Hi Nils, > So what is happening here is that you convert the R, G and B bands to L > mode each on their own, and then merge the three L-mode images together > into a three-band RGB mode image.. which is indeed a color image pretty > indistinguishable from what you started with.
Thank you for this explanation, this sounds plausible for me. I converted to grayscale because I thought that this would be easier to deal with at the beginning (just need to calculate with one value instead of three and so on) but if each band is processed individually this makes it much easier for me because I don't have to care what mode the image has. Thank you, best regards, Philipp Bender _______________________________________________ Image-SIG maillist - Image-SIG@python.org http://mail.python.org/mailman/listinfo/image-sig
