What is it?

The main research and innovation of the proposed project is the bridging of the gap between the raw information that remote sensing images provide and the knowledge gained from the marine application domain to retrieve relevant to the semantic query data.

The main objective of SEO-DWARF is to realize the content-based search of earth observation (EO) images on an application specific basis. The marine application domain and data from Sentinels 1, 2, 3 and ENVISAT will be used. Queries such as “Calculate the rate of increasing chlorophyll in the baltic sea ” will be answered by the SEO-DWARS, helping users to retrieve the appropriate EO images for their specific needs or alert them when a specific phenomenon occurs.


Earth observation satellite images Natural Language for interaction Semantic Organization of contents Specific marine phenomena identification

Intelligent Retrieval and Alerting System

The research comprises the:

  • ontology formalization for the specific research topics,
  • determination of the semantic queries for the application domains,
  • algorithm development for extracting metadata from the EO images,
  • design of an architecture of the platform to perform the semantic image retrieval and storage and management of the extracted metadata.

All four aspects will be integrated in an innovative and user-friendly web based platform enabling the users to retrieve images for marine applications or register for a semantic alert.

What are the challenges?

Technological advances in Remote Sensing have increased the availability of satellite images with different spatio-temporal and spectral characteristics.There is difficulty for retrieving the most appropriate data for each user’s needs. One key challenge is to connect the quantitative information of the EO images with the qualitative (high-level user queries) and be able to mine these connections in big archives. An inherent question arises; how to retrieve EO images based on user semantically aware questions. Content based EO image retrieval techniques have been introduced for bridging the gap between low-level image features and high-level queries. The main constraint of the existing approaches is the generalization of the problem.The formulated ontologies are not focused on the constraints of EO images.

Who are we?

A strong and experienced research team, of 4 academic and 5 industrial partners, coming from Greece, Italy, Germany, France, Cyprus and Bulgaria constitute the project’s consortium.

TWT GmbH Science & Innovation Planetek Italia S.r.l. Laboratory of Remote Sensing,
National Technical University of Athens (NTUA)
Cyprus University of Technology (CUT) I-SEA SAS Department of Marine studies,
University of the Aegean
Planetek Hellas EPE University of Bari “Aldo Moro”
Department of Computer Science
CloudSigma AG