Detecting generalized salt and pepper noise image based on standard deviation

摘要

The goal of this research is to detect generalized salt and pepper noise in grayscale and color images, using strip window and standard deviation. We compared the performance of algorithms between proposed algorithm (PA), Rank-Ordered Absolute Differences (ROAD), and Rank-Ordered Logarithmic Differences (ROLD) in term of Peak Signal to Noise Ratio (PSNR). We found that PA is better than ROAD and ROLD.