Multi-scale image de-noising and de-cluttering using cross-scale mathematical morphology
Andrew D. Graham, Martin J. L. Turner, Terry Hewitt · Research Explorer (The University of Manchester) · 2007
We present an interactive method for de-noising and decluttering images based on applying morphological operations to the hierarchical tree formed from thresholded coefficients of the over-sampled dyadic wavelet transform of the image. We compare the method to simple and adaptive coefficient thresholding techniques and find that it denoises and de-clutters more completely than these other techniques.