Testing an OpNav Rig
An appropriate function that inputs the base-line, focal length, resolution & depth will help define the "hardware rig". You will still require test footage to be used for development of relevant control algorithms. Destructive testing of this rig is kind "easy" though prohibitively expensive.
However all we really need is video footage, at a reasonably high resolution, calibrated for accurate height from the "surface" - these when put together can simulate a free/controlled fall onto the lunar surface without actually destroying the test rig itself. So, the keywords are absolutely no damage to the test rig, Controlled fall/ Movement, Video footage, lunar surface simulations.
To make things easier, its easy to rotate the fall axis by 90 deg and make it horizontal, therefore instead of falling, the rig moves closer to the wall, which is pretending to be the floor :)
Add a second arm with lights on it to simulate the different lighting conditions, add as many degrees of freedom of movement to the rig itself and we are ready to non-destructively yet reasonably accurately test the rig.
This is software testing using in loop hardware and the device you are looking for to do the task mentioned above is called Test-Bed for Robotic Optical Navigation, TRON for short, great stuff @DLR - http://www.dlr.de/irs/en/desktopdefault.aspx/tabid-6883/