Today, drones equipped with high-resolution cameras and integrated high-performance GPUs are increasingly available, even at affordable prices. Such sensory and computational capabilities, combined with recent advances in deep learning and computer vision, now offer the possibility of implementing decision-making systems directly on board the drone, opening up a scenario in which drone flight is entirely autonomous. Countless applications, from video surveillance to precision agriculture, could benefit from using drones, as they can provide a low-cost alternative to traditional methodologies. This discussion paper reviews some of the research we are conducting in this area, particularly in crowd detection for safe landing and rescue mission support. The paper concludes with an overview of the current research we are carrying out and potential future directions.
Computer Vision Meets Drones: Our Research Experience
Giovanna Castellano;Corrado Mencar;Gennaro Vessio
2022-01-01
Abstract
Today, drones equipped with high-resolution cameras and integrated high-performance GPUs are increasingly available, even at affordable prices. Such sensory and computational capabilities, combined with recent advances in deep learning and computer vision, now offer the possibility of implementing decision-making systems directly on board the drone, opening up a scenario in which drone flight is entirely autonomous. Countless applications, from video surveillance to precision agriculture, could benefit from using drones, as they can provide a low-cost alternative to traditional methodologies. This discussion paper reviews some of the research we are conducting in this area, particularly in crowd detection for safe landing and rescue mission support. The paper concludes with an overview of the current research we are carrying out and potential future directions.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
