Influence of the Wavelet Family in the Compression-Denoising Technique on Synthetic and Natural Images

A denoising procedure based on a thresholding function has been integrated into the JPEG2000 part II standard which is designed to support a wide variety of applications, including the compression and transmission of medical images. This standard is based on the discrete wavelet transform using the Daubechies(9,7) biorthogonal wavelet, also named the Cohen-Daubechies-Feauveau 9/7 wavelet. However, is this the best choice concerning quality at any time? In this work we compared the results from 36 different variations of wavelets compression schemes to explore their difference related to image content and quality. Two groups of well known images of different levels of complexities and content are used to evaluate the relation among fidelity, image content and noise level. The first group is formed by natural images and the second group is formed by synthetic images. They are considered on two different resolutions (128 × 128 and 256 × 256) and with high level of Additive White Gaussian Noise. The performance of each approach is evaluated by means of fidelity, comparing the originals without noise versus the same images after compression/denoising and decompression. The goal is to define which combination present the best and the worst results considering the final visual quality.