To segment a image with strongly varying object sizes results generally in under-segmentation of small structures or over-segmentation of big ones, which consequences poor classification accuracies. A strategy to produce multiple segmentations of one image and classification with support vector machines (SVM) of this segmentation stack afterwards is shown.

This paper presents preliminary results of change detection studies using TerraSAR-X spaceborne data. Change detection is important for a variety of applications like updating maps (several years following the creation of the base map), emergency assessment following a natural catastrophe, and site monitoring. The first two cases often rely on existing geospatial information (i.e. maps) in the ...

Monostatic pursuit refers to the operating mode formed by two monostatic SAR systems that follow an identical orbit with a separation in time of several seconds. The detected changes between scenes seconds difference are most likely caused ground moving targets. Hence, this opens opportunity detect targets change detection methods. paper investigates possibility using and combine GMTI for GMTI....

Advanced SAR sensors such as the German TerraSAR-X instrument provide a variety of different image products. Typically, the products range from single look complex data to various detected data formats and interferometric products. Among the detected data formats, one can find stripmap and spotlight products derived from single and multi-polarization mode data and with different geometrical pro...