Week 7 Summaries

Tabletop Autostereo

This paper is about tabletop autostereoscopic display and its real-time rendering method. Their system allows viewers to see a different subset of pixels through the random-hole screen. Its advantage is to provide stereo perception without having users wear glasses.

Their work is based on a previous concept named “Random Hole Display”. The authors build the system that can support multi-user collaboration and they imagine the future application of the system. They developed it based on a high density display panel. They print the random-hole mask on a thin polyester film and attach it to the display surface to generate separating layers. As a result, certain area of pixels are visible through the random hole on the mask from a specific viewing position.

The authors later talk about their rendering strategy and some algorithms that can be used to produce accurate color for each view and avoid conflicting pixels. They also come up with the concept of “priority rendering”, which means people in a specific point of view can get better image quality. It needs a lot of efforts to calibrate the system.

The main application of this kind of system is real objects manipulation and teleconferencing. Finally the image quality and image degradation are evaluated by carrying out a series of  experiment.

Question: How to decrease the visual noise caused by the the masks?

 

Scape

This is a paper about a system which can provide a shared space (3D synthetic environment) for multiple users to collaborate on some work, viewing the space from their own perspective. Its core display components are head mounted projective displays (HMPD). The HMPD technology supports collaboration of multiple users by enhancing users’ senses with 3D synthetic information, providing the capacity to create an arbitrary number of perspectives and retaining face-to-face communication among local participants.

The limitation of Scape is the characteristics of retro-reflective materials. A retro-reflective material can only work well for a specific set of angles. Screen shape, screen position and room size are talked about in this paper. Factors that may affect the use of the system such as cross talk, environment light and portability problems are introduced here.

In the section of system implementation, the authors show their work on hardware and software separately. The hardware is combined with a workbench, an immersive cage, multiple HMPDs, interface devices, computing facilities and networking tools. The software is mainly the toolkit named Scape which is a platform controlling over the Scape workspace.

At the end of the paper they discuss how they explore Aztec with their system. They get the 3D model of Aztec from 3D cafe and enhanced it with texture mapping. Each user is provided with a unique ID and a virtual avatar to experience an immersive walk through environment.

Question: Since the system can support multiple users, how do its interface devices deal with the  conflicts of users’ instructions?

 

Comments are closed.