The closer we get to something, the more complex it seems -- until you can finally break it down to subatomic bits. And even then, it will probably turn out far more complex than we currently believe.

The orbital and surface cameras used for planetary photography need to be calibrated once their journey is completed, whether that be in orbit, or on the planet surface. This is done through imaging a picture with known color values. Minor glitches in the process continue to plague imaging companies, though (i.e., the "known image" may have altered/faded to a fine degree, the lens of the camera may have clouded to a fine degree, the camera was setup wrong in the first place...), and results in a process that has an "acceptable level of error".

People might wonder why the same is not true for orbiting Earth imagers. Well, it is, but the regions being imaged are generally accessible by other means, and the colors (sometimes through the use of average colors from various orbiting sources) aren't so difficult to determine.

Dan