Iterative median filtering for restoration of images with impulsive noise

Abstract

Iterative median filtering for restoration of images corrupted by impulsive noise is considered. A modified version of median filtering that can be also applied iteratively is also proposed. The methods are compared with an iterative method suggested by Marvasti for images with high pixel loss. The results show that applying the median filter iteratively results in well sighted resulting images. Moreover, despite the implementation simplicity, the proposed modified median filtering scheme provides a considerably higher convergence speed which intends a lower numerical complexity. © 2003 IEEE.