ESA > Join & Share > DAMATS


Introduction and Objectives

The main objective of the Data Mining For Analysis And Exploitation Of Next Generation Of Time Series (DAMATS) project is to study, develop and demonstrate a prototype for the extraction of information in Satellite Image Time Series (SITS). To achieve this, the following activities shall be addressed:
- Quick and effective generation of SITS;
- Definition and categorisation of classes with the same evolution in time (“categories of evolution”);
- Fast semantic searches of defined classes.
- The selection of SITS specific algorithms for the extraction of inherent information content;
- Benchmarking activities for relevant SITS, including additional information sources;
- Implementation and validation of prototypes with relevant user communities, with related services made available over a virtualized environment.

Expected Results

The DAMATS prototype shall contain several modules. The main ones are:
- Interactive, inteligent selection module, for generating SITS in accordance with several criteria: area of interest, time interval, sensor type, maximum precentage of cloud coverage, minimum percentage of a paticular Corine Land Cover class etc.
- Toolbox with instruments for extracting and analyzing information from SITS: change detection maps, classification in terms of "categories of evolution", single class definition through query by example techniques etc.
- Various visualization modes, for viewing original SITS and derived products.
- A set of application scenarios, to prove the usefulness of the prototype; it is expected that scenarios like deforestation, flooding, urban area extension, land use/cover evolution are to be found among selected issues.

Project Schedule

MilestoneDate Place Description


TERRASIGNA, Romania (Prime)
University Politehnica of Bucharest - Research Center for Spatial Information (UPB-CeoSpaceTech), Romania
EOX, Austria

Contributors to this page: Michele Iapaolo .

Page last modified on Monday 09 of April 2018 12:04:00 CEST by Michele Iapaolo.