As Martin has noted this artifact:

-Appears as a dark rim just inside the limb of the planet
-Varies as a function of the aperture of the telescope
-Varies as a function of the wavelength used in the capture.
-Is apparently a diffraction effect caused by the telescope aperture itself.

Below are images of Callisto and Ganymede captured 28 Oct 2012 between 1107 and 1120 UTC. Note that, where you can see detail,
each moon shows a very similar feature,  that feature being a dark ring inside the limb of the moon. This is actually the rim artifact noted by
Martin Lewis.

Details of Capture:
Celestron C14, Televue 2.5 x Powermate. Final FL (measured) 9200 mm. Camera Point Grey Flea3.
Image scale as captured 0.12"/px
The filters used were Astronomiks 2c R, G, B and Astronomiks IR (with a passband >742nm)
Images as shown here have been resampled by a factor of 4 (so the image scale as shown is 0.03"/px)
Captured with Firecapture (2000 frames per filter)
Processed with AVIStack2, best 200 frames used
Registax and Photoshop for further processing
Detail or Artifact?
Under conditions of excellent seeing amateur astronomers have captured very fine detail of Juptier. In those same images one often sees
apparent surface features on the Jovians moons as well. But are those actual features on the moons or artifact?

this rim effect was an artifact of processing, but Martin has very good evidence it is there at capture:
Because the moons are of much smaller diameter than the planets studied by Martin, the distance of the artifact from the limb dominates
complex artifacts in a composite RGB image.

Can one, with amateur equipment, resolve real surface features of the Jovian moons or similar diameter bodies?

Below is an image of Ganymede taken with the Hubble as Ganymede is just appearing from behind the disk of Jupiter. Next to this is a
USGS map of the surface of Ganymede, in Mercator projection but cropped to show the visible face of Ganymede.
You can resolve some identifiable surface features in the Hubble image, but note that (even with the Hubble) the dark areas are of low any
real feature.

Can one eliminate the rim artifact? One thought would be to image the moon repeatedly as it rotates night to night. The rim artifact won't
change for a given filter, aperture and focal length (at least over the short term, while the apparent angular diameter of the moon is fairly
constant). Perhaps the artifact could then be subtracted out.

In any case, to have confidence you are capturing real surface features, not artifact you will have to compare the images to known surface
features on the face of the moon visible at that time.

One can get maps of the surface features of Ganymede and Callisto from the USGS. Search "USGS Map Callisto" (or Ganymede). or by
using the Graphics Ephemerides feature of WINJUPOS.