Advanced Rendering of Line Data with Ambient Occlusion and Transparency

Area of Science:

• Computer Graphics
• Scientific Visualization

Background:

• 3D lines are crucial for visualizing complex data in fields like fluid dynamics and fiber tractography.
• Global illumination and transparency enhance the perception of 3D features and reduce occlusion.

Purpose of the Study:

• To develop an efficient, high-quality GPU-based rendering approach for 3D line data.
• To incorporate ambient occlusion and transparency effects for improved data understanding.

Main Methods:

• GPU-based raycasting of rounded cones (geometric primitives with spherical endcaps).
• Object space ambient occlusion using efficient voxel cone tracing.
• A novel fragment visibility sorting strategy for handling large datasets.
• Exploitation of hierarchical opacity maps for performance enhancement.

Main Results:

• Achieved efficient, high-quality GPU rendering of 3D lines with ambient occlusion and transparency.
• Enabled interactive visualization of line datasets containing millions of segments.
• Demonstrated improved perception and reduced occlusion in complex 3D line data.

Conclusions:

• The proposed rendering approach significantly enhances the interactive visualization of complex 3D line data.
• The combination of rounded cones, voxel cone tracing, and fragment sorting offers an efficient solution for large-scale scientific visualization.