Ming Jiang (Lawrence Livermore National Laboratory), Michael Andereck (Ohio State University), Alexander J. Pertica (Lawrence Livermore National Laboratory), Scot S. Olivier (Lawrence Livermore National Laboratory)
Keywords: Scientific visualization, computer graphics, and out-of-core algorithms
Abstract:
Visualization plays a crucial role in Space Situational Awareness, because it has the expressive power to provide insights to support the assessment and planning process of SSA. As computers become more powerful, SSA modeling and simulation in testbed environments are producing massive amounts of data, including more debris particles at higher fidelity and longer orbital propagations with uncertainty quantification. What is lacking in current visualization systems for SSA is the ability to visualize and exploit these massive amounts of modeling and simulation data in real-time. The main reason for this lack of capability is the underlying assumption that the entire data set can be completely loaded into the main memory before any processing or visualization, which is certainly not the case anymore. One effective strategy for dealing with massive amounts of data that cannot fit in main memory is to use out-of-core, or external memory, algorithms, which operate on smaller subsets of the data, while keeping in memory only as much of the data as needed by the algorithms. In this paper, we present a scalable visualization system for SSA simulation data that operates in an out-of-core fashion for the data access and the rendering process. We utilize an out-of-core octree to spatially partition simulation data points, and we use a realtime rendering approach that incorporates techniques for view frustum culling and discrete level-of-detail. We provide experimental results to demonstrate the efficacy of our proposed visualization system.
Date of Conference: September 14-17, 2010
Track: Posters