2131 RIQ FCa : High-pass filtering the coronagraphic images for a better SNR ?

[astronomyk]

See METI-2241, MET-2131

[gotten]

RIX 2131:

Experience showed that, for point source detection, high-pass filtering prior to the application of classical ADI is quite beneficial; in particular to create a SNR map more in line with the Gaussian or t-student distribution of the post-ADI residuals. Is it somethong considered here for the METIS coronagraphic imaging modes ? It may also be beneficial to remove part of the sky background ?

**RIX answer: **

We didn't initially consider this as baseline functionality however an FFT-based high-pass filtering approach with a certain cutoff frequency would seem to be easy to implement (default would be off). Can be discussed with HCI lead and PIP team during PIP FDR.

We try to avoid increasing the number of supported functionality, but this would likely give a good improvement to point source detections and functionality that users might expect to be present.

Action Item 2241:

As a major action item: high-pass filtering will be added as an option to the pipeline. Documentation for this will be updated.

[gotten]

Todo:

Email:

• Talk with Olivier Absil to understand which high-pass filtering technique should be implemented (fourier was not OK)

Text correction in description of algorithms

• add optional high-pass filtering step which takes certain (TBD) parameters

[gotten]

Emailed Olivier Absil about high-pass filtering. He forwarded my email to Faustine and Valentin Christiaens (dev of VIP). Faustine will be in Liege this week so he will debrief the FDR meeting with her. Valentin and Faustine will get back to me.

[gotten]

The author of VIP answered that he had good results with Laplacian or a standard median filter (sliding window). No reply from reviewer yet on this email.

[gotten]

Added a section on high-pass filtering at the end of Section 6.7. Added text to the flows regarding the parameter and algorithm step.

Need to update the figures with an extra High-pass filtering box.
Need to add the corresponding recipe.
Need to ask the reviewer if she is OK with the choices for the high-pass filtering (median and LaPlacian).