From: uselton@nas.nasa.gov (Samuel P. Uselton) Subject: Re: TECH: GPS in helmet tracking? Date: 03 Apr 1996 16:44:01 -0800 Message-ID: Organization: MRJ at NAS Systems Division From: uselton@nas.nasa.gov (Samuel P. Uselton) In article <315BB22A.4708@hplb.hpl.hp.com> Simon Poole writes: > > > Does anyone know if any helmets are being manufactured that use GPS > > for positioning? Since GPS technology is getting so small and > > accurate you would think someone would incorporate it into a helmet > > tracking system. > > I forget the figures, but GPS will not locate your position to > anything near the resolution needed for traditional VR - i.e. it could > maybe say which room you're in in a building, but it wouldn't tell you > where you are in the room. My understanding is that the military have > much higher resolution systems, but we (the general public) cannot > access them. I don't have the precise figures in front of me either, but I think I can give the orders of magnitude. About a year or two ago, I was at a talk which described, among other things, a tracking system for sailboats that used the GPS. What I recall was that the military can get resolution under 1 meter, but that "for security reasons" the signal available to civilians is INTENTIONALLY distorted, so that accuracy is only about 5 or 10 meters. Years ago when photosensitive CCD array chips were new, a fellow grad student (and some faculty of course) worked on a project using a tracking system that depended on intersecting three planes determined by the edges of shadows cast by a light onto three (1 dimensional) CCD arrays in widely separated places in a room. Some back-of-the-envelope calculations yield a relationship between resolution of the array, size of the working volume, and precision of the positioning. As I recall, the arrays were 256 elements. With a working volume of (say - to make math easy) 2.56 meters in each of three dimensions, the precision of positioning comes out to a cube of uncertainty about 1 cm per side. You need to see three such light sources, in RAPID succession (and with known relative geometry) to get position and orientation for head tracking. Can you figure the range of uncertainty of head angle if two lights are 10 cm apart and each is 1 cm uncertain? It is clearly WAY more than enough to make head tracking an exercise in user discomfort. The position sensing hardware was described in a paper, but I don't think the rest of the system was, because it never worked satisfactorily. However, one of the faculty members, after moving to a different University, was (is?) involved in a system that used (uses?) head mounted CCD array cameras to observe ceiling mounted lights or patterns as a means of head tracking. So, back to the question. GPS is a LONG, LONG way from giving the precision needed for human scale VR tracking. My 2 cents plus a buck...... Sam Uselton uselton@nas.nasa.gov employed by MRJ working for NASA (Ames) speaking for myself