Sep 5, 2018 · 1 min read
very cool… now all we need is either a) accurate registration of a parallel video camera with your pixel depth values, or b) your OS-1 to output color as well as intensity at every pixel. Then we have a great tool for compositing live action video with CG. I have long dreamt of recovering a true 3D reconstruction of a video scene and being able to model the objects and replace them with CG.