PixInsight Forum (historical)

PixInsight => Tutorials and Processing Examples => Topic started by: fatboy1271 on 2019 August 15 22:04:52

Title: Range_Mask + CloneStamp?
Post by: fatboy1271 on 2019 August 15 22:04:52
Sorry if this Post is in the wrong section...

I'm attaching an image of a range_mask created with RangeSelection. My question is whether or not I need to CloneStamp out the black dot leftovers in the stars? They range from "big" to pretty small and I don't know how to approach this.

Title: Re: Range_Mask + CloneStamp?
Post by: ngc1535 on 2019 August 16 23:21:21
The "need" depends on how important the mask is in terms of the operation/process you are applying.
The general answer I suspect is yes since whatever you might do- you want it do equally across a star.
There are many way to fix this particular issue... I might apply morphological dilation- or perhaps
better use the morphological selection with a greater than .5 (maybe .75) parameter. This will fill in
the holes with the surrounding white values assuming you choose a large structure element.

Give it a try.

You could probably also use Histogram Transformation and simply lower the white point (right most adjuster) on this mask...
This would probably work as well.

-adam
Title: Re: Range_Mask + CloneStamp?
Post by: fatboy1271 on 2019 August 22 22:50:45
Thanks adam!

My "biggest" issues right now are knowing the importance of any Process I'm doing and then not knowing all of the Batch and Process options. Example, I had not known about applying morphological dilation.

On my RangeMask I tried bumping up the HistogramTransformation for the whites and it seemed to introduce more than I wanted. This could all be because I did something incorrectly earlier in my Project, but I don't know...

fat