Recent developments in stack filtering and smoothing

Stack smoothers and stack filters have emerged as a useful class of nonlinear filters which overcome many of the limitations of linear filters. At first, the stack smoother structure was limited to have low-pass frequency characteristics. Recently, these structures have been generalized to allow not only low-pass frequency characteristics but also bandpass and high-pass filtering characteristics. Stack smoothers and stack filters compose a very large class of filters. Among these, the most popular and efficient in terms of implementation are the so-called weighted median (WM) filters. It has been shown that these nonlinear filtering structures can effectively address a number of signal and image problems that require frequency selection type of applications. It was also shown that when a recursive structure is incorporated into the WM framework, the resultant filtering structure offers near-perfect stopband characteristics and robustness against noise.