The tricky part in solving the Galex challenge will be selecting stars that will be bright in Galex's bandpass. This isn't too hard for Near UV (NUV), but it's harder for Far UV (FUV).

Here are the first 300 Galex sources. Note the large number of sources near the edge of the field.

 http://trac.astrometry.net/attachment/wiki/GalexChallenge/galex1.png?format=raw

Here's the overlap with an index that we built just by taking bright blue stars.

The NUV overlap is decent, which allows this field to solve quickly and easily (we only need to look at the 20 brightest sources!)

 http://trac.astrometry.net/attachment/wiki/GalexChallenge/galex3.png?format=raw

The FUV overlap is terrible:

 http://trac.astrometry.net/attachment/wiki/GalexChallenge/galex2.png?format=raw

(these plots were produced with ~/stars/RUNS/52/galex_plot.m)

Next, I took the known positions of the Galex FUV sources and searched for stars in the AN catalog that are nearby (within 5 arcsec), and looked at their stats. I also took the first ten thousand stars from that healpix as a background sample.

Bands B and V come from Tycho-2. B covers about 350 to 505 nm, V about 455 to 675 nm.

http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_1.png?format=raw

http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_7.png?format=raw

Bands O and J are the "blue" bands from USNO-B:

http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_4.png?format=raw

http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_6.png?format=raw

Bands E and F are the "red" bands from USNO-B:

http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_2.png?format=raw

http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_3.png?format=raw

Band N is the infrared(?) from USNO-B:

 http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_5.png?format=raw

Here's the USNO-B "red" versus "blue" plots for background vs. FUV stars. (Note that the choice of red and blue dots is purely coincidental and a bit unfortunate).

 http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_16.png?format=raw

Finally, here are histograms for different "epsilon" settings. We seem to get a bit of separation between FUV and background, but the "signal" is small compared to the "noise".

http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_17.png?format=raw	http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_18.png?format=raw
http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_19.png?format=raw	http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_20.png?format=raw
http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_21.png?format=raw	http://trac.astrometry.net/attachment/wiki/GalexChallenge/galexcut_ana_22.png?format=raw