Rapid Digital Monitoring of River Ecosystems

The objective of Fast-DEM is to propose efficient software solutions for monitoring watershed ecosystems, using information captured by multispectral sensors and exploiting high-performance computing (HPC) techniques, in order to promote the digital transformation and automation of watershed ecosystem monitoring, currently a labour-intensive task. The project will move towards automation of this process. Different efficient solutions for the digitisation and automation of processes will be proposed, which can be framed in four major associated domain themes: data pre-processing, spatio-spectral data enhancement, classification for monitoring and, finally, tracking changes over time. Different alternatives will be evaluated to select the most appropriate hardware solutions for each problem in order to produce solutions that are accessible to the end-user and thus have a real impact on the digitisation of this domain. Heterogeneous systems including multi-core CPUs and different types of GPUs and low-end GPUs will be considered.