The Earth Observation Image Librarian (EOLIB): The Data Mining Component of the TerraSAR-X Payload Ground Segment
Daniela Espinoza-Molina, Vlad Manilici, Corneliu Octavian Dumitru, Christoph Reck, Shiyong Cui, Henry Rotzoll, Mathias Hofmann, Gottfried Schwarz, Mihai P. Datcu · elib (German Aerospace Center) · 2016
In this paper we present the Earth Observation Image Librarian (called EOLib) as a new generation of Image Information Mining Systems. EOLib is operated in the Payload Ground Segment of TerraSAR-X. The main goal of EOLib is to provide semantic annotations of satellite image content and offer to the end user a semantic catalogue via a web user interface. Moreover, EOLib has more functionality such as searches based on image metadata and semantics, visual exploration of the image archives, metadata extraction, etc. The system consists of components such as a query engine, knowledge discovery in databases, visual data mining, epitome generation, and user services. EOLib is able to ingest a TerraSAR-X scene with 8000×8000 pixels in about three minutes. The EOLib workflow starts with the ingestion of a scene, it continues with the semantic annotation of the image content based on machine learning methods, and it ends with publishing the semantic catalogue and enabling the search by metadata and semantic image descriptions.