Methods and Algorithms for Understanding of High Resolution SAR Images: Towards a SARINT Concept

Abstract

Synthetic Aperture Radar (SAR) imagery, in the last two decades, has become increasingly popular as some of its properties are favorable and complementary to optical imagery. With the increase of the SAR sensor performance up to sub-meter resolution, a more detailed analysis and a finer description of SAR images over scenes, mainly object of human activities areas, are needed. The high diversity of man-made structures combined with the complexity of the scattering processes makes the analysis and information extraction, from high resolution SAR images over such areas, non-trivial. Since SAR provides N-dimensional complex valued signals, i.e. the information on the scene or target is modulating in amplitude, frequency or phase the radar echoes, the SAR data interpretation requires very specific techniques, to adapt to its “non-visual” nature. In addition the “sub-meter” resolution makes possible the detailed observation of man-made structures, thus the importance of SAR information is growing dramatically. In this frame is needed to develop expertise and tools for semi-automatic or automatic support to SAR data understanding and SAR Intelligence (SARINT). The tutorial aims at presenting advanced methods for SAR image understanding based on interactive and automatic algorithms. The methods are intended for the analysis of the most used products: detected and single look complex images, and supported by coherent multiple observations in InSAR modes. The inclusion of model knowledge obtained from collections of pre-recorded physical target data leads to a comparison of the acquired data with representative models. Similarities and deviations revealed during the comparison allow a detailed high resolution interpretation of the image data and lead to a full image understanding. The lecture will give practical examples using TerraSAR-X images.