Attribute-Aware RBFs: Interactive Visualization of Time Series Particle Volumes Using RT Core Range Queries.

Area of Science:

• Computational physics
• Scientific visualization
• Computer graphics

Background:

• Smoothed-particle hydrodynamics (SPH) is crucial for simulating volumetric media but faces visualization challenges due to massive datasets.
• Existing Radial Basis Function (RBF) interpolation methods are computationally expensive and limited to density fields, hindering interactive analysis.

Purpose of the Study:

• To develop an efficient and interactive visualization technique for large-scale SPH simulations.
• To enable visualization of color-mapped attributes alongside density fields.
• To accelerate scalar field reconstruction and data updates for dynamic simulations.

Main Methods:

• Leveraging GPU ray tracing for accelerated scalar field reconstruction.
• Implementing a novel RBF interpolation scheme integrating per-particle colors and densities.
• Utilizing GPU-parallel tree construction and refitting for dynamic updates.
• Employing a Hilbert reordering scheme to optimize tree memory consumption.
• Adopting a spatio-temporal blue noise sampling for noise reduction in volumetric shadows.

Main Results:

• Achieved significantly faster and more detailed visualization of volumetric SPH datasets.
• Enabled interactive manipulation of particle attributes and simulation parameters.
• Demonstrated effective integration of color and density fields for richer data representation.
• Reduced memory footprint through Hilbert reordering and improved visualization quality via blue noise sampling.

Conclusions:

• The proposed GPU-accelerated RBF interpolation method enhances the interactivity and detail of SPH simulation visualization.
• This approach facilitates new insights into complex physics simulations by providing a more comprehensive view of volumetric data.
• The technique overcomes limitations of traditional methods, paving the way for advanced scientific discovery.