AI4Sen2Cor

Supporting the monitoring of the Earth’s condition by observing its changes and variability is the main target of the S2 mission.

In the spirit of the S2 goals, AI4Sen2Cor is designed to extend the capability of the Sen2Cor_3 processor by applying a synergic approach that combines systematic AI-based algorithms with the existing and available Sen2Cor quantities and qualities.

The AI4Sen2Cor study’s first goal is adding Geospatial Detection capability to Sen2Cor_3 with a systematic production of AI-enhanced and spectral-based single-sensing-time (Static) Augmented Scene Classification (S-ASCL). The second main goals is to produce a tool to analyse a series of S-ASCL belonging to different observations (Sensing Times) of a given tile to produce a Temporal-ASCL (T-ASCL), where temporal variations can be analysed and associated statistics produced.

Conference proceedings:

• Francesco C. Pignatale, Davinder P. Singh, Satish Madhogaria, Bodo Werner, Patrick Griffiths  “AI4SEN2COR: A SEN2COR ENHANCEMENT FOR GEOSPATIAL DETECTION” Proc. of the 2023 conference on Big Data from Space (BiDS’23)