Design of Digital Image Enhancement System Using Noise Filtering Techniques

Abstract

Noise Reduction in digital image is one of the most important and difficult techniques in image research. The aim of Noise Reduction in digital image is to improve the visual appearance of an image, or to provide a better transform representation for another automated image processing. Many images like medical images, satellite images, aerial images and even real life photographs suffer from viewing, removing blurring and noise, increasing contrast, etc.[34]. Reducing noise from the digital image is a challenge for the researchers. Several approaches are existed in the literature for noise reduction,generally there are common types of noises found in most digital image [35]. This project proposes filtering techniques for the removal of Gaussian,Salt-pepper and Speckle noise from the digital image. MATLAB software is used to simulate and implement the filtering techniques under study. The efficiency of each filter is evaluated based on image metrics PSNR and MSE,MAE and Histogram analysis. Results suggests thatWiener Filter performance is suitable for all Gaussian and Speckle noisy image and the Medianshows better results with Salt & Pepper noisy images.