Abstract
Arctic coastal regions are strongly endangered by changing climatic conditions and anthropo-genic utilization, for example due to permafrost degradation, or near-shore oil drilling. Since SAR systems are operating independently from external illumination or cloud shadowing, they are suitable for continuous area-wide monitoring of arctic environments over the whole year. In this work we show the application of dual co-polarized (HH/VV) TSX data for unsupervised wa-ter and land cover classification of arctic tundra environments by means of Kennaugh Matrix, Two-Component Dualpol and Alternative Entropy Alpha Dualpol Decomposition. Classification of both water and land is derived by means of iterative unsupervised classifiers. Overall accura-cy for water is approx. 94 \% and 74\% for examined basic land cover types.
Users
Please
log in to take part in the discussion (add own reviews or comments).