Plan

End goal

We want to have a podman container with a REST API that does a few things:

1. request images from the acquisition camera
2. process those images to determine a pointing error
3. return a pointing offset
4. This should ultimately be configurable using a text file

Discrete chunks of work to be done

Infrastructure setup (build the container, etc.)

• just need to deal with this, can wait until later

REST API (pretty easy, can wait)

Need endpoints for:

• triggering this whole thing (probably want to be able to specify how many frames to coadd)
• provide (or look up from TCS) star brightness to inform how many images to acquire
• Or maybe just auto-exposure
• calibration
  • ra/dec
  • cass ring angle - complication for later

AutAq python script to actually do this - spend most of our time here

1. request images from the acquisition camera
2. process those images to determine a pointing error
3. send a command to TCS to fix the pointing error

Data simulation

Must simulate

• fpa size (nrows x ncols)
• fpa pixel binning
• different sized "stars" -> simulates different amounts of turbulence
• random motion of star
  • small drift (already done)
  • jitter about a center point
    • specify star center in frame
    • amount of jitter (random deviation from center frame-by-frame)
• how many frames to produce
• stars with different brightness
• blobs as gaussian (https://andykee.github.io/prtools/generated/prtools.gauss.html)

Things that will add reality to the simulation

• optical propagation through turbulent phase screen
• trying to predict realistic radiometry
• "jitter" statistics evolve over time

Data processing algorithm (autaq)

• Python script that ingests some number of images
• Does some level of image processing?
• Coadd
• Computes centroid
• Calculates dRA and dDec values
• So we need a Starlink finder? - should at least do calc to see how quickly one would pass through FoV

Calibration

• Refine calibration plan
• Get design FPA to sky mapping from Jenny or Kent
• Poke the telescope by some known amount
• Calculate resulting centroid motion
• Do this for a couple of cass ring angle positions
  • Make slightly modified version of data simulator script that allows us to generate simulated data with a known FPA to sky sensitivty
  • Use this new simulator to generate simulated calibration data
  • Evaluate our ability to compute the FPA to sky sensitivity