A Simplified Technique Of Design For Double Density Wavelets With Enhanced Denoising APPLICATIONS

Abstract

Denoising Images is a crucial step in noise removal in any data or image communication problem. Recently two image denoising techniques were presented that depends on bivariate analysis. These two techniques were called Double Density Discrete Wavelet Transform (DD DWT) and Double Density Dual Tree Complex Wavelet Transform (DD CWT). They relayed on the decomposition of noisy images with either DD CWT or DD DWT decompositions. This decomposition was after the bivariate based shrinkage technique was applied to exploit the wavelet parent and children correlation for better denoising performance. In this paper we present a novel filter design technique for a DD DWT structure for denoising applications. It composed of 3 dimensional cascaded orthogonal sections. This proposed 3-channel Double Density Wavelet decomposition structure is simple and is fast in terms of convergence. It overcomes most problems for other filter design techniques of being slowly iterative and relays on randomness in parameter optimization. The proposed filter design DD-DWT structure guarantees the Perfect Reconstruction PR and the Alias Cancellation AC conditions and is suitable for VLSI iterative designs. Illustrative examples for the usage of the proposed DD-DWT based design in different image denoising applications are shown in comparison with other denoising structures from recent literature.