This paper presents a class of adaptive ranked-order morphological filters. The implicit formula of the order operation is expressed. The adaptive processing of structuring element and percentile is implemented based on a minimizing criterion such as the mean square error (MSE) or the mean absolute error (MAE). The filters can not only efficiently suppressed various noise in signals but also preserved the geometrical features of signals. The results of computer simulation show that the new filtering method is quite effective in restoration to noisy images.