The objective is to explore effective virtual worlds that immerse the visitor into the VR world. The visitor is immersed when the can interact only with the virtual reality and not distracted by the real world. When successful the VR experience has the potential to teach, inform, and train the visitor.

Part of feeling presence is “ease of interaction” with the virtual environment, such as navigating the environment. Navigating in the natural world is by walking about, which is also possible in virtual environments. However, some virtual spaces, such as university campus, are larger than the available lab space. 

As there is not enough room to walk everywhere, one solution is to “fly” through the environment. Similar to a video game, the visitor uses a joystick to move and turn through the environment. Although this is possible, the visitor’s VR experience is diminished because vital cues are missing to immerse the person. Unlike the joystick, the person can have a better perception of how far they walked. I also believe walking allows the visitor to build a more complete understanding of the environment (still to investigate).

The solution for the person to walk in a large environment is to use redirected walking, which is investigated by Sharif. Redirected walking gives the visitor the allusion of walking in a straight line for long distances. The path in reality is back and forth between two points. While the visitor is stopped, distracted by looking around, the virtual world is turned. When the person continues to walk it will be in a different direction.

I hope to demonstrate redirected walking in a virtual world of the UNC campus. The visitor will walk, for instance from Sitterson to Davis Library. Along the way they have to stop and identify the surrounding buildings and unknowingly will be physically rotated. 

The redirected walking experiment also explores the question “how much reality is needed to have presence?” The campus environment is rendered from limited information about the environment. Only the foundations of the buildings will be known. The walls of the buildings come from stretching the foundations upward. Details used to identify the building will be missing, such as total height, windows, doors, stairs, building material. Very generic buildings might not be enough for the person to easily identify the buildings. 

Original suggestion

Campus environment

The virtual world consists of the campus buildings. A SVG file stores information about each buildings’ foundation and name. The units can be scaled to be life size. The SVG file format is in XML and is read by the program with the Apache XML parser. 

The campus buildings are composed of walls that are rectangles. All buildings have the same height. Each building has a random color which helps distinguish them. 

The buildings are easier to recognize with labels that appear in front of the buildings. Although it is not realism it will greatly improve building recognition. In OpenGL the labels are created with text bitmaps. 

Navigation

The simplest navigation to implement is joystick mode. In this mode, the entire virtual world can be explored, such as from 5 miles above or on the ground. It is interesting to compare the different perspectives. At 5 miles away, the campus looks like a map and it is easy to recognize campus. On the ground it is much more difficult to get around. 

Redirected walking will be the next form of navigation. This will require placing markers, which are the points the visitor has to stop and be turned around. At the stopped point, they have to be distracted and look about.

I have a pointer to Sharif redirected walking, so when I have time, I can experiment with it.