Image Denoising Using Adaptive Bivariate-Bayesian Threshold in Multi Wavelet Packet Transform

Abstract

 The basic aim of Image Denoising is to reduce the noise and modify the important features present in the corrupt signal. Researchers have been doing it in one way or the other since long time. In this paper, we present a new adaptive Bivariate- Bayesian soft threshold denoising scheme using multi wavelet packets. Bayes estimation and Bivariate wavelet soft thresholding overcome the shortcomingsof each other. Our algorithm is capable of dealing with highly contaminated images. Simulation results show that Bivariate-Bayesian soft threshold has better perform over existing adaptive threshold denoising techniques viz. Bayesshrinkage, Maximum A Posterior (MAP) shrinkage, Ogawa shrinkage, etc.The quality parameters likePeak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), L2 Norm and Structure Similarity (SSIM) Map were used for the comparison of the performance.