On Saturday, May 7, 2016 7:05:16 PM CEST johannes hanika wrote: > ooh, nice! papers to implement :) > > one that's been on my list for quite a while but i never got around to > do it is this: > > http://people.csail.mit.edu/sparis/#cacm15 > http://people.csail.mit.edu/sparis/publi/2015/cacm/Paris_15_Local_Laplacian_ > Filters.pdf > > which can separate coarse/fine detail without introducing halo effects > a little bit better than what we currently have in, say, the > shadows/highlights module. this is a building block that may replace > the bilateral filter in a couple of places.
Thanks for the nice welcome and the reference. I read the paper and it sounds interesting to me. Under [1], Matlab source code of the filter is provided. I quote its licence at the end of this mail. As it is not exactly GPLV3, I kindly ask someone to check if I am allowed to read that code. I would offer to summarize the main points of the paper to bring everyone (including me) on the same page. We than can discuss details. Does this work for you? I am looking forward to work an this filter. Cheers, Wolfgang Licence of the Matlab implementation of the Local Laplacian Filters "Start quote" Copyright (c) 2011 Sam Hasinoff Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. "End quote" [1] http://people.csail.mit.edu/sparis/publi/2011/siggraph/ matlab_source_code.zip > > re:denoising. there is a little bit to be gained if we make use of > negative values, passed on from the input with black point subtraction > up to the denoiser. i'm not sure how well this would work with the > Anscombe transform we currently do for noise stabilisation (initial > results were only quite marginally better), we may want to add a Fisz > transform wavelet mode to the profiled denoiser (i have some code for > this in a branch). > > let us know what you fancy, then we can discuss details.. > > cheers, > jo > > On Sat, May 7, 2016 at 9:03 AM, Moritz Mœller <virtualr...@gmail.com> wrote: > > What is missing is porting one of the contemporary black and white > > conversion plugins to be available in DT. > > > > The monochromes module is just very ... traditional. > > > > Some of the recent papers have C code/Gimp plugins. One of them should be > > ported to DT. I nan send your some suggestions. > > > > .mm > > > > > > > > __________________________________________________________________________ > > _ > > darktable developer mailing list > > to unsubscribe send a mail to > > darktable-dev+unsubscr...@lists.darktable.org > > ___________________________________________________________________________ > darktable developer mailing list > to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
signature.asc
Description: This is a digitally signed message part.
