提出一种新的用于图像斑点噪声抑制的滤波算法。此算法应用了形态滤波和模糊逻辑理论,首先以一方向结构元素进行广义形态闭-开、开-闭操作,获得不同方向结构元素的滤波版本,然后根据设计的模糊规则计算所有版本像素的隶属度,以具有相应最大隶属度的像素构成最后的滤波结果。实验结果证明:本文算法优于Lee滤波算法、F.Safa算法和加权形态滤波算法。 A new filtering algorithm for image speckle noise suppression is proposed. In this algorithm, the morphological filtering and fuzzy logic theory are applied. Firstly, the generalized form closed-open, open-close operation of structural elements in one direction is used to obtain the filtered versions of the structural elements in different directions. Then the membership of all versions of pixels is calculated according to the designed fuzzy rules Degrees, the pixels with the corresponding maximum degree of membership form the final filtering result. Experimental results show that the proposed algorithm outperforms the Lee filtering algorithm, the F.Safa algorithm and the weighted morphological filtering algorithm.