Image information mining for earth observation at EAS

D'Elia, Sergio

doi:10.1109/igarss.2004.1368972

Cited by 2 publications

(1 citation statement)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance, the German Aerospace Center (DLR) expects to launch along year 2014 a hyperspectral satellite mission, the Environmental Mapping and Analysis Program (En-MAP) [11], wich will generate a huge amount of hyperspectral data. Content Base Image Retrieval (CBIR) systems are relevant to the geosciences because they provide automated tools to explore and understand the contents of large and highly complex images [17,9,7]. There have been several efforts along this decade to develop CBIR tools for remote sensing images.…”

Section: Introductionmentioning

confidence: 99%

Validation of a Hyperspectral Content-Based Information Retrieval (RS-CBIR) System Upon Scarce Data

Veganzones

Graña

2011

Advances in Intelligent and Soft Computing

View full text Add to dashboard Cite

V alidation of Remote Sensing Content-Based Information Retrieval (RS-CBIR) systems requires innovative strategies due to the scarcity of labelled data. CBIR systems validation by means of precision/recall measures based on either user feedback or a-priori known categories, are hard to apply to RS-CBIR systems. We propose to apply a data-driven (unsupervised) quality assessment strategy analogous to the DAMA strategy applied for the validation of classication methods used in thematic mapping. The strategy is intended for quality assessment when little or no ground truth is available. The proposed strategy deals with the RS-CBIR validation problem by giving a quantitative and qualitative evidence of the relative (subjective) quality of RS-CBIR systems without need of a-priori knowledge. We apply the proposed strategy to validate a Hyperspectral CBIR system.

show abstract

Section: Introductionmentioning

confidence: 99%