In the framework of the AI4Copernicus open call, we have secured funding for the development of a micro-project, LIFT Sentinel – LIFT Sentinel AI Terrain Detector.
The project’s aim is to develop an automated classification system for satellite imagery using deep learning to identify aquatic, urban, rural, and forested areas. This system will enhance the existing module for statistical and advanced analyses in the LIFT software and will be validated using real-world datasets.
The objective of our project is to utilize the AI4Copernicus service, “Deep network for pixel-level classification of S2 patches,” to develop an automated classification system for satellite imagery using deep learning methodologies to identify water, urban, rural, and forest areas. We will utilize the Sentinel-2 L2A data provided by the DIAS provider to enhance our existing module for statistical and advanced analyses. This module encompasses a wide array of statistical analyses, predominantly based on attribute data with its associated location. The upgraded module will feature added functionality for automatic recognition of urban and rural areas, as well as other use-cases such as measuring water or forest areas within spatial or administrative units.
From a technical standpoint, we intend to integrate our existing services, including our LIFT software and the analysis module, with the AI4Copernicus service to ensure automated classification of satellite imagery. From a business perspective, this project will benefit our remote sensing department by streamlining their workflow and improving the accuracy of the results.