Some thoughts on an optical solution.
The problem of navigation is basically two parts. Where are we and which way are we pointing.
The ability to look at landmarks, planetary positions and relative object size can give us a determination of our position. Unfortunately that is not a very precise measurement and other, maybe simple methods work better. It is probably better to just provide a radio ranging solution based on the communications system or similar equipment.
Attitude, or which way we are pointed is another matter. Taking angular measurement off of stars, planets, moons, or other landmarks is a very accurate method of determining a spacecrafts attitude. Some differential radio measurements can be used, but the wavelengths involved make these sensors much larger than an equivalent optical sensor (camera or less)
The basic concept is to probably take a picture of a star field and identify known stars. measure the position of the stars in the cameras field of view and translate that back into a vehicle attitude. This attitude is then used to update the internal attitude reference.
The most easily identifiable star is the sun. Unfortunately it is not always visible (darkside) and when it is it either blocks out all the other stars or damages the camera. A simple sun sensor will give 2/3 of a vehicles attitude, and also tell when the camera is safe to use. The idea is to protects the camera with a de-polarizing shield akin to an automatic weldor hoods.
So we need a hardware system (camera) and software to determine spacecraft attitude occasionally.
So, sounds like an astronomy problem. Any takers ?