A new class of nonlinear filters
Qiansheng Cheng, Xiaobo Zhou · 2002
To attenuate random noise is important in recovering signals with change points for signal processing. We give a new class of nonlinear filters. We give left prediction and right prediction, left smoothing and right smoothing. We estimate the current signal value using the smoothing values and predictive values (NCNF). In simulated experiments, it is shown that the presented method is effective.