Identify the effect of filters on noises

Abstract

Removal of noise from the image is the preleminary step when image pre-processing is done. Many fields such as weather forecasting, 3D graphics in which working on images is done by enhancing the image so that it can be used furthur. Various noises due to temperature, environment conditions, sensors enter into the image. When noise is removed from the image it highlights the particular parts of the image that are required for furthur analysis. In this paper we studied the effect of  three filters (Median, Gaussian, Average) on the four noises ( Gaussian, Salt & Pepper, Speckle, Poisson). This paper gives overview on the effect of fillters by chnaging  detailing on the level of filtering. Performance parameter used is entropy. MATLAB 2016a was used to create the GUI so that effect of filteration is seen on one output screen.

Keywords – gaussian, salt & pepper, speckle, poisson, average, median