Magnitude-based streamlines seed point selection for unsteady flow visualization

• Main Author: Yusoff, Yusman Azimi
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://eprints.utm.my/id/eprint/98104/1/YusmanAzimiYusoffPSC2020.pdf; http://eprints.utm.my/id/eprint/98104/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:143761

Flow visualization is a method utilized to obtain information from flow data sets. Proper blood flow visualization can assist surgeons in treating the patients. However, the main problem in visualizing the blood flow inside the aorta is the unsteady blood flow rate. Thus, an unsteady flow visualization method is required to show the blood flow clearly. Unfortunately, streamlines cannot be used by time-dependent flow visualization. This research aims to propose an improvement for the current streamline visualization technique and appearance by implementing an improved streamline generation method based on structured grid vector data to visualize the unsteady flow. The research methodology follows a comparative study method with the Evenly-Spaced Seed Point placement (ESSP) method as the benchmark. Magnitude-Based Seed Point placement (MBSP) and selective streamlines enhancement are introduced to produce longer, uniform, and clutter-free streamlines output. A total of 20 visualization results are produced with different streamlines separation distance. Results are then evaluated by comparing streamlines count and uniformity score. Subsequently, survey and expert reviews are carried out to strengthen the analysis. Survey questions are distributed to respondents that have data visualization knowledge background in order to get feedback related to streamlines uniformity and enhancement. In addition, experts review is conducted to get feedback based on current researches and techniques utilized in the related fields. Results indicate that streamlines count for MBSP are higher, but the differences are neglectable. Uniformity analysis shows good performance; with 80% of the MBSP results have better uniformity. Survey responses show 65% of respondents agreed MBSP results have better uniformity compared to ESSP. Majority of the respondents (92%) agreed that selective streamlines is a better approach. Experts review highlights that MBSP can distribute streamlines better in 3-dimension space compared to ESSP. Two significant findings are identified in this research: magnitude is proven to be an important input to locate seed points; and selective streamlines enhancement is a more effective approach as compared to global streamlines enhancement.