Monitoring the Resilience of Seagrass Meadows from Satellite Meadows from Satellite Imagery Using Machine Learning

Seagrass meadows, found in coastal marine ecosystems worldwide, play a vital role in enhancing coastal biodiversity. This project aims to contribute to the automatic characterisation of the status of \emph{Posidonia oceanica} seagrass meadows in the Balearic Islands, and potentially in other Mediterranean locations, using multispectral satellite imagery. The integration of these images has the potential to improve predictions of the state of health of the seagrass. To this end, supervised neural networks have been trained using pattern images generated directly through theoretical models of seagrasses' growth. Subsequently, the trained neural network will be used to classify pattern types and estimate plant mortality. Throughout the training process, various hyperparameters of the network were optimised.

TFM Supervisors: Manuel Matias and Alex Giménez-Romero

TFM Committee: Tomàs Sintes, Miguel C. Soriano, and Manuel Matias

This online event can be followed online at:

https://us06web.zoom.us/j/87561621346?pwd=h8ewJujLOJ6hpwzWAHLjE9AmXYLj3B.1