Abstract
The organised storage of spectral data described by metadata is important for long-term use and data sharing with other scientists. Metadata describing the sampling
environment, geometry and measurement process serves to evaluate the suitability of existing data sets for new applications. There is a need for spectral databases that serve as repositories for spectral field campaign and reference signatures, including appropriate metadata parameters. Such systems must be (a) highly automated in order to encourage users entering their spectral data collections and (b) provide flexible data retrieval mechanisms based on subspace projections in metadata spaces.
The recently redesigned SPECCHIO system stores spectral and metadata in a relational database based on a non-redundant data model and offers efficient data import, automated metadata generation, editing and retrieval via a Java application.
RSL is disseminating the database and software to the remote sensing community in order to foster the use and further development of spectral databases.
Hueni, A