Hi,
as I recently mentioned I am working on an automatic haze removal module
for darktable. Adopting the useless module I added a new module to the
correction group and placed this new module into an appropriate place in
the pixel pipe. The only missing part is the actual implementation of
the process method. Here, I got stuck with the following problem:
A central step in automatic haze removal is to estimate the color and
intensity of the global diffuse background light. This is done by
estimating the amount of haze in each pixel by investigating the local
neighborhood of each pixel. Then the 2%, let's say, brightest pixels
among the 2% of the most hazy pixels of the whole image yield the
global diffuse background light. The problem now is that, only some
region of interest is accessible in the process function, but I need
access to the whole (possibly scaled) image data. Thus, my question is
how can I extract global image features in the pixelpipe?
If I would determine the global diffuse background light just on the
basis of the region of interest the result of the haze removal process
would depend on the chosen image section in the darkroom preview, which
is very undesirable.
Heiko
--
-- Number Crunch Blog @ https://www.numbercrunch.de
-- Cluster Computing @ http://www.clustercomputing.de
-- Professional @ https://www.mpi-hd.mpg.de/personalhomes/bauke
-- Social Networking @ https://www.researchgate.net/profile/Heiko_Bauke
___________________________________________________________________________
darktable developer mailing list
to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
