자료유형 | 학위논문 |
---|---|
서명/저자사항 | Visualization and Analysis of Vector Field Simulation Data. |
개인저자 | Liao, Yangguang. |
단체저자명 | University of California, Davis. Computer Science. |
발행사항 | [S.l.]: University of California, Davis., 2019. |
발행사항 | Ann Arbor: ProQuest Dissertations & Theses, 2019. |
형태사항 | 126 p. |
기본자료 저록 | Dissertations Abstracts International 81-06B. Dissertation Abstract International |
ISBN | 9781392438985 |
학위논문주기 | Thesis (Ph.D.)--University of California, Davis, 2019. |
일반주기 |
Source: Dissertations Abstracts International, Volume: 81-06, Section: B.
Advisor: Kreylos, Oliver. |
이용제한사항 | This item must not be sold to any third party vendors. |
요약 | A fundamental challenge for visualizing vector fields is that the complexity of the field causes structural and depth perception obstacles, which impede transmitting information accurately and effectively. In fluid dynamics, there is a long history of focusing on how to visualize the flow as a way to understand it. The line integral convolution (LIC) method is frequently used to visualize two-dimensional (2D) data but has rarely been used in three-dimensional (3D) scenes. To address the demand of applying 3D LIC in scientific flow visualization, my dissertation introduces a new solution to visualize a particularly challenging data type: large-scale simulation of the flow fields in planetary cores and the resultant magnetic fields. Geodynamo theory is able to describe the flow motion in a planet's liquid core, which generates the planetary magnetic field. Numerical simulations make it possible to explore and understand planetary magnetism by producing high-resolution vector field data. A good visualization of the model output is essential to understand the geodynamo process. Motivated by this multi-dimensional flow field visualization challenge and the demands of comprehending geodynamo simulations, a LIC-based visualization method is introduced to depict vector fields as a form of volumetric representation. Developed through the collaboration between visualization and domain scientists, this application is motivated by both actual domain usage and the need for technical improvement. To address the need for highly efficient and scalable parallel flow visualization methods, I developed a flow visualization system for large unstructured simulation data using parallel 3D LIC. The main consideration for a parallel LIC implementation is a trade-off between the additional memory cost of replicating cells at sub-domain boundaries, or the communication cost of exchanging those data among computation nodes. To improve scalability, A load-balancing scheme is introduced that partitions datasets based on estimated LIC computation time. I also introduce a data-driven sub-domain extension scheme that determines which external cells at a sub-domain boundary need to be added based on current boundary cells. This new scheme reduces memory overhead because the same visual quality can be achieved with a significantly smaller number of replicated external cells. In my dissertation, I evaluate and conclude that the new parallel 3D LIC visualization method is an efficient and scalable approach to the visualization of large and complex 3D vector fields.To solve the challenging issues of occlusion and overlapping caused by 3D LIC visualization, I introduce a progressive visualization in an immersive environment. The new virtual reality based visualization can help to address the structural problem in 3D visualization. In my dissertation, I define a new progressive visualization experience by using an innovative 3D brush tool for interaction to help users locate and extract targeted information quickly. This new 3D LIC visualization provides flexibility to apply varying parameters on the same data domain. By using this new LIC brush tool, experienced users are able to locate target features efficiently by reducing interfering information. |
일반주제명 | Computer science. |
언어 | 영어 |
바로가기 | ![]() |