More about this network
A pressing requirement, at the time this network was launched, was for the integration of the various computing methodologies such as pattern recognition, information retrieval, and data analysis, into a common framework for analysing the ever increasing amounts of data generated by space-borne missions, ground based observatories, and large wide-field surveys.
Hence, this Network aimed to help achieve this by coordinating the transfer of relevant computing expertise between different European laboratories and departments. Limited linkages between centres of expertise already existed, and the Network aimed to build on these and provide an ongoing commitment to the field not available at the time from any other European organisation involved in either astronomy or associated computing methodologies.
Among computing problems for modern astronomical research at the time were:
pattern recognition, needed, for example, in automated galaxy counts where it is necessary to distinguish galaxies from other sources such as individual stars in our own galaxy; new techniques for retrieving images from databases quickly and easily; ability to cross-correlate between large catalogues of information, both for scientific analysis and as a basis for missions (e.g. the Hipparcos Input Catalog, INCA); ability to integrate data associated with different wavelengths for a particular part of the sky to build a complete picture; ways of classifying and storing the increasing volumes of data collected, which already run into many terabytes.
This Network aimed to concentrate on three central topics to consolidate the various techniques that solve these problems.
From Vision Models to Image Information Retrieval
Existing methods such as fuzzy logic, mathematical morphology, and multi-resolution approaches, need to be moulded together to develop new ways of accessing information from large image databases. Methods of identifying images by specifying their content, or their type, need to be found.
The Data Life-Cycle - Methodological Aspects
The aim here was to consolidate standards that help address the various stages of the astronomical data life cycle. Increasingly data is captured digitally on CCD electronic detectors, it is subjected to image processing and statistical analysis, and finally it is either archived, or published, or both. The final stage of electronic publishing or archiving in digital libraries is increasingly important in astronomy. To support these various stages of the life cycle, the Network helped coordinate the different data storage, access and interface standards.
From Data-Integration to Information Integration
Astronomy involves observations made at different wavelengths, including visible light, infra red, radio waves, X rays, and gamma rays. Integrating data generated by observations at different wavelengths to provide a composite understanding of objects such as galaxies, spanning the whole electromagnetic spectrum, is particularly important for large space and ground based observation projects. Such projects to integrate data at different wavelengths have been undertaken by NASA, by ESA, and by observatories and research organisations, in recent years, and these have identified issues and concerns that can be solved with improved methods.
Within this section, the Network also looked at new ways of classifying large data collections, and deciding which data is worth storing in the very long term, as it is increasingly impossible to keep everything.
