[INSTRM-629] Refine centroiding/spot detection algorithms based on commissioning run data Created: 14/Mar/19  Updated: 17/Dec/21  Resolved: 17/Dec/21

Status: Done
Project: Instrument control development
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Normal
Reporter: karr Assignee: karr
Resolution: Done Votes: 0
Labels: EngRun
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified


 Description   

Based on the data from the commissioning run the centroiding and spot detection algorithms need to be updated for robustness and accuracy with strongly non-Gaussian PSFs and variation of the PSF shape over the field.

Using SeXtractor + PSFEx with PSF fitting provides more than sufficient accuracy and robustness, but is much too slow for on the fly analysis. Therefore, I am adapting SeXtractor methods to find a balance between accuracy/robustness and speed. 



 Comments   
Comment by hassan [ 14/Mar/19 ]

Moved from INFRA to the INSTRM project.

Comment by naoyuki.tamura [ 29/Mar/19 ]

Just for a record Jeniffer circulated a report of updates in the ongoing works by e-mail PFS-ics: 01631.

Now I seem to remember (and perhaps I should more formally have put this heads-up sooner) that Robert commented on PFS slack that there is a code available that was developed in the LSST project. If that does the same jobs faster, it should be worth looking at it in parallel.

rhl could you comment again and give information here?

 

Comment by karr [ 02/Apr/19 ]

The code has been translated into C and integrated into the cython/MHS framework and the commissioning python notebooks. The timing values are within required parameters. 

Next I'll tidy up the code a bit and add comments, work out the robust method for automatically determining input threshold/spot size values, and run it through some more tests for robustness. 

Comment by karr [ 17/Dec/21 ]

Finished some time ago, so will close.

Generated at Sat Feb 10 16:27:01 JST 2024 using Jira 8.3.4#803005-sha1:1f96e09b3c60279a408a2ae47be3c745f571388b.