Gianfranco Di Pietro

Earth Obserbation and remote-sensing technologies have contributed significantly to spatial data analysis. Acquisition and dissemination of multispectral data and images is growing exponentially. EO science has need for increasingly high-performance artificial intelligences capable of facilitating a rapid and organized digital transition in order to foster innovative processes in the field of big data analysis. In the last decade, developments in machine learning applied to multispectral data has generated an increasing diffusion of specialized methods and models. These are tools capable of investigating and determining geological, geophysical, environmental, demographic parameters and more... through multi-temporal recognition, detection and quantification of significant pattern. Most of the methods developed involve massive use of data from satellite sensors combined with surveys on the ground or lab-data.

Ongoing research aims to develop software solutions and prototypes (or cloud platforms) for multispectral data analysis applied to land management and risk assessment cases.The research focuses on the use of data from significant time series for the analysis of land evolution that has occurred due to both natural and anthropogenic causes.

Taking advantage of what has already been developed in the use of algorithms for image classification from recent research in the field of remote-sensing, it is intended to develop systems that autonomously, or semi-autonomously, analyze spatial changes on a temporal and localized scale. The effectiveness of the studied solutions and the developed software prototypes will also be evaluated. In particular, the feasibility of deployment in non-academic settings and their active functionality in digital transition processes within the workflows of small and medium-sized administrations should be assessed.

One of the results expected is the development of a cloud-based geographical information system of Posidonia presence effect for morphodynamic patterns of the beach reconstruction process following storm surges in South-Est of Sicily. Other results expected are development and distribution of open-source software, systems and information technologies designed to be enjoyed by non-specialists.