Automatic Detection of Surface Deformations by DInSAR

Abstract

Due to the noise of various sources present in the interferograms and the differential interferograms, the detection of surface deformations by differential interferometry (DInSAR) is thereby a difficult and complicated task. However, the adoption of a visual inspection method for the measured results interpretation can lead to false interpretations, more often in the case of small deformations; that is due to the total dependence to the detection model conceiver. In this context, the work that we present is a contribution to the methods of surface deformation detection by DInSAR, the automatic process that we propose is based on a multi-scale analysis by the discrete wavelet transform. In order to establish this latter, we generated deformation signatures by simulating a surface subsidence and integrating them into pairs of ERS1 / ERS2 sensors images acquired in the region of Ouargla in Algeria.