possible bug in image registration

[gvanhau]

Hello

There is something strange happening with image registration in latest distribution

During the process I saw at the console output that all but the first image had 0 rotation, and with few exceptions dx and  dy were also 0:

scale       : 1.000
scale-X     : 1.000
scale-Y     : 1.000
rotation    :     +0.00 deg
dx          :     +0.00 px
dy          :     +0.00 px


I Knew that I had a pier flip, so at least 30% of my images sould have rotation angle of 180°

When I inspected the resulting images, it was like they were same as original (no aligment occured).

I repeated this twice, with same results

Regards
Geert

[Juan Conejero]

Hi Geert,

I have just made several tests on Linux and Windows, and StarAlignment works fine.

Meridian flips are no problem at all with SA. However, if you have mirrored images, you have to select triangle similarity instead of polygons as star matching descriptors (triangle similarity is selected by default on the BatchPreprocessing script). Anyway, this cannot be the cause of the problem you are seeing. Can you try with a different data set?

[georg.viehoever]

You should also check which stars StarAlignment actually sees. Use one of the diagnostic mode. I once had the problem that all my images registered to hot pixels, with no shift and rotation...
Georg

[gvanhau]

Ok, I solved the issue.
It´s like Georg said, all the images were registered to hot pixels  
This had never happen before.

Thanks for your help Juan and Georg.

Geert