Adaptive Order-Statistic Filters for optimal non-Gaussian noise suppression

Abstract

This paper presents an adaptive Order-Statistic Filter (OSF) that maximizes the gain in image SNR. In particular, this distribution-independent non-linear filter approximates the optimal filter when the noise is not Gaussian (e.g., speckle-type clutter, Gamma noise, etc.). Simulation results quantitatively demonstrate the superior performance of the adaptive OSF over popular linear and non-linear alternatives in the presence of non-Gaussian noise.