Real-Time Volume Rendering of FCC Datasets Using Box-Splines on the Mobile Platforms

Abstract

Visualizing volumetric datasets using real-time volume rendering technique involves a large number of interpolation operations that are computationally expensive. This situation used to restrict real-time volume rendering methods to be used only on high-end graphics workstations or special-purpose hardware. This paper presented a real-time direct volume rendering (DVR) implementation of face centered cubic (FCC) datasets with box-spline interpolation on mobile devices. The latest version of OpenGL ES (Open Graphics Library for Embedded System) (3.0) is used for implementation to leverage cutting-edge 3D graphics technology, and it shows interactive performance (2.40 frame per second (FPS)) for moderate-sized volume datasets (64×64×64). Aims: To present a real-time direct volume rendering (DVR) implementation of face centered cubic (FCC) datasets with box-spline interpolation on mobile devices. Study Design:  Study is based on research conducted in computer lab, University of Seoul, South Korea. Place and Duration of Study: Computer Science Lab, Department of Computer Science, The graduate College, University of Seoul, South Korea, between June 2014 and April 2015. Methodology: The latest version of OpenGL ES (Open Graphics Library for Embedded System) (3.0) is used for implementation to leverage cutting-edge 3D graphics technology, and it shows interactive performance (2.40 frame per second (FPS)) for moderate-sized volume datasets (64×64×64). Results: In our implementation, we calculated the opacity using front–to-back composting whereby the viewing rays are traversed from the eye point into the volume. We also compared different volume sizes, having the same density. Conclusion: We have presented a real-time volume rendering technique for FCC datasets on mobile devices that efficiently evaluate spline value. Our work has proven that mobile devices constitute a valid program to achieve interactive volume visualization, despite the fact that the rendering capabilities are concentrated in comparison to desktop solutions, due to their inherent autonomy limitations.