A modified trimmed median filter technique for noise removal in an image

Abstract

A modified trimmed filtration that is actually median when it comes to recovering the original images that are dishonored by the salt & pepper noise is actually presented here.The refined the pixels tend to be checked for high or least pixel values for example, 0 or 255 in this suggested method. Then this expedient pixel is recognized as the noisy one otherwise maybe not a noisy one in the event that processed pixel is actually 0 or 255. A two-dimensional 3*3 window can end up being selected with this specific noisy pixel as center component and being a running pixel. In this 3*3 window, then this algorithm can replace the value of the loud pixel as mean of these 9 elements in the 3*3 screen if all 9 aspects in the window are values like 0’s or 255’s. Then this modified trimmed median filtration technique can be applied if most of the 9 elements tend to be not 0’s or 255’s. This formula can eliminate the 0’s and present that is 255’s the 3*3 window and change the loud pixel value as average of mean and median of this remaining pixel in 3*3 screen. Simulation results imply that the proposed modified trimmed filter that is actually median can be works really in the event that picture can end up being influenced by salt and pepper sound.The outcome obtained from the recommended modified trimmed filter is compared with the AWMF, DBUTVF, and MDBUTMF. This proposed MTMF filtering technique is actually analyzed on various images for different quality testing parameters like PSNR, MSE.