Data Mining and Knowledge Discovery for the TerraSAR-X Payload Ground Segment

Abstract

Earth Observation (EO) imaging satellites continuously acquire huge volumes of high resolution scenes thus increasing the size of image archives and the variety and complexity of EO image content. Therefore, new methodologies and tools that allow the end-user to access large image repositories, to dynamically find and retrieve collections of desired images, and to extract and infer knowledge about the patterns hidden in the image archives are required. In this context, this paper presents the Earth Observation Image Librarian (EOLib), which is a modular system offering data mining and knowledge discovery functionality for the TerraSAR-X Payload Ground Segment and is serving to setup the next generation of Image Information Mining (IIM) systems. It implements novel techniques for image content exploration and exploitation. The main goal of EOLib is to create a communication channel between Payload Ground Segments and the end-user who receives the image content enriched with annotations and metadata as well as coded data in an understandable format associated with semantic categories being ready for immediate exploitation. EOLib is composed of several components offering new functionality such as ingestion and feature extraction from SAR images, metadata extraction, semantic definition of the image content based on machine learning and data mining methods, advanced querying of the image archives utilizing data content, metadata and semantic categories, as well as 3D visualization of the huge and complex image archives. EOLib will be interfaced to and operated in DLR’s Multi-Mission Payload Ground Segment (PGS) of the Remote Sensing Data Center at Oberpfaffenhofen, representing at the same time a general new concept for the operations of Ground Segment infrastructures.