A Point-Based Remote Visualization Pipeline For Large-Scale Virtual Reality

Authors: Ge, J.

Publication: EVL, PhD Dissertation

Jinghua Ge’s PH.D. Dissertation - Submitted as partial fulfillment of the requirements for the degree of Doctor of Philosophy in Computer Science in the Graduate College of the University of Illinois at Chicago, 2007.

State-of-the-art Virtual Reality technologies such as Varrier™ bring better interaction and comprehension into visualization experience. But VR applications are still limited in the area of large-scale scientific visualization mostly because of the intensive graphics computation for VR viewing.

The goal of this thesis is to design and implement a distributed visualization framework which combines VR technologies and remote computing resources through a high speed network, so that large-scale scientific datasets can be visualized in real-time on local VR devices.

The framework is designed to be a scalable distributed system with pipelined data retrieval, computation, and visualization for various datasets. Scalability makes the system adaptive to the available computing and visualizing resource configurations. Each subsystem of the distributed system can perform either cluster-based parallel computing or single workstation-based sequential computing. The pipeline configuration can be optimized based on a balanced granularity as the ratio of computation to communication. The pipeline is an MIMD design which explores computing and networking parallelism along with the data flow.

Date: May 1, 2007

Document: View PDF
Point patch (L) and Point Cluster (R) representation of a point packing - J. Ge, EVL

Related Entries



Related Categories