Loading...
 
ESA > Join & Share > EO Resources > Knowledge-centred Earth Observation
Print

Knowledge-centred Earth Observation

KEO Web Page


The Image Information Mining Working Group promotes technology research activities in support of automated and user-centred extraction of information from EO products and image archives. As a step in such a direction, ESA financed the Knowledge-centred Earth Observation (KEO) project. The resulting KEO prototype system aims at expanding the use of Earth Observation (EO) data by:

  • Encouraging the use of a common scientific cooperative environment;
  • Supporting automation of identification and extraction of user relevant information;
  • Providing a large set of EO data processing tools (bridging the gap between data and information);
  • Easing the access to the information extracted from EO data.


The KEO prototype allows users to identify already available processing components suitable for specific applications, create new processing components to extract new information from EO images, chain components into more complex processing chains and run them on internal and external computing nodes.

The KEO system is mainly made of three sub-systems:

  • The Knowledge-based Information Mining (KIM), for interactive exploration of collections of EO images through Probabilistic Information Mining techniques;
  • The Component-based Processing Environment (CPE), for distributed processing and graphic programming via processing components;
  • The KEO Application Operating on Services (KAOS), the user client interface access all system functionalities.

KIM

KIM permits to create and explore collections of images, automatically extracting Primitive Features and reusing them interactively to define user-oriented semantic features.
The system automatically extracts the following Primitive Features:

  • Spectral signature;
  • Texture information;
  • Geometric parameters;
  • Discrete Cosine Transform.

After the interactive training, KIM permits to explore image collections and export the extracted information (features and maps) also for reuse in OGC Web Servers.

CPE

The distributed CPE permits to:

  • Create and semantically identify internal or external Processing Components;
  • Use Processing Components provided by the system or added by the user;
  • Graphically chain Processing Components into more complex Processing Chains;
  • Work with EO data coming from different sources.

KAOS

All KEO system functionalities for EO data processing are provided through a unique Graphical User Interface, the KAOS client application.
Processing Components can be either Software Modules or Feature Extraction Processors (FEPs). The Software Modules, deployed on KEO or user machines, can be:

  • Written in Java;
  • Written in any programming language but wrapped by a Command Line Interface (CLI);
  • Provided as Web Services (WS).

The Software Modules can be executed within the CPE only if embedded into FEPs. A FEP can include one or more Software Modules and / or combinations of other FEPs. The CPE core is a FEP Engine, which activates FEPs via centralised or remote FEP Actuators according to the way in which they were chained using KAOS.

The CPE provides a large number of (currently about 300) Processing Components for:

  • Calibration and Classification of single images
  • Objects / Features Detection from single images
  • Signal Processing
  • Inter-equalisation and co-registration of time series of images
  • Change Detection and Hot Spot Monitoring
  • Basic processing (format conversion, segmentation, etc.)


Contributors to this page:
Michele Iapaolo
livia ranieri .

Page last modified on Monday 28 of April 2014 15:15:16 CEST by .