Automatic Detection of Subsidence Troughs in SAR Interferograms Based on Circular Gabor Filters

This letter presents a method of automatic detection of subsidence troughs in the synthetic aperture radar (SAR) interferograms. Construction of such an algorithm is motivated by the very large amount of freely available SAR imagery from Sentinel mission. The proposed method is based on the convolution of the SAR interferogram with a bank of circular wavelets. Spatial locations with high values of the convolved images correspond to centers of troughs. The method has been tested on differential interferograms generated based on Sentinel-1 SAR images. The detection rate varies from 30% to 53% with a relatively low number of false alarms.