NavList:

That's a great image.

The bright limb is useless, so this would be a good time to update my tool with support for features internal to the Moon's disk, which is probably better overall anyway.  Already on the image one can see numerous small, high-contrast craters suitable for centroiding (and whose longitude and latitude are presumably known to high accuracy).  The libration is predictable, so that can be unwound from the image, but I'm not current on all the data sources needed for this, nor really the correct chain of coordinate systems from Moon-fixed to ICRF.  (All that's really needed is [craters, time] --> [center of mass].)

Alternatively, the dim limb could be used right away.

About 18 stars are available---that should be fine to recover the stellar background frame.  As always, the JPEG artifacts are unfortunate and will degrade the positions (some of the dimmer stars should be deweighted).  Refraction is probably pretty high (an estimate might be available from the stars, but because the field is only 1 degree or so, it will be very noisy).

Perhaps the day is not far away when such images can be taken handheld aboard ship.  This image appears to track the Moon, but consumer cameras are unlikely to support that, so a series of shorter inertial subexposures will have to do.

Cheers,

Peter