Three-Dimensional Interactive Environments Using VRML


Nothing has captured the imagination of the world like the Internet. Information and communication are more direct, interactive and informative. And nothing has evolved on the Web as dramatically as VRML, the Virtual Reality Modeling Language. Since its inception in 1994, VRML has been encouraging an open forum development. This has created a kind of grass roots basis of support and interest. VRML supplies an immersive technology to the Web and gives subscribers the ability to explore information never experienced prior to the Internet.

VRML allows the developer to describe 3D objects, combining them into scenes and environments. Now, 3D worlds may be mapped and explored using the World Wide Web. Users may create simulations incorporating animation, sound, motion physics and multi-user participation.


Motion Capture (MoCap) is the process of capturing the movement of an actor and applying it onto a computer generated object. Usually, MoCap is used to create synthetic actors by capturing the motions of real humans. In this case, special markers are placed over the joints of actors. Then, a special hardware samples the position and/or orientation of those markers in time, generating a set of motion data, also known as motion curves.

MoCap has been studied since the 1980's. It has been used for medical purposes, as in the study of physiological movement. MoCap also has been very successful with design and engineering applications. An example is the use by jet engine manufacturers to use the immersive technology to study the performance of jet engines. Also, this technique has been used by special effects companies to produce incredible, realistic animations in movies such as Star Wars I and II, Titanic, Starship Troopers, Species, Batman, Terminator 2, and others. The great advantage of MoCap over traditional animation techniques such as keyframing and simulation is the ability to generate natural-looking animation.


The goal of this project is to study VRML 2.0 and create a 3 dimensional environment incorporating a couple of characters acting out a scene. The team will meet and design their environment, describe a story about the actors and the environment. The choreography has to be well-defined, most important of which are timing, relationship of the actors to other actors, etc. The project group will select one or two team members to be the actors during the mocap session. On Thursday of the first week the team will meet in ACCAD's (The Advanced Computing Center for the Arts and Design) Motion Capture Lab. The team will be given a short orientation on mocap. Then, members of the team will be assigned responsibilities for the session. These assignments will include director, timer, and actors.

After the mocap session, the team has little time to build the environment in which the virtual actors will perform. So it will be "shoulder to the mouse" and "nose to the keyboard" time for the remainder of SI03.


The modeling language is VRML 2.0. We will be using the text, The VRML 2.0 Handbook, by Jed Hartman and Josie Wernecke. We will delve into 3D computer graphics and understand the concepts behind 3D modeling and surface treatment. VRML 2.0 will be taught like the other programming languages such as FORTRAN and C. As with other programming languages, structure, scope and hierarchy are important to VRML 2.0. One difference is that we don't have to compile. We will be viewing the models through a choice of network browsers.

The models will be developed with a "hands on" approach, as in "we're not going to use any interface". The most useful tool will be the word editor. (My editor of choice is emacs.)

Surface treatment will involve discussing polygon, vertex and texture maps. We will be using gimp, an image/paint processing application, with which we will create unique surfaces as applied through texture maps.

Effective worlds are only as effective as the light that illuminates them. We will be using the constructs in VRML 2.0 to light our models. Vertex normals will be an important tool to illuminate our models as well as giving us more tools to create interesting appearances to our models.

Performance is a very important aspect to effective VRML. We will be looking at supplying different level of details (LOD) for our models. Models further from the observer do not require the detail as they would require if they were closer. By specifying simpler models for those further away, performance is enhanced appreciably. This will be another area of investigation.

Other topics we will consider are the inclusion of sound and animation, as time and weather permit.







Pete Carswell is the coordinator for the VRML project. His office is in OSC, cubicle 420-7, phone 292-1091.

For assistance, write or call 614-292-0890.