It would require covering the surface with lenses or "pixels" that receive, transmit and reflect light from the object's surroundings.He's trying to make the brute-force-and-ignorance approach work. In concept, this is the obvious way to "cloak" something. And while it should be fairly simple to make work for a fixed (and known) observer position, I think it would be extremely difficult, perhaps impossible, to make work for an arbitrary observer position.
Dave: That's essentially chroma-key video replacement (green-screening); it's completely different from this article.
Cees: Some sacrifices must be made for perfect cloaking
We've discussed this before:
Japan's Invisible Coat