On the Use of Complex Flower Pollination Algorithm for Coherence Optimisation In Polarimetric SAR Interferometry

Abstract

This paper presents a new interferometric coherence optimisation approach based on the flower pollination algorithm, called flower pollination interferometric coherence optimisation (FPICO). FPA is a newly developed algorithm based on the pollination process of plants that is used to solve constrained and/or multiobjective optimisation problems. The proposed approach gives us the possibility to control the scattering vectors to be optimised by setting one of its appropriate parameters beforehand. Moreover, the optimised vectors are conforming to the scattering parametrisation, in contrast to previous coherence optimisation methods that are based on eigenvalues problem. Comparisons of FPICO with classical optimisation approaches have shown that the residue rate decreases by more than 70%, which leads to less deformation in the interferometric phase. Furthermore by the enhancement of the phase quality, the results are interesting and provide additional possibilities in contrast to the classical methods such as the optimization of a specific scattering mechanism which may help in scattering decomposition.