BeHere

General Exam Project
under the direction of
Prof. George Liaropoulos-Legendre

Introduction

The BeHere is an exploration in visually blending together two remote spaces from the perspective of one person in the room. The result is a new perspective and form for the combined space for the observer in his or her local space. The perspective looks correct to a select person of the space. Others in the space view a dynamically changing space that is tied to that one person's movement in the space.

Process

Images
Five walls of a room were photographed using a low distortion wide angle lens. The film was developed into slides which were in turn scanned to produce digital images of the walls.

Model
These images were than texture mapped onto the five walls of a rectangular solid with the same proportions as the original room. A rendering of the model is shown below.

Figure 1. OpenGL rendering of the space looking
at the origin.

Sensing
The location of the select person in the room was determined using a Flock of Birds system. The Flock of Birds receiver was held by the selected observer. From this, the system tracked the users position with respect to the transmitter. This position was then sent to the SGI that was responsible for the rendering. The position obtained with the Flock fluctuated. Some filtering was done to maintain a more steady image.

Rendering
The illusion of the other space is projected onto a flat projection screen that represents the joining plane of the two spaces. The room is rendered using a perspective-view frustrum and is then projected onto the near flat plane. The scene is rendered ccontinuously from each sampled position of the user as they move about the room holding the receiver. The user holding the transmitter sees an accurate rendering of two alligned spaces from their point of view. Others in the space see uncontinuous rectangular volumes.

Figure 2. Three other views of the volume in Figure 1.

Research Issues

Framing
The framing of the space helps create the illusion of a continuous volume. Currently, the right wall of the physical space and the virtual volume are alligned. Framing the projection screen to appear to have this similar wall aids in maintaining this illusion.

Figure 3. Projection screen. The left portion
of the screen is adjacent to a parallel wall
in the physical space. A perpendicular sheet
of foam core is place to the right to represent
a wall perpendicular to the projection screen.

Figure 4. Half of a cardboard frame is placed onto perpendicular
wall extending into the physical space. The purpose of this frame
is to eventually bleed into the virtual space of the extended room.

Bleeding
So far, the two places have been joined, yet there is a clear divide between the two spaces as represented by the projection screen. In an attempt to allow the virtual space to bleed into the physical, an object is placed in the physical space and is represented in both spaces. For example, a picture frame is placed on the wall. Half of it appears on the right wall of the virtual space; the other half would appear on the right wall of the physical space.

Figure 5. Image as seen by the tracked observer.

Figure 6. Image seen from another non-tracked location.

Figure 7. Image as viewed by tracked observer from another location.

Figure 8. Image with tracked user.

Figure 9. Image as viewed from a non-tracked viewer.

Future issues

This project brings forth interesting conceptual issues. To further this idea, it would be interesting to also have virtual objects bleed from the virtual into the real space. In the example above, the frame might begin in the virtual space and flow into the physical space. This can be accomplished through the use of another projector in the physical space projecting objects that can be static or dynamic onto the physical walls.

Similarly artifacts of objects from the real space can bleed into the virtual space. One such example might involve placing a stool in the physical space near the projection screen and seeing its shadow in the virtual space. This would require complex vision, tracking, and modelling systems.

The next step would be to have two physically existing spaces instead of one physical space and one virtual space. People in each of these spaces would each see a different dynamic space before them. How would this affect space usage and interaction? Which forms would be most observed? What forms would be created for the others in the space?

References

Ascension Technology Corporation. The Flock of Birds: Installation and Operation Guide. 1995.

R. Evans The Projective Cast: Architecture and its Three Geometries. The MIT Press. 1995.

M. Kubovy. The Psychology of Perspective and Renaissance Art. Cambridge University Press. 1986.

M. Woo, J,Neider, and T. Davis. OpenGL Programming Guide. Addison Wesley. 1997.