Concepts elaboration and system architectures for mining very large image archives [Elektronische Ressource] / von Inés María Gómez Muñoz

Concepts Elaboration and SystemArchitectures for Mining Very LargeImage Archivesvom Fachbereich Elektrotechnik und Informatik derUniversit¨at Siegenzur Erlangung des akademischen GradesDoktor der Ingenieurwissenschaften(Dr.-Ing.)genehmigte DissertationvonDiplom-Ingenieurin In´es Mar´ıa G´omez Muno˜ z1. Gutachter: Prof. Dr.-Ing. habil. Otmar Loffeld2. Gutachter: Prof. Dr.-Ing. habil. Mihai DatcuVorsitzender: Prof. Dr. rer. nat. Rainer Bruc¨ kTag der mundlic¨ hen Prufung:¨ 08.07.2009gedruckt auf alterungsbest¨andigem holz- und s¨aurefreiem PapierTo my grandparentsAbstractDuring the last decades, satellite technology has been outstandingly improved, providinghugeamountsofEarthObservation(EO)datatobeprocessedandstored. Theavailabilityofveryhighresolutionsensorshasencouragedthebirthofnewdomainsforremotesensingapplications. Relatively new fields in this frame are Image Information Mining (IIM) andContent Based Image Retrieval (CBIR). These fields are born to provide solutions forquerying very large EO archives by content. This dissertation tries to contribute on theIIM domain, providing new image processing algorithms and optimization processes formining image databases.The study of an IIM system can be focused on signal processing methods, data com-pression, semantic knowledge discovery, human-machine interaction or system architec-ture design.