Level Set Restricted Voronoi Tessellation for Large scale Spatial Statistical Analysis.

Area of Science:

• Scientific visualization
• Computational geometry
• Statistical analysis

Background:

• Analyzing large-scale multivariate volumetric data presents challenges in scale, complexity, and occlusion.
• Existing methods in topological segmentation and feature extraction offer partial solutions.

Purpose of the Study:

• To introduce a novel spatial statistical decomposition method for efficient analysis of complex volumetric data.
• To develop a parallel implementation and interactive visualization system for large-scale data analysis.

Main Methods:

• The method utilizes level sets, connected components, and a restricted centroidal Voronoi tessellation.
• Data structures are organized into a nested hierarchy for efficient region-of-interest extraction.
• An efficient parallel implementation and interactive visualization system are developed.

Main Results:

• The new method facilitates a coherent nested hierarchy of features.
• The approach supports flexible and efficient out-of-core region-of-interest extraction.
• The system was successfully applied to turbulent combustion data.

Conclusions:

• The combined approach enables an interactive spatial statistical analysis workflow for large-scale data.
• It facilitates a top-down, multi-level-of-detail analysis linking phase space statistics with spatial features.
• This method enhances the exploration and understanding of complex scientific datasets.