A ROBUST ADAPTIVE ESTIMATOR FOR FILTERING NOISE IN IMAGES

In this correspondence, we provide three new methods for restoring ima ges corrupted by additive noise, One is the adaptive mean p-median fil ter for preserving the details of images when restored from additive G aussian noise. Another is the minimum-maximum method for removing outl ier noise, The third method, the robust adaptive mean p-median filter, is based on a combination of the previous two methods, In the past, p roposed restoration methods have generally proven to be inadequate for both detail preservation and noise suppression, but the new adaptive mean p-median filter is shown to be good at both of these tasks, while the robust adaptive mean D-median filter can give good performance ev en in the presence of outliers, Degraded images are processed by the p roposed algorithms, with the results compared with a selection of othe r median-based algorithms that have been proposed in the recent litera ture.