Integration of images taken with different scopes.

[MarcoS]

Good day!
I am trying to find out (without success at the moment) how to integrate two sets of frames of the same area taken with different scopes.
Camera EOS1100D
one set taken with a 90mm FL600 apo triplet
The second taken with Meade LS6" 152mm reduced at FL1280.

is this possible?

Thanks and best regards
Marco

[Carlos Milovic]

StarAlignment should be able to register them without problems. Then use ImageIntegration to average them.
One thing to consider is the trade-off between sharpness and SNR. I would add them together if the seeing and spatial sampling is similar. Otherwise, it might be better just to work with the data captured with the larger scope.
(Another possibility is to sepparate the wavelet layers of them... average medium and large scales, and keep the small scale layers from the set with better resolution.

[Alejandro Tombolini]

(Another possibility is to sepparate the wavelet layers of them... average medium and large scales, and keep the small scale layers from the set with better resolution.

...interesting!

[MarcoS]

Hi Carlos
thanks for the tip, I will try again with star alignment.
Regarding the other possibility you mention  to separate the wavelet layers of them.......
can you indicate were I can find instructions on how to do it?
thanks and clear skies
Marco

[Carlos Milovic]

First of all, you'll have to use the image with the higher resolution as reference. The other should be upsampled/registered to this. Then, to split the wavelet layers, use the MultiscaleLinearTransform. Make a clone of the image, and disable in it the first 2 or 3 layers (you'll have to experiment). This will create the large-scale version of your image. To get the small scale version you'll need to use PixelMath. Use the following equation: " (original-large)/2 + 0.5" and create a new image. This should create an image with a mean value of 0.5 (neutral gray) with all the small scale features. Now, average your large scale version with the low resolution one. Finally, add again the small scales with this equation "averaged + (small-0.5)*2". You should change the name of the images accordingly. Also, you may need to rescale the final image in PixelMath. Play with the number of layers, and you may change too the weight of the small scales (the *2 factor).

[MarcoS]

Hi Carlos,
Thanks for the explanation, I will try tounderstand and see what I can make out of it

clear skies Marco

[lucchett]

What happens if the FOV of the two scope is different?
Is there a way to increase Snr where the two frames overlap and retain the wide field frame for the outer area?
Thank you,
Andrea

[Carlos Milovic]

Yes, ImageIntegration can handle that situation. Nevertheless, if the intersection of the frames get a much better SNR, this could be a problem. You could have an SNR map, and try to use TGVDenoise to smooth that (use that map as Local Support), but in most cases it is just simpler to just crop.

[lucchett]

Thank you Carlos,
What do you mean for Snr map?
How do you produce that?

Thank you!
Andrea