Detecting water using UAV-based GNSS-Reflectometry data and Artificial Intelligence

GNSS Reflectometry (GNSS-R) is a consolidated remote sensing technique that exploits the back-scattered GNSS signals to retrieve information about the Earth surface. By the use of GNSS-R sensors onboard UAVs, this tool can bring significant advantages in agriculture, especially for the detection of the presence of water/moisure in specific rural areas. However, to perform this detection, GNSS-R needs a priori calibration of a threshold on the received reflected power in order to distinguish the presence/no prence of water. This is a limitation for the adoption of this approach at large scale. The work presented in this paper aims at overcoming such a limitation, proposing a novel approach based on artificial intelligence for the automatic water detection on the Earth surface, avoiding a priori, empirical, thresholding.