Building edge detection from SAR amplitude and phase data

Building edge detection exploting Synthetic Aperture Radar (SAR) multi-channel data is a research topic of increasing importance in years. The basic idea of the presented algorithm is to estimate edges using jointly both SAR amplitude and phase data. The technique exploits also Markov Random Field (MRF) theory. A correlation map, from which building edges are detected, is computed using both phase and reflectivity map (i.e. using the amplitude and the phase of SAR images). Respect to amplitude based edge detectors and to phase based ones, the proposed method shows better performances in terms of detection accuracy and false alarm rate. The algorithm is tested on simulated data, showing its effectiveness.