Advanced InSAR coregistration using point clusters

In this study, we introduce a refined algorithm for the fine InSAR image coregistration which could be used in highly decorrelated scenes. The refinement is introduced at the point of selection of points necessary for the estimation of the offset vectors between master and slave image. A new approach for point selection based on the Harris corner detector algorithm is presented. The new point selection algorithm results with the clusters of point candidates for the offset vectors over a scene. Consequently, the number of points and their spatial distribution are improved, which results in a better global quality of the coregistration model