A Markovian approach for urban change detection in multitemporal complex SAR images

In this paper a new technique for change detection in urban areas using SAR data is proposed. Starting from two datasets, the pre event and the post event ones, the technique first computes the relative spatial correlation maps and then compares the similarity between them using a statistical test. The hypothesis is that, if a change occurs between the pre and the post event acquisition, the statistical distribution of the spatial correlation maps changes. The evaluation of the distance between the statistical distribution of the two maps provides an index of changes. The computation of the spatial correlation is carried out using statistical estimation techniques, while the similarity comparison is based on the Kolmogorov-Smirnov test. The algorithm is tested on real COSMO SkyMed data acquired on the area of Naples, showing interesting and promising results.