Hi Per,

The "camera" is a time-of-flight imager. It measures the time between the firing of the pulse and the return of individual photons (or photon clusters, I'm not sure which). The YouTube video demonstration says that they have 0.6 mm resolution of the time-of-flight, which implies a 2 ps window of "exposure" (sampling) frame rate.

And before you ask, I have not at all wrapped my little mind around the algorithm that reconstructs the object based on time-of-flight. That would be the fascinating part. Well, that and the whole femto-second laser pulse with a pico-second sampling frame rate.

I guess if you know the distance to the initial wall that you strike, and you know the distance to the second wall you would strike, that gives you some information about when you would expect reflected photons to return. You could then start calculating possible surface locations of any that return at an unexpected time. Any photons that return before you'd expect if they made it as far as the second wall would have to come from something closer than the second wall. And photons that reflected off of the rear surface of the object would return later than those that only hit the two walls. I suppose after multiple pulse/samples you'd start to see an image sharpen up, especially if you're also measuring the position (angle maybe?) of the returning photons in a 2-D field of view.

Anyway, if you haven't looked at the video, give it a look. It's pretty impressive.

Joe