Scientific Visualization and Virtual Reality are increasingly being used for the design of complex systems. These technologies offer powerful capabilities to make decisions that are cost and time effective. The next logical extension is to collaborate with these visual models in real-time, where parts of a design team are geographically separated. Specifically, visual collaboration enables ideas and proposed changes to be discussed exactly on a virtual model of a product. However, high-end visualization hardware and Internet technologies impede widespread use of real-time visual collaboration due to the large amount of data from which these representations are created. These data are typically in the form of 3D vertex-based models, which offer a high degree of realism when displayed, but at a price of storage, rendering speeds and processing efficiency. The more realistic the representation desired, the larger the number of vertices required and hence the higher the file size. In this paper, we propose a new data modeling and handling technique where traditional vertex-based models are converted into combinatorial B-Spline based wire-frame models that allow realtime visual collaboration in the context of typical virtual reality systems. Using appropriate filtering methods, parametric equations are computed for each curved segment in a vertexbased representation and bundled together with sampled linear segments of the model. The computed parametric equation based models occupy only a fraction of the size when compared to the original vertex-based models. These lightweight models can easily be transmitted over the Internet, in real-time, for viewing with a platform independent visual client program. The proposed methods were tested on several example data files to prove the method’s effectiveness.
展开▼
