ELIMINATING HIGH DENSITY SALT AND PEPPER NOISE FROM GRAYSCALE IMAGE USING ALPHA TRIMMED MEAN-MEDIAN FILTER

Abstract

The use of images has increased over the previous decade, and they have the potential to be effective communication tools, similar to social media. In social media, uploading visual information or images seems to be becoming more popular. The appearance of noise disturbs the original information in the image. Thus, removing the noise before using the image for subsequent tasks is necessary. The approaches for image restoration are based on a mathematical model of image deterioration. Alpha trimmed mean median filter (ATMMF) is proposed as a new method for removing salt and pepper noise in digital images. The basic principle behind this method is that it starts with noise detection and then moves on to a filtering strategy. The experimental process was performed with 12 samples of grayscale images with a variable salt and pepper noise density ranging from 10% to 90% to compare the proposed method to other widely used methods. Afterward, PSNR and SSIM were taken as the quality measurements. The proposed filtering technique is simple to use and implement. Experimental results show that the proposed method has successfully reduced salt and pepper noise in high noise density. It outranks all the previous filtering methods regarding visual effects and quantitative measure results.