The Use of Drones and Deep Learning to Identify Igneous Rocks and Fractures

Abstract

Igneous (or magmatic) rocks may be critical elements in the petroleum systems. They may be related with generation, migration and trapping of oil and gas and also generate hydrothermal processes that can affect the reservoirs. In Campos and Santos basins, they occur as volcanic and intrusive bodies, affecting the pre-, intra- and post-salt sequences. Associated with fractures and faults, igneous intrusives were emplaced during reactivation of preexisting discontinuities or weakness zones, although they also cross cut them. In the continental area, adjacent to the Campos and Santos basins, intrusive igneous bodies crop out as dykes, sills and stocks. The area selected for this study is the "Pontal do Atalaia", in the city of Arraial do Cabo (RJ). The use of drone images allowed to identify the structural framework of the igneous bodies and fractures patterns associated with them. By developing tools of deep learning, it was possible to analyze the fractures and compare the results with visual interpretation. In both cases, the results indicate the predominance of fractures along roughly N-S and E-W trends. The outcomes allow the comparison and correlation to the offshore geological features to better understand the role of the igneous rocks and fractures in the tectonic evolution of Campos and Santos basins.