Noise reduction with a multiscale edge representation and perceptual criteria

The wavelet multiscale edge representation of signals developed by S.G. Mallat and S. Zhong (1991) provides a new tool for signal and image processing. A tree structure was built from wavelet-transform (WT) maxima, and metrics for analyzing the WT maxima tree were developed. These metrics fall into two classes corresponding to two perceptual criteria (scaling and spatial stabilities) for discriminating features from background noise. Identified noisy branches are trimmed off the WT maxima tree. This technique of noise reduction preserves edges well while suppressing noise. The results of processing simulated and medical images with this technique are presented