A Method of Detecting and Deleting Noise Based on the Maximum Value Iteration of the Wavelet Transform

The singular points in signal often include some important information, whose detection and localization have significance in many practical problems. In this paper, according to the relationship between the singular points in signal and the maximum absolute value of the signal wavelet transform, the noise can be deleted through the maximum absolute value iteration. The effectiveness of deleting noise by using this method is experimentally demonstrated.