Scale Wavelet Coefficients Research Articles

Efficient wavelet-based image denoising algorithm

Characterising the statistics of wavelet coefficients is a critical issue in image compression and denoising. Many powerful approaches have been investigated, but accurate modelling suffers from high computational complexity. In this work an efficient adaptive algorithm to capture the dependency of both inner and inter scale wavelet coefficients is proposed. Experimental results show that compared with other algorithms, in the case of higher noise variance, greater PSNR performance gain may be obtained.

Fast adaptive wavelet packet image compression

Wavelets are ill-suited to represent oscillatory patterns: rapid variations of intensity can only be described by the small scale wavelet coefficients, which are often quantized to zero, even at high bit rates. Our goal is to provide a fast numerical implementation of the best wavelet packet algorithm in order to demonstrate that an advantage can be gained by constructing a basis adapted to a target image. Emphasis is placed on developing algorithms that are computationally efficient. We developed a new fast two-dimensional (2-D) convolution decimation algorithm with factorized nonseparable 2-D filters. The algorithm is four times faster than a standard convolution-decimation. An extensive evaluation of the algorithm was performed on a large class of textured images. Because of its ability to reproduce textures so well, the wavelet packet coder significantly out performs one of the best wavelet coder on images such as Barbara and fingerprints, both visually and in term of PSNR.

Acuity-matching resolution degradation through wavelet coefficient scaling

A wavelet-based multiresolution image representation method is developed matching human visual system (HVS) spatial acuity within multiple regions of interest (ROIs). ROIs are maintained at high (original) resolution while peripheral areas are gracefully degraded. Variable resolution images are generated by selectively scaling wavelet (detail) coefficients prior to reconstruction. The technique is equivalent to linear interpolation MIP-mapping which involves smooth subsampling (decomposition) prior to texture mapping (reconstruction). Multiple ROI degradation is achieved through wavelet coefficient scaling following Voronoi partitioning of the image plane.

Wavelet based image denoising: VQ - Bayesian technique

An image denoising algorithm for the wavelet-transform domain of an image is presented. In the proposed method, a Bayesian estimation procedure is used to restore the higher scale wavelet coefficients, while vector quantisation is used to realise a spatially adaptive Bayesian estimation procedure for the lower scales. The proposed algorithm provides an improved noise suppression performance as compared to that of the denoising algorithm of Mallat and Hwang (1992) and also that of Malfait and Roose (1997).