摘要： 传统滤波器的优点是简单快速和不需图像的先验知识,但其效果不佳。neuro-fuzzy滤波器的优点是滤波效果较好,但滤波时需要一段训练学习时间。针对这点,文章提出了一种新的模糊滤波器,该滤波器除了具有上述两类滤波器优点的同时,又避免了它们各自的不足。实验表明该方法滤波效果优于传统的滤波器和其它模糊滤波器。

关键词: 模糊加权平均滤波器, 脉冲噪声, 区域信息, 图像恢复, 模糊参数

Abstract: Conventional filter has the benefits that it is simple and quick, but its performance is not good. Neuro-fuzzy filters has the benefits that its performance is good, but it needs a lot of training and learning time. In the light of this, a new fuzzy filter is proposed in this paper. The filter has advantages of both kinds of filters besides avoiding shortcomings of them. Experimental result shows that new-filter gives superior performance compared with conventional filters and other fuzzy filtes.

中图分类号: 

• TN911.73