On the statistical similarity of synthetic aperture radar images from COSMO-SKYMED and TerraSAR-X

Singh, Jagmal; Espinoza-Molina, Daniela; Schwarz, Gottfried; Datcu, Mihai

doi:10.1109/igarss.2014.6947549

Cited by 1 publication

(1 citation statement)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our case, we concentrate on an annotation scheme for X-band SAR images with meter scale resolution. Typical instruments delivering such images are TerraSAR-X, COSMO-SkyMed, or airborne SAR images with many similar characteristics [50];…”

Section: Scientific Backgroundmentioning

confidence: 99%

Land Cover Semantic Annotation Derived from High-Resolution SAR Images

Dumitru

Schwarz

Datcu

2016

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

Users of remote sensing images analyzing land cover characteristics are very much interested in classification schemes that define a consistent set of target categories. Up to now, a number of established classification schemes are mainly being used by interpreters of medium resolution optical satellite images focusing on large scale land cover. In contrast, we concentrate in this publication on the definition of a new classification scheme for high resolution synthetic aperture radar (SAR) images that are mostly taken over built-up areas. Here we can see many small details of buildings, industrial facilities, and infrastructure that have to be classified. However, the appearance of details in high resolution SAR images is often difficult to understand for human observers and therefore calls for an automated semantic annotation of the target objects that has to follow a number of specific scientific guidelines. We demonstrate that a selection of representative SAR images with subsequent feature extraction and relevance feedback classification during the generation of a classification scheme leads to a reliable definition of a new high resolution multi-level SAR image classification scheme that can be applied globally for semantic annotation in an automated chain. Index Terms-Annotation; land cover; semantics; SAR I. INTRODUCTION comprehensive in-depth content-oriented analysis of images needs detailed information about their semantics (i.e., the meaning of the local and global image content taken from a thesaurus of semantic labels). When we look at satellite images as we know them from a variety of Earth observation missions, we are able to extract much more knowledge from our image archives once we can query, for instance, the temporal evolution of airport runways within a given country, or the Manuscript

show abstract

Section: Scientific Backgroundmentioning

confidence: 99%