CompuServe Thread

3D scanning leads?

2 messages in this thread
#33345From: Monte DavisJan 25, 1993 3:24 PM
Has anyone here any acquaintance with techniques for acquiring a 3d scan -that is, a DXF or whatever of a 3d object's surface with all its "topography", at least as seen from one direction? I'm interested in, say, source objects that would fit within one cubic foot, at a resolution of up to 100 dpi. I know that reconstructive plastic surgeons have such scanners, but I'm told they start in six figures. How much processing would be needed to do it with a 180-degree "wraparound" set of photographs, using an appropriate light source at a known distance for each? Or take one time exposure from the front as strobes go off behind spaced slits to either side? Or best of all, some kind of small-scale rastering radar — oops, I think I'm creeping back into six figures… I'd be very grateful for any leads to publications, people, or firms involved in this area.
#33624From: John K. JordanJan 27, 1993 10:36 PM
Monte, One image-based method for 3D digitizing is called Phase Measuring Profilometry (PMP). PMP is a non-laser based directed-light method relying on phase difference calculations between several frames of a object imaged with light thru a sinusoidal density grating. For more info you could check out R&D magazine, May 90, or contact Maurice Halioua, director of the Center for Optics, Lasers & Holography at the New York Institute of Technology, Old Westbury, NY 11568. Where I work, we are developing a PMP-based inspection system that will cost well into the dreaded six-figure range. There is no reason, however, that a little ingenuity could not trade some precision and ease of use for a big pile of money. Most of the problem is in delivering the light appropriately and the image caputure (which abounds in Targa Land). The math (software) is not difficult. JKJ