Automatic dataset builder for Machine Learning applications to satellite imagery

摘要

In this paper, the architecture of an innovative tool, enabling researchers to create in an automatic way suitable datasets for Artificial Intelligence (AI) applications in the Earth Observation (EO) context, is presented. Two versions of the architecture have been implemented and made available on Git-Hub, with a specific Graphical User Interface (GUI) suitable for non-expert users. The tool has been designed to work with different types of sensors, but up to now has been tested with Sentinel-2 and Sentinel-1 data. We strongly believe that this tool will be of great usefulness for researchers applying AI to EO and Remote Sensing (RS). At the best of our knowledge, there is not a similar freely available tool, collecting the same benefits.