Week 7 Summary : AR supported collaboration
The two papers we read this week had a common theme that they enable collaboration, allowing multiple people to participate on a singular “workbench” though enable a personal view for each viewer. Each paper uses a different technique to enable head-tracking of each user.
SCAPE: Stereoscopic Collaboration In Augmented and Projective Environments.
Multiple users can do collaborative work with people present locally or even remotely. There are two modes, outside-in or the inside-out view. The system has a workbench setup with enables the outside-in view where people could see a detailed view of the setup and collaborate on. It also uses a four-walled mural display surrounding the workbench which enables the inside-out system. The Scape system is an elaborate setup usinga wireless DataGlove, HPMD (head-mounted projective display) and retro-reflective materials on the display, however, the system seems less practical since the retro-reflective firm must be deliberately applied to physical surfaces, which reduces the scalability of the system. I particularly liked the magnifier widget, which is associated with a macroscene at a higher
detail level than the bench view, bascically showing a magnified view of the spot of interest.
A Practical Multi-viewer Tabletop Autostereoscopic Display:
The tabletop auto-steroscopic display featured in this paper is a rather cool take on the problem. The system supports multiple simultaneous users, without the need for stereoscopic glasses, however, the user needs to be tracked. Each person has its own perspectively correct view and can move independently of the other. They utilize the Random Hole Concept which uses a dense pattern of pseudo-random placed holes on top a tradional LCD screen with a good viewing angle. This allows every user to view a subset of the native pixels of the LCD screen behind it. When there is one user, then the display is dedicated to that user and has a good video quality. As number of users being tracked increases, the quality of the video stream diminishes a little. When multiple viewers are present, some pixels are visible to more than one viewer, but due to the non-uniform hole pattern, the RHD is able to diffuse the errors resulting from these conflicts across the image, turning them into pseudo-random white noise. With finger tracking enabled, the possible interactive applications are many where, multiple people can collaborate on the shared workspace with virtual objects, with each having its own perspective.
Comments are closed.