio das Velhas, quadrilátero ferrífero, MG

Abstract

Poor outcrop and deep weathering characterize the Rio das Velhas Greenstone Belt in the southeastern from Brazil. This paper summarizes the use of high-density airborne survey for mineral exploration studies based on interpretation enhancements of magnetic, radiometric and frequency domain electromagnetic data using image-processed methods and an unsupervised classification. The generated products provide new insights and an excellent tool for mapping and trace individual lithological units, improving the information content of the single geophysical channels. The geophysical images were processed using different combinations. The best product was the analytical signal amplitude and inclination integrated by IHS transformation. Information extracted from this image maps the geology and lineament patterns at both regional and local scales. The K-means technique using ten classes was also applied to the geophysical data. These results enhance the lithologies mapped by the field geologists at the 1:100.000. Also shows important host rocks and different gold mineralized geological domains. Such domains host the known gold mineralization, illustrating the utility of these techniques to improve the geological knowledge in the study area.