IVIS INTERACTIVE VOLUME VISUAL USE MULTIRESOLUTION OBJECT SPACE POINT SPLATTING

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Interactive visualization of large volume data sets is a very important research problem and is critical in a number of application domains in scientific visualization and medical imaging. A major challenge in volume visualization is to overcome scalability problems associated with very large data sets that can reach thousands of samples in each dimension, and thus due to the exponential growth of the volume data result in intractable size. Therefore, it is critical to develop rendering algorithms and systems that scale well with size while maintaining interactive frame rates at high display quality. Direct volume rendering is amid the most effective and direct techniques to get insight into the 3D structure of volumetric scalar field data set. In this project we will develop a novel high-quality and efficient interactive direct volume rendering approach based on hardware accelerated point splatting. In particular, it includes the investigation and analysis of blending functions for smooth interpolation of the voxel grid, and the development of multiresolution volume models and error metrics for efficient level-ofdetail (LOD) based rendering. Furthermore, we will implement our volume rendering techniques on top of a scalable cluster rendering system. This research project will result in a novel and fast volume visualization technique that allows LOD-based direct volume rendering and isosurface visualization as well as support shading. Through hardware accelerated splatting we will achieve interactive frame rates, and the hierarchical organization as well as the cluster rendering system will provide scalability to large data sets.