My research aims at bridging the gap between optical remote sensing and applied marine ecology by developing novel techniques to map underwater ecosystems, with a particular focus on macroalgal forests and seagrass meadows. The resulting cartographies represent the extent and cofiguration of key habitats, supporting the analysis of underwater landscapes, ecological status assesments, and the detection of early signs of ecosystem degradation or collapse.
I take a multiscale approach, integrating 3D underwater models with UAS and satellite imagery. To transform raw imagery to ecosystem maps, I apply a range of classification methods: from manual digitalization and classic machine learning algorithms to deep learning techniques. I particularly emphasize the critical role of high-accuracy field data to validate AI-driven outputs. As a result, my work spans the full research pipeline: from image acquisition and code development to precise in situ habitat validation in both intertidal and subtidal environments.