Skip to Main Content
A new filtering scheme using generalized morphological operators and fuzzy logic 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 acquiring the several filtered versions with different directional structuring elements respectively, then computing 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 proves that performance of the proposed scheme is superior to that of lee's filter, F. Safa's algorithm and weighted morphological filter.
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on (Volume:8 )
Date of Conference: 18-21 Aug. 2005