LinearStarnet

Hi Edoardo, i have tried the script and it works well. is it possible to expose the "starmask" control from StarNet?
 
Hi Edoardo, i have tried the script and it works well. is it possible to expose the "starmask" control from StarNet?
Download the new version from my site.
It has a different file name, so uninstall version 1.1 and install version 1.2.
The starmask is also linear, it is the difference between the original image and the starless one.
 
Fantastic tool!

It would be a pain for me to try and train a new neural network for linear images, so tools like this are more than welcome!
 
Fantastic tool!

It would be a pain for me to try and train a new neural network for linear images, so tools like this are more than welcome!
I'm happy you like it Mikita: you made an awesome job with Starnet and this script is only a drop in an ocean.
I'm still giving the final touch to the script code to cleanup some code specific issues, then I will ask to Juan if we can include it in the standard PixInsight distribution.
 
Looks like the latest PixelMath updates have broken the script? I tried it for the first time today and get errors:

ProcessContainer: Processing view: Rose_clone_ClippingMap

Writing swap files...

5570.506 MiB/s

PixelMath

*** Error: Disassociated subexpressions:

mtf(0.7233512823967073,Rose_clone_ClippingMap-0.030990026949463855)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^

Reading swap files...

4572.189 MiB/s

<* failed *>



ProcessContainer: Processing view: Rose_clone_starless

Writing swap files...

5918.179 MiB/s

PixelMath

*** Error: Disassociated subexpressions:

mtf(0.7233512823967073,Rose_clone_starless-0.030990026949463855)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^

Reading swap files...

4533.994 MiB/s

<* failed *>
 
Looks like the latest PixelMath updates have broken the script? I tried it for the first time today and get errors:

ProcessContainer: Processing view: Rose_clone_ClippingMap

Writing swap files...

5570.506 MiB/s

PixelMath

*** Error: Disassociated subexpressions:

mtf(0.7233512823967073,Rose_clone_ClippingMap-0.030990026949463855)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^

Reading swap files...

4572.189 MiB/s

<* failed *>



ProcessContainer: Processing view: Rose_clone_starless

Writing swap files...

5918.179 MiB/s

PixelMath

*** Error: Disassociated subexpressions:

mtf(0.7233512823967073,Rose_clone_starless-0.030990026949463855)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^

Reading swap files...

4533.994 MiB/s

<* failed *>
Thank you for reporting, I will check tomorrow
 
ColorMask is also broken which would point to the issue being in the new PM module... otherwise it seems every script will need updates?
 
ColorMask is also broken which would point to the issue being in the new PM module... otherwise it seems every script will need updates?
Yes, it seems that now PixelMath requires spaces between operator and operands.

if you add a space before and after the - operator, the expression works.
I don't know if this behavior is intended or not.
I will investigate with Juan.
 
Guys, when I change the Stride and compare images, the image remains the same, can you confirm "Stride" this is working?
 
Bug Fixed (I hope).
Please let me know if it works.
 

Attachments

  • LinearStarnet.zip
    6.8 KB · Views: 207
Back
Top