A generalized phase unwrapping approach for sparse data

Phase unwrapping is usually considered for data defined on a rectangular grid. In this article, the case in which data are available on a sparse set of points is considered. The authors formulate a generalized unwrapping approach to deal with sparse data. The method can require the use of an approximate model of the differences of the unwrapped phase function to be reconstructed. Given a parametric model of the unwrapped phase differences, a technique to determine the parameter values that make the model fit "best" with the sparse wrapped phase data is also proposed. Tests performed on SAR differential interferometric data confirm the validity of the approach.