Abstract

Earth Observation (EO) mining systems aim at supportingefficient access and exploration of large volumes of imageproducts. In this work, we address the problem ofcontent-based image retrieval via example-based queriesfrom Petabyte-scale EO data archives. To this end, wepropose an interactive data mining system that relies ondistributing unsupervised ingestion processes onto virtualmachine instances in elastic, on-demand computinginfrastructures that also support archive-scale contentindexing via a "big data" analytics cluster-computingframework. In particular, we focus on the analysis ofpolarimetric SAR data, for which target decompositiontheorems have proved fundamental in discovering patterns indata and in characterizing the ground scattering properties.Experiments are carried out on the publicly availableUAVSAR full polarimetric data archive, whose basicproducts amount to about 0.64 PB of storage. We report theresults of the tests performed by using a public IaaS. Theobtained measures appear promising for data mapping andinformation retrieval applications.

Autore Pugliese

• Nico Giovanni

Tutti gli autori

• L. Mascolo; M. Quartulli; P. Guccione; G. Nico; I.G. Olaizola

Titolo volume/Rivista

Non Disponibile

Anno di pubblicazione

2014

ISSN

Non Disponibile

ISBN

978-92-79-43252-1

Numero di citazioni Wos

Nessuna citazione

Ultimo Aggiornamento Citazioni

Non Disponibile

Numero di citazioni Scopus

Non Disponibile

Ultimo Aggiornamento Citazioni

Non Disponibile

Settori ERC

Non Disponibile

Codici ASJC

Non Disponibile