Shape from depth discontinuities under orthographic projection

We present a new method for reconstructing the 3-D surface of an opaque object from the motion of its depth discontinuities, when viewed under orthographic projection as the object undergoes rigid rotation on a turntable. A novel shape completion scheme is introduced to fill in gaps in the recovered surface, which would otherwise be impossible to reconstruct from the motion of the depth discontinuities alone. To verify the basic theory, we construct a large-format orthographic multi-flash camera capable of recovering the depth discontinuities using controlled illumination. We analyze the limitations of multi-flash depth edge detection using orthographic imaging with both point sources and directional illumination. We conclude by considering future applications for the shape completion scheme and the specialized hardware introduced in this work.