ESA > Join & Share > Technology Projects > EOLIB Project

EOLIB Project

Earth Observation Librarian

The Earth Observation image Librarian (EOLib) project serves to setup the next-generation of Image Information Mining (IIM) systems, implementing novel techniques for image content exploration. EOLib will exploit information about Earth Observation (EO) product contents which is usually hidden in raster data, image time series and metadata, thus enabling content-based search in very large archives of high resolution EO data. EOLib will be interfaced to and operated in the DLR Multi-Mission Payload Ground Segment (PGS) of the DLR Remote Sensing Data Center, representing at the same time a general new concept for the operations of Ground Segment infrastructures.

Project a Glance

With the increased resolution offered by modern space borne imaging sensors, the amount of available information calls for new mechanisms and processing techniques to handle, understand, and discriminate information. In current high-resolution optical satellite imagery individual objects such as buildings, vehicles, or bridges are depicted separately and oftentimes by more than one image pixel. Classification approaches applied to lower resolution optical multi-spectral imagery, which were traditionally based on pixel-based analysis of multi-spectral signatures, are replaced by object-based analysis methods in high and very high resolution image data. Rule-based context information provides additional input for correctly classifying objects in their environment. Modern SAR imaging systems expose a number of individual objects, which are comparable to the number of objects distinguishable by the human eye, and therefore, a solid comprehension of a high resolution (HR) SAR scene demands that hundreds of object/scene classes be defined and possibly identified.

Objectives and Benefits

Through EOLib, ESA and DLR aim at enlarging the IIM frame for a more complete exploration of EO data sources establishing large scale Information Mining functions within the multi-mission Payload Ground Segments which currently is and will be operated for missions such as TerraSAR-X/TanDEM-X, Sentinel-1/-2, and similar high resolution SAR and optical missions.

The EOLib system will allow users to find EO products of interest for their specific application based on information image content, to prototype EO applications by locally applying Information Mining tools to selected products, and to identify unusual and yet undiscovered patterns in large EO datasets including time series. EOLib aims at developing and implementing a tool for sustainable long term and efficient utilisation of EO data content.
The objectives of the EOLib project are listed below:

  • Analyse the scientific and technical Data and Information Mining achievements for their use in an oper-ational PGS context on well-defined very large EO datasets;
  • Analyse Information Mining use cases for their uniqueness and expected success as opposed to (or in collaboration with) other approaches of accessing large EO data holdings;
  • Extend the IIM concepts to methods for mining heterogeneous EO sources, e.g., EO images, image Time series, their metadata, and other related geo-information;

Develop relevant state-of-the-art and elaborate new Information Mining algorithms up to operational maturity;

  • Elaborate a multi-mission multi-temporal Image Information Mining system architecture ensuring PGS integration;
  • Develop, demonstrate and evaluate EOLib as a system implementing selected IIM and KDD functions for selected use cases on selected datasets.

Application and Result Expected

EOLib is a concept and system (search engine), to be integrated as an organic component within an EO PGS system. It aims at interpreting image content, associate it with other information, create categories, and understand user' inquiries. It will interact with the user to refine his expression of needs, comments and informs users about data content, and suggests the most appropriate products including alternative interpretations. EOLib is a proposal for establishing the next generation EO search engine which adds value in comparison with current systems by integrating “live” products and image and geo-information intelligence.

Contributors to this page: Michele Iapaolo .

Page last modified on Wednesday 23 of April 2014 09:55:05 CEST by Michele Iapaolo.