My proposal is for a repository of data and assets for the INDI drivers.
This stuff may be useful for automated stuff or as machine learning models.
this repository should contain various stuff:
star shape models:
airy patterns
deconvolution patterns
aberration patterns
double star models
starlight timeseries
scintillation timeseries
obscuration timeseries
novae timeseries
double variable timeseries
etc.
spectra models:
star class spectra
nebulae spectra
laboratory on-earth references (for redshifts)
crosscorrelation models:
double star models
AGN models
Zernike models
radioastronomy models:
radio spectra
pulsar timeseries
radar timeseries
Some volunteer should put some of this material into the new repository or take them from archive documentation.
Yes,
This is meant for automated observatories or who wants assets or models to compare with their observations.
This way observers can find or recognize objects or take them automatically into account for further investigation.
Spectra are also useful to determine distances or more complicated stuff like star masses or radial velocities.
Some of the code of INDI contains the needed functions, still not perfect but at least automated observatories can do some guess in-loco instead of having a client with the relevant processing delay times or effort.
Including a machine learning core into INDI needs some effort however, but having a ready models set could be a good start!