This is really interesting. Back in the early 1990s I had the opportunity to use a Rolleimetric close-range photogrammetric system in development, through a company in Canada and Grossman & Associates, who at the time was writing the yearly Western hemisphere archaeology summary for the Encyclopedia Britannica yearbook. It was before Windows 95 and it worked inside AutoCad and then stood alone. There were a couple of programs in development. This worked by a series of photos a parabola of an object ideally, with a camera that had been documented with its specific lens for and abnormalities in regard to the reseau marks (small star or cross marks on a glass plate) the film was held against at the time of five fixed-focus settings. The photos were registered on a digitizing tablet and after calculation would produce 3 dimension coordinates traced with a digitizing puck into a file that could be used for drawings, very accurate, if the "error ellipse" for each point calculated was within needed tolerances. For archaeology however, though it was used in the archaeology of an EPA National Priority Superfund cleanup, where you might want as little personal contact with the materials recorded, it has some drawbacks in that it requires correct lighting, color reproduction, etc. For accident investigation, fly over of remote or about to be covered wrecks it, an aircraft in Gander of US military transport, the reason investigated by Canada, could be used from a helicopter with photogrammetric precision, with specific questions to be answered, i.e., did that part of the nuclear reactor move since recorded five years ago and how much, it was very good. It was used for integrating animation and live action in "Starship Troopers" film at first but replaced by laser imaging when some of the photogrammetric recording of the prop cave allowed some of the animated bugs' legs and feet to disappear into the rock. It was also compared to Lidar recording in a test at the Institute of Archaeology at Bryn Mawr, and though comparable results arguably, the Lidar system was much more efficient for recording the test subject an ornamental doorway one might want recorded for HABS/HAER for example. These handheld 3D imaging devices appear to be the best of all possible worlds, having also at Greenhouse Consultants, Inc. back in the mid-1980s, Joel Grossman had the beta of a pointer for early desktop computers, basically two "super" potentiometers like in joysticks that could record 3D of an object, and project for an example the diameter and projected form of a clay vessel from a sherd. Just recently, for cash, Trimble, the GPS providers, bought Rolleimetric of Germany. (
here to MelleniuM or the scanners
http://www.creaform3d.com/en/default.aspx )
