An improved coherence optimization method in polarimetric SAR interferometry

Because the important approximation in the optimization algorithm with one projection vector proposed by Colin is not proper in general cases, it is only applicable to process the special data with no temporal decorrelation. To avoid the shortage, an improved approach, which relies on iterative estimation to find higher coherence, is presented in this paper. In order to find higher coherence with same polarization states, the new method takes the results obtained by original one projection vector algorithm as initial estimate. Then it makes use of refreshed two parametric variables to yield new estimation until the coherence can not be improved any more. Finally, the improved method and the original one are applied to L band SIR-C data to demonstrate the better performance.