Noise is always presents in digital images

No abstract

No abstract

Image ac q uisition is a common task in ever y image processing operation. Noise is entered during image ac q uisition from its source and once entered it degrades the image and is difficult to remove. In order to achieve the noise cancellation in an image, non-linear filter works better than linear. This paper presents the joint scheme of Wavelet Transform using iterative noise densit y and Median Filtering to remove Salt and Pepper Noise in Digital Images. The first part of the paper derives the wavelet coefficients with slight increase in noise densit y and in second part these coefficients are further modified b y median filter. The algorithm shows the remarkable improvement over Gaussian noise model and removes most of the nois y part from the image and maintains the visual q ualit y . The level of wavelet decomposition is restricted to three. The renowned indexes Peak Signal to Noise Ratio (PSNR) and Root Mean S q uare Error (RMSE) demonstrate marked improvement of image denoising over Gaussian method.

No abstract