Speckle suppressing based on fuzzy generalized morphological filter

Abstract

A new filtering scheme using fuzzy generalized morphological operators is proposed for suppressing speckle noise in images. The algorithm employs generalized morphological close-open and open-close operations with a directional structuring element, and acquires the several filtered versions with different directional structure elements respectively, then computes the fuzzy membership of the versions' every pixel according to the designed fuzzy rule. The final filtered image is composed of all the pixels with corresponding maximal membership. Experiment result shows that performance of the proposed scheme is superior to that of lee's filter, F.safa's algorithm and weighted morphological filter. © Springer-Verlag Berlin Heidelberg 2006.