Shape adaptive discrete wavelet transform for denoising of images

Abstract

The image acquired from a sensor is always degraded by some form of noise. The noise can be measured and eliminated by the process of denoising the image. Recently, Shape Adaptive methods of denoising have gained popularity. The Shape Adaptive Discrete Wavelet Transform (SADWT) transforms and codes the arbitrarily-shaped regions obtained by a segmentation of the image. The arbitrary shapes preserve the edges, articrafts and produce a high quality images. The features of the SADWT’s include the number of pixels in the original visual images is same as the number of coefficients after SADWT’s, the spatial correlation, locality properties of wavelet transforms and self-similarity across sub-bands are maintained well. For a rectangular region, the SADWT is similar to the traditional wavelet transforms. In this paper, the SADWT is evaluated for various images by comparing in terms of peak signal to noise ratio and improves the signal to noise ratio.