Immersive Visual Analytic of Earth Observation Data

Kurzfassung

New trends in exploration and visualization are highly demanded in dealing with the massive amount of collected Earth Observation (EO) data. In this work, we propose an immersive visual analytic of EO data utilizing Virtual Reality technology. Precisely, we introduce the visualization of large scale SAR images in an immersive 3D environment and also introduce an interactive learning algorithm for the data clustering using Nonnegative Matrix Factorization framework. We conduct our experiments on a dataset of SAR images represented by different features and show that the proposed interactive clustering outperforms the others.