SAR segmentation algorithms: a quantitative assessment

Abstract

~ This paper compareethe performance of two segmentation tttgaithms, the MUM (Merge Usinghrloments) and RWSEG, using simulatea3yn+h& Aperture Radar (SAR) data. This amparison is performed from b t h the qualitative and quantitative-pohrs o€ *and carried out using synthetic contain regions WTth different homogeneity applications. The process for onsists of creating an idealized which summarizes thhe-mi%n -m e t r i c and topologic charac€er&ks Q€ Amazonian targets.observations from each dmis expesiaceis performed.Then a stati%tkz& pmgosed, and a M