The aim of the contest is to check existing methodologies at the research or
operational level to solve remote sensing problems using data from different
sensors.
In 2008 a hyperspectral data set was available and the task was to obtain a
land cover and land use map as accurate as possible with respect to ground truth
data.
The Tor Vergata Earth Observation Lab and GeoInformation PhD students team,
led by Giorgio Licciardi, has ranked first, both in the obtained accuracy (0.99)
and in the contribution (0.94) to the fused result.
In 2007 the contest was focused on Synthetic Aperture Radar and optical image
classification for urban area mapping.
The Tor Vergata Earth Observation Lab team, led by the GeoInformation PhD student
Fabio Pacifici, was the winner, able to reach an accuracy of 0.94 in terms of
k-coeff with respect to the unknown ground truth, from sets of satellite radar
and optical images (ERS amplitude and Landsat multi-spectral data).
The contest is sponsored every year by the Data Fusion Technical Committee of
the Geoscience and Remote Sensing Society (GRSS)
of the Institute of Electrical and Electronics Engineers (IEEE)."
|