High Density Impulse Noise Removed Depending on Nearest Interpolation and Median Algorithm

Abstract

   In this paper, a Nearest Interpolation and Median (NIM) algorithm is proposed to remove high density salt & pepper noise from digital images. First stage in this algorithm the noisy pixels are detected and in the second stage is calculated the absolute deferent between median and mean value in the kernel, after noise values are eliminated, if the region is  homogenous, the center of the kernel replaced by the nearest interpolation value or  not it replaced by the median value, according to absolute deferent. The proposed algorithm shows significantly better image quality than a simple median filter (SMF), Adapted Mean Filter (AMF), Decision Based Algorithm (DBA) and Decision Based Unsymmetric Trimmed Median Filter (MDBUTMF). The proposed algorithm is tested with different gray scale image and it gives better Peak Signal-to-Noise Ratio (PSNR) and Image Enhancement Factor (IEF).