二维自适应排序滤波器及其实时实现的几点考虑

Two Dimensional Adaptive Rank Order Filter and Some Considerations in Real time Implementation

本文首先提出了二维自适应排序滤波器２ＤＡＲＯＦ（ＴｗｏＤｉｍｅｎｓｉｏｎａｌＡｄａｐｔｉｖｅＲａｎｋＯｒ－ｄｅｒＦｉｌｔｅｒ）的结构和算法，然后分析了收敛系数和不同扫描方式在图像处理应用中对收敛和滤波性能的影响。考虑到实时处理的需要，本文还讨论了在实时实现时的几点考虑。最后，给出了２ＤＡＲＯＦ在图像去噪和恢复中的应用实例。

In this paper,the structure and algorithm of the two dimensional adaptive rank order filter(2DAROF)is proposed firstly,then the effects of convergence parameter and some scanning methods to the convergence property and filtering performance are discussed in the case of image processing.Finally,some important problems of real time implementation scheme are studied,and two examples about image noise cancellation and restoration with 2DAROF are illustrated.