Abstract
Earth Observation (EO) mining systems aim at supporting
efficient access and exploration of large volumes of image
products. In this work, we address the problem of
content-based image retrieval via example-based queries
from Petabyte-scale EO data archives. To this end, we
propose an interactive data mining system that relies on
distributing unsupervised ingestion processes onto virtual
machine instances in elastic, on-demand computing
infrastructures that also support archive-scale content
indexing via a "big data" analytics cluster-computing
framework. In particular, we focus on the analysis of
polarimetric SAR data, for which target decomposition
theorems have proved fundamental in discovering patterns in
data and in characterizing the ground scattering properties.
Experiments are carried out on the publicly available
UAVSAR full polarimetric data archive, whose basic
products amount to about 0.64 PB of storage. We report the
results of the tests performed by using a public IaaS. The
obtained measures appear promising for data mapping and
information retrieval applications.
Anno
2014
Autori IAC
Tipo pubblicazione
Altri Autori
Luigi Mascolo, Marco Quartulli, Pietro Guccione, Giovanni Nico and Igor G. Olaizola
Titolo Volume
Proceedings of the 2014 conference on Big Data from Space (BIDS'14)