Study of Noise Removal Techniques for Digital Images

Abstract

Images very often get contaminated by different types of noise like impulse noise, Gaussian noise, spackle noise etc. due to malfunctioning of camera sensors during acquisition or transmission using the channel. The noise in the channel affects processing of images in various ways. Hence, the image has to be restored by applying filtration process before the high level image processing. In general the restoration techniques for images are based up on the mathematical and the statistical models of image degradation. Denoising and deblurring are used to recover the image from degraded observations. The researchers have proposed verity of linear and non-linear filters for removal of noise from images. The filtering technique has been used to remove noisy pixels, without changing the uncorrupted pixel values. This chapter presents the metrics used for measurement of noise, and the various schemes for removing of noise from the images.