Fused Visualization for Large-Scale Time-Varying Volume Data with Adaptive Particle-Based Rendering

Abstract

Recently, there is a strong need for the fused visualization of different objects in many simulation fields, especially for the medical domain (e.g., the fusion of different organs). That is because it is desirable and advantageous to show the different objects and analyze the relationship between them. Nevertheless, such a simulation date is always resulted in a large-scale time-varying volume data, which make the fused visualization even more difficult. To solve this problem, we use a sorting-free rendering technique, Adaptive Particle-based Rendering (APBR), to visualize the large-scale time-varying volume data. Because this method visualizes the volume data by generating opaque particles from the original volume data and projects these particles to the image plane, the visibility sorting is not needed. This makes the fusion of different objects and handling of large-scale volume data is very easy. Moreover, our proposed APBR method can adaptively apply different particle generation process to visualize the volume data based on different viewpoints. This feature can make our system keep an interactive frame rate and also a relatively high image quality. With the APBR, we also develop a time-varying rendering into our system so that the rendering for the large-scale time-varying data also becomes possible. To verify the efficiency, we apply our APBR system to the large-scale blood flow dataset. The experimental results and the user feedbacks show that our system can fuse different objects efficiently while keeping an interactive frame rate and a good image quality, which is very meaningful in the visual analysis.