From: John Wann Subject: Re: TECH: telerobotic/presence application questions Date: Mon, 2 Sep 1996 12:37:58 +0000 Message-ID: >>And as for vergence... >>after what distance does it become unnecessary to verge the sources >>onto the subject.. pls let me know! Marc Bernatchez suggested: > Stereopsis data become useless around 100 feet. > Depending on your remote sampling resolution, a 100ft may be the case for a machine vision system, but for human perceptual judgments any advantage of binocular viewpoints is negligible after 6m (This is relative to IPD - e.g. if you double the remote camera seperation [without vergence], then you will get useful information out to 12m - but there appears to be quite rapid adaptation to this rescaling so its not a magical-fix). Most pertinant is the consideration of the consequence of verging remote cameras and then presenting the images to a human observer via an HMD/screens. A few previous applications have got the projection geometry badly wrong on this and you need to think the problem through carefully. We have a UK project funded to tackle the remote stereo problem in interfacing a human observer to a remote (high-speed) stereo robot head. John W ********************************************************* Dr John P. Wann < J.P.Wann@rdg.ac.uk> ********************************************************** Action Research Laboratory: Development, Rehabilitation, Immersive Virtual Environments (Wann, Mon-Williams, Smyth , Langaas) ARL - DRIVE Department of Psychology University of Reading PO Box 238, Reading RG6 6AL, UK +44 1734 318532 Fx +44 1734 316604 ***********************************************************