The solar system at 1 kilometer per pixel: Can you identify these worlds?

The image below contains 18 samples of terrain on solid worlds across the solar system. Each square covers the same area: the squares are about 500 kilometers on a side, and when fully enlarged, the pixels are about 1 kilometer square. Before you even try to identify worlds, I'd like you to just step back and appreciate the diversity of terrain; even when things look superficially similar (i.e. lots of craters), there are major differences in the size distribution and shape of those craters. Remind yourself: all of these are the same scale.

Many of you are probably already trying to guess which square represents which world. I'm going to give you all the weekend to try to figure that out. I've included every solid-surfaced world that is (a) large enough to completely fill the square and (b) for which we have imagery at or very close to the target 1 kilometer-per-pixel resolution. (There are two worlds here for which the squares actually show about 430 kilometers on a side because images that have exactly the right compromise between resolution and areal coverage aren't available; I figured that was close enough for the purposes of this comparison.) Each world is on here only once. Almost all the images were taken in visible wavelengths, but there are two worlds for which we have appropriate imagery only in infrared or radar. Sadly, none of the Uranian satellites is present, because all our imagery has a resolution too coarse for inclusion here, except for Miranda, which is too small to be included.

How many of these worlds can you identify? Don't give it away in the comments; let other readers enjoy the mystery. And come back next week for me to reveal their identities to you. I'm not sure how I would do on this challenge -- there are many I could guess right off the bat, but several that I'd have trouble telling apart.