Nonlinear filters for noise reduction in image signals

Abstract

In this study, we investigate the performance of several nonlinear filters for noise reduction in image signals. We find that the total variation (TV) regularization filter exhibits the best performance for effective noise reduction and picture quality preservation. We analyze the behavior of the TV regularization filter and elucidate how noise is reduced effectively without losing edge sharpness or picture minuteness.