Multiwavelet Image Denoising Based on VisuShrink Threshold Function

ABSTRACT This paper presents a new algorithm of noise removal from an image corrupted with Gaussian noise based on discrete multiwavelet transform (DMWT) and multisoft threshold values using threshold function proposed by Donoho which is propose the concept of wavelet threshold denoising method (VisuShrink method). DMWTis better than Discrete wavelet transform (DWT) because it has all properties of scaler wavelet transform as well as some features such as short support, orthogonality, symmetry, and higher order of vanishing moments which are a good feature in image and signal processing. Scaler wavelet can’t implement all these features at the same time but multiwavelet can. All results in this paper compared with results got from VisuShrink method. The experimental results show that the denoising results by a new proposed algorithm is better and PSNR larger than the results got from VisuShrink method for the same image. Keywords:Gaussian noise, Denoising, Soft Threshold, DMWT, DWT.