A digital image is often ruined by noise that comes from the process and equipment used to make it or that is added to the image while it is being sent. Different kinds of noise, like impulse noise, uniform noise, Salt-and-Pepper noise, and additive Gaussian noise, can show up in images. Using a good image de-noising method, noise can be taken out of an image to bring back its details. In this study, we propose and test a new way to remove noise from images by using a median filter (MF) in the wavelet domain. When testing the proposed method, different types of wavelet transform filters are used along with the median filter to get better results for the image de-noising process and, in turn, to find the best filter. Wavelet transform is a powerful way to analyze images. It does this by looking at the frequencies of the sub-bands that are split from an image. Based on these tests, the proposed method works better than using either the wavelet transform or the median filter alone. The MSE and PSNR values are used to measure how much an image has been cleaned up. With more and more digital images being made, we need better ways to remove noise from images. No matter how good a camera is, improving the image is often a good idea to make it more useful.
