My research can be roughly classified into two directions: 1) data visualization
and 2) geometric modeling and geometry processing.
For data visualization, I am interested in investigating scalable
solutions to the analysis and representation of higher-ordered scientific data, such
as vector and tensor fields, using topology-based and other innovative
methods. Our solutions have found applications in science and engineering,
including computational fluid dynamics (CFD), automatic and aircraft design,
mechanical engineering, climate study, oceanography, earthquake engineering,
fusion science, and medical practice.
For geometric modeling and geometry processing, I aim to develop efficient
and robust techniques for the generation and optimization of volumetric meshes,
especially structured meshes, for the subsequent computation tasks, such as
simulations. Meshing is an indispensable step for many critical scientific
computation. The quality of the meshes often determines the speed and accuracy
of the computation. My research seeks to improve the state-of-the-art meshing
techniques substantially. At the same time, I am also interested in developing
innovative techniques for people to create and control various vector
and tensor fields for many computer graphics and simulation tasks. This may
provide an active learning environment for experts and general audience to
gain new knowledge through interactive creation of different dynamic effects.
My research in the above directions has resulted in numerous publications.
Please refer to the "Publications" page for those papers. The following provides
the overview of some of my research projects.
Research Projects
Multi-scale Coherent Structure Extraction and Tracking for
Modern CFD Data Analysis
Description:
Coherent structures are persistent and recognizable patterns that can be found in fluid
flows. In turbulent flows, coherent structures are closely related to a diverse range of
physical phenomena, and understanding their behavior is crucial for characterizing,
predicting and controlling these flows. However, reliable identification and characterization
of coherent structures is challenging due to their diversity and complex inter-relations
across different space- and time-scales. This project brings together experts from both the
data visualization and fluid mechanics communities to investigate novel solutions to
multi-scale coherent structure extraction, separation, tracking, and visualization.
Robust and Automated 2D and 3D Structured Mesh Generation and Optimization
Description:
Structured meshes, such as quadrilateral meshes in 2D and surfaces and hexahedral meshes in 3D,
are preferred by many critical applications in engineering, science, and medicine that require
solving certain partial differential equation (PDE) systems accurately to simulate various physical
phenomena (e.g., blood flow, heart beating, heat diffusion, structural analysis of mechanical
components and buildings, and more). The quality of the structured meshes on which those
computations are performed directly impact the success and accuracy of the simulations.
However, there currently lacks of robust and automated framework that can produce high quality
structured meshes for arbitrary input, especially for arbitrary 3D volume.
This project aims to address this challenge from two aspects: (1) the procedural optimization
of the initial (likely unstructured) meshes that are easily obtained to produce high quality
structured meshes and (2) the procedural adaption of the structure to the input model starting
from a simple but topologically equivalent space.
Sponsor:
This is a project that I am really passionate about since I joined UH.
I have been trying to apply for
funding (from both federal agencies and companies) to support this direction.
Unfortunately, I am yet to secure any fundings for it. But I believe in this direction
and will continue trying.
Project period: September, 2012 -- present
[Project webpage (under construction)]
Generating Hierarchical Vector-Valued Data Summaries for Scalable Flow Data
Processing, Analysis and Visualization
Description:
Most existing analysis and visualization techniques for vector fields are not scalable
to the real-world data that continue increasing in size and detail. This situation is
worsen given the finite resolution and dimensions of modern displays and the limited
capacity of human visual perception channel, all of which prevent us from understanding
this inherently complex data as a whole or in fine detail. To address these challenges,
this project aims to develop an effective summary represetation for vector fields.
This summary will enable a number of applications for scientific discovery and education
including the scalable and knowledge-assisted exploration of flow data, vector field
comparison, and vector field synthesis gaming. To construct such a summary, the proposed
research will first investigate the inherent relation among the physical properties
of the vector fields and their geometric appearance, and then develop a novel,
dimension independent intermediate representation that seamlessly integrates various
flow information characterized from different perspectives and in various scales.
From this intermediate representation, a hierarchical summary for vector fields will
be defined. This research represents one step towards a unified framework of knowledge
discovery and integrity from heterogeneous data sources.
Advanced Web-based Visualization for Drilling Data Analystics
Description:
This project aims to develop effective and interactive visual analysis system(s) for
the visualization and monitoring of drilling well log data. Two separated but related
systems were developed: one for the anomaly detection from the live streamling drilling
log data and the other for the visualization of the bore hole wearing measuring.
Sponsor: Shell (CW173096)
Project period: October, 2018 -- September, 2019
Define and Construct an Enhanced Graph Representation
for Multiscale Vector Field Data Summarization
Description:
Vector field data analysis is indispensable for many applications in science and engineering,
ranging from climate study, physics, chemistry, automobile design, to medical practice.
Most existing analysis techniques for vector field data are not scalable to the real-world
data with ever-increasing sizes and complexity. More importantly, the inherent limited visual
perception channel largely constrains the ability to understand the complex geometric and
physical behaviors of vector fields as a whole or in detail. To address these challenges,
this project investigates a graph-based vector field data reduction for the subsequent extraction
of a multiscale vector field data summary. In addition to the construction of the enhanced graph
representation, a number of novel vector field analysis techniques are developed to enrich the
current state-of-the-art of the vector field analysis and visualization.
