If this CCD camera at this secondary thing is doable, then maybe it's worth looking at correction programs. Camera as the one-and-only viewing system.... with a CCD and $200 eBay laptop computer...
A correction program should be able to undo distortions of most types.
Back during the Cold War it was common rumor that the C.I.A. could do near perfect ray tracing on camera lenses.
The spooks could take a photo and the camera that made it. Analysis of the camera let them get a corrected photo as good as the theoretical limit of the film.
So... point a telescope/camera at a known object. The full Moon comes to mind. The image is distorted by systemic telescope/camera errors.
Not to speak of atmospherics, but multiple overlays should cancel out that one. A 50-ply overlay image sounds good.
Then run a correction-determination program.
Compare the overlay thingy with a known "perfect" image. A Hubble moon image would do, most likely.
You get the idea...
The resulting correction mask would function mathematically to adjust for telescope/camera distortion.
A near-perfect optical system ??????
Vane spikes are a different issue. Seems to me that's more like a sampling problem. Preprocessing the image input looks good for spike eradication.
Anybody know beans about this stuff ?
I'm a database "guru"-guy. Not a scientific programmer, image processing supercoder, etc...... ;^)
I looked for "correction" in the CCD forum. And "image correction" with "software" overall. Not much.
