Skip to Main Content
We propose application of the Morlet reconstruction to eliminate artifacts as they are introduced by standard wavelet denoising algorithms. Our algorithm generates less correlated results than DWT or dyadic wavelet transform based schemes. The quality is measured by the autocorrelation of the reconstructed signal. We find the results better suited for further processing by algorithms that rely on statistical features, as for example in the problem of mass detection in digital mammograms.
Date of Conference: 15-18 April 2004