Sentinel-2 Semantic Data Cube Austria

The project is a very relevant step towards the endeavour to build an Austrian data & information cube.  The project will exemplarily show, that semantic content based image and information retrieval (SCBIR) is possible in big EO databases, allowing human users to query and analyse EO data on a higher semantic level (i.e. based on at least basic land cover units and encoded ontologies). This includes:

  • Develop a system and algorithms for systematic and automated semantic enrichment for Sentinel-2 images based on fully automatic pre-classification and a convergence of evidence approach to reach a semantic level up to land cover types that is ready for semantic content based querying.
  • Identify suitable database technologies to develop spatio-temporal modelling and querying techniques in semantically enriched EO databases using encoded ontologies to decrease the complexity of queries for human user interaction.
  • Develop a Web interface for human-like queries based on spatio-temporal relationships and explicitly specify spatio-temporal models for different envisaged user groups
  • Demonstrate the generic potential of the data & information cube in future service development. Up to four service types will be exemplarily outlined based on the generic use cases and shown in a proof-of-concept environment for Austria, e.g.:
    • semantic content-based image retrieval
    • generation of cloud-free mosaics/composites
    • location-based queries through time
    • polygon-based/object-based land cover change analysis through time

Starting in July 2018, the Z_GIS project team will implement together with the project partners Spatial Services, AMA and ZAMG.

Instrument: FFG (ASAP 14)
Role: Coordinator
Project Volume: 323,774 EUR
Contact: Dirk Tiede
Researchers involved: Martin Sudmanns, Hannah Augustin, Stefan Lang