W.Zhou hat geschrieben: ↑29 Jul 2020, 11:49
Many thanks for the tutorial!
Now i am using a duo band filter on my color camera( O-3 and H-alpha)
As you mentioned, the color of stars is a huge problem when i try to stretch the curve (and also when i try to reduce noises with ATWT or MWT). May I ask if there’s a better way to get rid of that in PI? I tried serval times with mask and SCNR (well this process seems incorrect) but there’s still some strange halos around the star .
Many many thanks!
One thing you can do is to split the channels and work on them independently as grayscale images:
1 - Use Channel Management > ChannelExtraction
to split Red and Blue (Green you can ignore or treat the same as green)
2 - Take the blue channel image (which will have the largest halos) and do a Star Mask
(increase Scale to 7 to make sure big stars are kept)
3 - you need to "grow" the star mask or the darker parts of the halo will not be corrected/reduced; the easiest way to do this is to do a HistogramTransformation
and brighten it a bit (or you can try Morphological Dilation but that's probably overkill)
4 - Apply the mask back to the blue image and open the Morphological Operations
5 - Set Operator to Morphological Selection, set selection to ~0.25 and 2 or 3 iterations, apply and see if it's not too strong. If it is, either reduce the iterations, or set selection to ~0.35
6 - Use Channel Management > Channel Combination
to put back the color channels
For SCNR I guess you did:
1 - Invert the image colors
2 - Use SCNR to remove green
3 - Re-Invert the image
The issue might be that if you are using a due band filter, your halo might not be really magenta, but rather "a lot more cyan than necessary" (basically you're not boosting Sulphur). In this case the method above should help as you're fixing the halo before combining the colors.
Hope this helps!