Browsing by Subject "Direct Volume Rendering (DVR)"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Open Access Fast direct volume rendering of unstructured grids(Bilkent University, 1997-09) Berk, HakanScientific computing has become more and more important with the evolving technology. The vast amount of data that the scientific computing applications produce need new ways to be processed and interpreted by scientists. The large amount of data makes it very difficult for scientists to extract useful information from the data, and interpret it to reach a useful conclusion. Thus. visualization of such numerical data as an image, which is known as Scientific Visualization, is an indispensable tool for researchers. Volume Rendering is a very important branch of Scientific Visualization and makes it possible for scientists to visualize the 3-dimensional (3D) volumetric datasets. Volume Rendering algorithms can be classified into two categories: Indirect and Direct methods. Indirect methods are faster, but direct methods are more flexible and acurate. Direct methods can be classified into three categories: image-space (ray-casting), object-space (projection) and hybrid. The efficiency of a direct volume rendering (DVR) algorithm is strongly related to the way that it solves the underlying point location and view sort problems. Although these problems are almost trivial ones to solve in structured grids, they become more complex ones to deal with for unstructured grids. Researchers have tried to speed up the volume rendering of unstructured grids by using special graphics hardware, and parallel architectures, but the need for software solutions to these problems will always exist. These thesis is involved in solving those problems in unstructured grids via software methods. It investigates three distinct categories, namely image-space methods, object-space methods and hybrid methods for fast direct volume rendering of unstructured grids. The main objective of the thesis is to identify the relative superiorities and inferiorities of the algorithms in these three categories. A survey of existing methods is enriched by a discussion of their merits and shortcomings. Three new and fast algorithms to overcome the existing inefficiencies are proposed, and one existing algorithm is investigated in detail for better comparision. All of the proposed algorithms are aimed at producing correct, high quality images. Two of the proposed algorithms are pure ray-casting based solutions that support early ray termination and can handle cyclic grids. The relative performances of the proposed algorithms are experimented on a wide range of benchmark grids in a common framework for software methods and they are found to be faster than the existing best DVR algorithms.Item Open Access Image-space decomposition algorithms for sort-first parallel volume rendering of unstructured grids(Springer, 2000) Kutluca, H.; Kurç, T. M.; Aykanat, CevdetTwelve adaptive image-space decomposition algorithms are presented for sort-first parallel direct volume rendering (DVR) of unstructured grids on distributed-memory architectures. The algorithms are presented under a novel taxonomy based on the dimension of the screen decomposition, the dimension of the workload arrays used in the decomposition, and the scheme used for workload-array creation and querying the workload of a region. For the 2D decomposition schemes using 2D workload arrays, a novel scheme is proposed to query the exact number of screen-space bounding boxes of the primitives in a screen region in constant time. A probe-based chains-on-chains partitioning algorithm is exploited for load balancing in optimal 1D decomposition and iterative 2D rectilinear decomposition (RD). A new probe-based optimal 2D jagged decomposition (OJD) is proposed which is much faster than the dynamic-programming based OJD scheme proposed in the literature. The summed-area table is successfully exploited to query the workload of a rectangular region in constant time in both OJD and RD schemes for the subdivision of general 2D workload arrays. Two orthogonal recursive bisection (ORB) variants are adapted to relax the straight-line division restriction in conventional ORB through using the medians-of-medians approach on regular mesh and quadtree superimposed on the screen. Two approaches based on the Hilbert space-filling curve and graph-partitioning are also proposed. An efficient primitive classification scheme is proposed for redistribution in 1D, and 2D rectilinear and jagged decompositions. The performance comparison of the decomposition algorithms is modeled by establishing appropriate quality measures for load-balancing, amount of primitive replication and parallel execution time. The experimental results on a Parsytec CC system using a set of benchmark volumetric datasets verify the validity of the proposed performance models. The performance evaluation of the decomposition algorithms is also carried out through the sort-first parallelization of an efficient DVR algorithm.