PIMS: Knowledge based Image Information Mining providing new functionalities in the TerraSAR Ground Segment System

Abstract

The paper presents the architecture and new functionalities implemented for the TerraSAR-X ground segment system by the synergy of DIMS (DLR’s Data and Information Management System) and KIM (Knowledge based Image Information Mining system). In the spirit of the KEO (Knowledge Centred EO) concept, this system aims at simplifying the access to, and therefore expanding the use of multi-mission EO data. This achievement comes from the use of emerging technologies for Image Information Mining (IIM), which enables also the selection of images via their content, and for simplified service publication, provision and chaining, as implemented in the ESA Service Support Environment (SSE). PIMS is interfaced with the Service Support Environment SSE on one end and with DIMS on the other. PIMS provides to SSE compliant Web Services to be easily accessed though its operational portal. These services externalize some of the functionalities provided by the PIMS system itself. In this way, PIMS is tailored to develop new services (modules) some of which could be “plugged” in the KEO architecture. Moreover, to be able to process data to get valuable information and services, PIMS will be fed by external data sources represented by either the DLR Ground Segment facility, as for instance TerraSAR-X or a third-party user machine for those end-users interested in using ingestion and processing services offered by the system. Therefore, the system provides a Machine Interface View, which will externalize the provision of services like data ingestion, data processing and catalogue access. Being a KEO component-based system, based on a SOA, the PIMS components are web services, in order to be orchestrated by the KEO workflow manager. Also PIMS takes into account the possibility to be manned by human operators; hence it provides a GUI to be used by end-users for a certain number of PIMS specific tasks.