Acerca da Detecção Automática de Silhuetas de Manchas de Óleo no Mar

Early detection of oil spills is paramount for quick response measures. The use of images acquired by UAVs (unmanned aerial vehicles), combined with computer vision techniques and deep-learning, presents automation opportunities to oil spill surveillance activities. However, proper training of models capable of detecting objects and region of interest in images often requires a large annotated dataset. In this work, we present a semi-automatic method to accelerate the process of annotating oil spill images generating silhouettes at pixel level based on the mean shift and simple linear iterative clustering segmentation algorithms. As a result, we built an annotated database of oil spill images by using the proposed methodology.