M.A. Thesis:

Fast and efficient median filter for remove salt- and-peppe rnoise

Digital images today are an inevitable part of everyday people’s life. Some of images have some very important information for us. One of the things that usually distorts the information in the images is the image noise. Digital image denoising is one of the most engaging and practical issues in digital image processing research for researchers One of the most common noises in the field of digital images is the salt and pepper noise. The most famous method of removing salt and pepper in digital image processing books is to use the median filter. In this thesis, a a method based on median filter for removing salt and pepper noise is presented. In the proposed method, in addition to not using all the pixels in the neighboring pixel in the image, the energy function is used to find clean pixels in the central pixel adjacency. The results of the implementation indicate improvement of the proposed method compared to the FEMF method.

Keywords: digital image, salt and pepper noise, median filter, energy function