ITR: Visualization and Interaction with Large Graphics Datasets over Networks

University Of Maryland, College Park, College Park MD

Investigators

Abstract

The last decade has witnessed a significant research thrust aimed at handling very large models in computer graphics. This project will extend those efforts by focusing on means to efficiently visualize, interact with, and manipulate large-scale graphics datasets over networks. The PI plans to develop techniques that will decrease latency, reduce bandwidth, and increase the interactivity for large graphics datasets over networks. The technical vision is to build a distributed rendering system that (a) takes advantage of the respective powers of image-based and object-based rendering for visualization of large-scale graphics datasets, (b)is well-suited for visualization-assisted collaboration and interaction across limited bandwidth links, and (c) takes into account the client, server, and network resources to deliver the best overall performance including times to transfer and render a dataset. Towards these ends the PI will explore techniques for high compression, progressive transmission, rendering from compressed data, handling time-varying datasets, and resource- aware rendering. The project work plan also seeks to develop tools for annotation, navigation, and sharing of educational spaces for use in synchronous and asynchronous learning environments that use three-dimensional datasets. This research will help push network-based collaborative design to the next level in a variety of science, engineering, and medical applications.