Secure UAV-based system to detect and filter sea objects using image processing

Abstract

In the last years irregular immigration has aroused again due to the fact of war in some regions like middle east and in continents like Africa. Because of that, in this paper we presented a novel solution based in new emerging technology to face off the immigration problem focused on the case of sea-type immigration. The presented system is composed by unmanned aerial vehicles (UAV), also known as RPAs or drones, used to fly autonomously over an area in order to detect unallocated potentials immigration boats. The method used is the image processing using combined algorithms and software like OpenCV2 to detect certain patterns in images to determinate whether it is a possible immigration boat or not. Besides, it has been included some prof of concepts to show the effectiveness of the presented system. As main results we was able to determine and detect using aerial pictures if a boat it is a boat with an 82.3% of accuracy (according to the F1 score accuracy test).