i think this is similar to comet processing, maybe, maybe not.
i managed to get 2005YU55 tonight, first with a series of 60s exposures at f/4 and then with a series of 30s exposures at f/4, using my widefield setup (200mm lens + canon 50d).
the 50d being a DSLR, despite being cooled, the individual subs are pretty noisy and when stretched enough to clearly see the asteroid, the background looks terrible.
i'm thinking perhaps to register everything and then integrate to get a nice noise-free version of the background. that leaves trying to pick out just the asteroid from each frame.
since the projection of its' orbit is pretty much linear, perhaps i could use the exposure start times to compute where the trail starts and ends and generate a series of masks that single out the asteroid, then grab that data and paste it back onto the smoothed frame.
or is there an easier way to do this? perhaps somehow utilizing the rejection maps from integration? ...ideas?
