I like the regularization parameters in the Deconvolution tool, it would be even more useful if they could vary as a function of iteration.
For example, suppose I have an image with noise at the smallest scale.
Removing small scale (single pixel size) details is most useful during the first iterations as the image starts out with noise at that scale. However, as deconvolution proceeds to reconstruct smaller details from larger variations in the image, these get removed as well. It would be useful to have the algorithm apply a great deal of suppression of single pixel size details during the first iterations but "less" suppression at later iterations.
My wish: being able to set the regularization parameters so they vary with iteration number.
