OPTICAL OCEAN MAPPING AND BENTHIC CLASSIFICATION
Project reference: TECSPR14-1-0050
IP:: Nuno Ricardo Estrela Gracias
TOTAL BUDGET : 88.077,00€
DURATION: : 01/01/2015 – 31/12/2016
This project addresses the problem of creating and interpreting large maps of the seafloor with special emphasis on computer vision. Two main research lines are proposed: (1) Large scale multi-modal underwater mapping by combining multiple Optical sensors (Standard and multispectral), and (2) the automated characterization and classification of benthic maps, by merging texture features with 3D shape information. A central objective is to improve the existing autonomous mapping technology, such that these tools can be used for habitat mapping and periodic monitoring, and become available for marine researchers and survey companies.