1Department of Electrical Engineering-Guilan University-IRAN
2Amirkabir University of Technology
Multifrequency polarimetric Synthetic Aperture Radar (PolSAR) imagery provides a rich data set of an earth surface’s single scene. These images are degraded by speckle noise as single PolSAR images. Moreover, the redundancies exist between different polarizations and bands. In this paper, Fast Fixed-Point Independent Component Analysis (Fixed-Point ICA) algorithm is used for separating of the speckle noise component from increased number of input PolSAR images. The theoretically analysis and the simulation results confirm that increasing the number of input PolSAR images leads to have a better separation of speckle noise and PolSAR images components using Fixed-Point ICA. Furthermore, this method can be effective for information compression and redundancy elimination. The simulation results and comparisons with the results of optimal weighting filter method results based on using NASA/JPL PolSAR imagery of P, L, and C bands in three polarizations confirms improvement in the separations.