-
Notifications
You must be signed in to change notification settings - Fork 12
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Plate Solving not returning a correct solution: more files #125
Comments
Thanks a lot for the thorough work @EAlexJ! First of all I notice that most of the time stars in your images are no well detected due to lots of artifacts on the detector or cosmics. To help with that I would recommend using: blocks.detection.PointSourceDetection(n=30, minor_length=5, min_area=10) This will filter out most of the spurious detections. May you try with that and let me know if that helps? I am happy to move forward and have a look at the cases for which this does not help. EDIT: Just so you know, the exact sequence I used is: plate = Sequence(
[
blocks.Trim((50, 50)),
blocks.detection.PointSourceDetection(n=30, minor_length=5, min_area=10),
blocks.catalogs.PlateSolve(debug=True),
]
) As trimming the overscan was necessary to have cleaner stars detection. With that, for the ones in
I'll try to check the ones with no solution but my guess is that the initial RA-DEC guess is far from the solution (could it be?) |
"ValueError: |
After getting to try the improved implementation with quite alot of new files from the same Instrument, I noticed that not all the files are being solved correctly.
For you to try to replicate my results I have provided all the files in question here. All of these files already have the correct WCS solution in their header.
There are three categories of issues I found:
First, the file labeled no_solve throws the following error:
Second, the files I provided in the wrong_solutions folder are ones where the solution returned by twirl is wrong. It has to be noted, that files originating from the same sighting but in different bands do return valid solutions by the looks of it, only these few are for sure not correct.
Third, the last folder no_solution is an entire sighting where the files can be solved by two different solvers, but twirl does not return a valid solution. They are riddled with quite a few artifacts, but it would be nice if twirl and the prose pipeline would be able to handle these as well.
Here is a snippet of how I ran the solver:
I am using prose 3.2.3
Please let me know if there are any further questions!
-EAlexJ
The text was updated successfully, but these errors were encountered: