Detecting Airport Activity from Sentinel-2 Imagery During COVID-19 Pandemic by Using Deep Learning

Since the global spread of COVID-19 in 2020, in order to reduce infections, the movement of people has been severely restricted. As a result, the economic environment in commercial aviation suffered an unprecedented impact. Therefore, it becomes important to study the impact of COVID-19 on commercial aviation. In this study, in order to understand the changing trend of the number of airplanes as an index of the airport activity, we applied a method that utilizes convolutional neural networks to effectively detect airplane in Sentinel-2 images. From the detection, we successfully obtained the changing trend of the number of airplanes in important airports around the world since the outbreak of COVID-19, and found different changing trends in different areas that may reflect different reactions to COIVD-19 situation in each country.