search

AdlerFarHorizons / GONet-Ecosystem

The main hub for GONet tasks and issues.
1 stars 0 forks source link

GONet calibration phase 2 #3

Open zhafen opened 7 months ago

zhafen commented 7 months ago

Due to the fisheye nature of the lens, the pixels near the zenith cover a much smaller area than the pixels near the horizon. This needs to be addressed when converting GONet pixel responses to SQM magnitude. There are two solutions to this:

  1. Assume the angular geometry of the lens is a perfect grid distorted into a fisheye. With this we can calculate the expected area covered per pixel. This is a first approximation.
  2. Calibrate the GONet cameras against a grid, i.e. take images of a perfect grid and use those images to determine the warp of the pixels. This is more effort, but will increase accuracy.

We might close this issue just by doing approach 1, but if so we should list this as one of the assumptions and make a note that this is one of the things we could improve.

zhafen commented 5 months ago

See image library that @jwalsh2me shared (SIRIL.org software).

zhafen commented 3 months ago

Charles is working on building the test hardware for this.