Interactive Exploration of Large-scale Streamlines of Vector Fields via a Curve Segment Neighborhood Graph

TL;DR

This paper presents a web-based interactive system for exploring large-scale streamline vector fields using a Curve Segment Neighborhood Graph (CSNG) to identify flow structures efficiently.

Contribution

It introduces a novel CSNG data structure and an interactive visualization approach supporting real-time, multi-level exploration of complex streamline datasets in a web environment.

Findings

Real-time performance on datasets with hundreds of thousands of segments

Effective identification of coherent flow structures using CSNG

Supports interactive, multi-level exploration and detailed analysis

Abstract

Streamlines have been widely used to represent and analyze various steady vector fields. To sufficiently represent important features in complex vector fields (like flow), a large number of streamlines are required. Due to the lack of a rigorous definition of features or patterns in streamlines, user interaction and exploration are required to achieve effective interpretation. Existing approaches based on clustering or pattern search, while valuable for specific analysis tasks, often face challenges in supporting interactive and level-of-detail exploration of large-scale curve-based data, particularly when real-time parameter adjustment and iterative refinement are needed. To address this, we design and implement an interactive web-based system. Our system utilizes a Curve Segment Neighborhood Graph (CSNG) to encode the neighboring relationships between curve segments. CSNG enables us to adapt a fast community detection algorithm to identify coherent flow structures and spatial groupings in the streamlines interactively. CSNG also supports a multi-level exploration through an enhanced force-directed layout. Furthermore, our system integrates an adjacency matrix representation to reveal detailed inter-relations among segments. To achieve real-time performance within a web browser, our system employs matrix compression for memory-efficient CSNG storage and parallel processing. We have applied our system to analyze and interpret complex patterns in several streamline datasets. Our experiments show that we achieve real-time performance on datasets with hundreds of thousands of segments.