一种复合顺序形态滤波百分位值优化方法

An Optimal Approach of the Percentile in Complex Rankedorder Morphological Filtering

本文在最小均方误差（ＭＳＥ）和最小平均绝对误差（ＭＡＥ）准则下，实现了复合顺序形态滤波的百分位值自适应处理。该滤波算法不仅可以有效地抑制信号中的各种噪声，而且较好地保持了信号的几何特征。通过计算机模拟，验证了算法的可行性。

In this paper, the adaptive processing of the percentile in complex rankedorder morphological filtering is implemented under a minimizing criterion such as the mean square error (MSE) or the mean absolute error (MAE). The filtering algorithm can not only efficiently suppress various noise in signals but also preserve the geometrical features of the signals. The invalidity of the algorithm is investigated by computer simulation.