AN INNOVATIVE REGION GROWING ALGORITHM BASED ON MINIMUM COST FLOW APPROACH FOR PHASE UNWRAPPING OF FULL-RESOLUTION DIFFERENTIAL INTERFEROGRAMS

摘要

We propose an innovative Phase Unwrapping (PhU) approach to unwrap a sequence of full-resolution multi-temporal differential interferograms through a Region Growing (RG) strategy. The algorithm is designed to handle noisy interferograms to carry out effective phase unwrapping process concurrently of large single-look differential interferograms. The key idea of our algorithm is to exploit a set of well unwrapped pixels (seed points) by applying existing multi-temporal differential technique and propagate the solution to successfully unwrap the neighbouring wrapped pixels (candidate points). To this aim, phase unwrapping of the candidate point network is effectively performed exploiting the Extended Minimum Cost Flow (EMCF) algorithm applied to a Constrained Delaunay Triangulation. The proposed algorithm has been applied to a set of 116 ERS-1/2 and ENVSAT SAR images acquired over the city of Rome (Italy) in 1992-2010. The achieved results demonstrate the effectiveness of the innovative proposed PhU approach.