Interactive Visualization: Visual Exploration, Manipulation, and Analysis of 3D Data


My research objectives are to explore and develop innovative interaction models and visualization techniques that will enable intuitive, efficient, interactive visual exploration, manipulation, comparison, and analysis of 3D (spatial) data. I am especially interested in 3D medical data sets although other types of data (for example, CAD meshes, 3D urban data, geographic data) are also of keen interest.

The resulting interactive visualization algorithms should be extremely easy to use, self-descriptive, aesthetically pleasing, should engage the user, and be underpinned by powerful data representation models and GPU-based data processing algorithms. To make effective use of all the available data in various applications, users must be able to quickly and effortlessly focus on and manipulate features of interest, while maintaining contextual views. This capability will be supported, in part, by creating "smart" software layers on top of the "raw" data, providing high-level access methods that will allow users to interact with the data using virtual analogies of familiar real-world actions. Examples of such interaction metaphors include cutting, pushing/pulling, painting, sliding, folding, and sketching. These "natural" input actions will be translated by the software into meaningful and efficient operations on the data, resulting in deeper insight.

Some of the interactive visualization tasks of interest are: object extraction, modeling, and 3D editing, visual comparison of two objects, object feature selection and emphasis, object manipulation in 3D, object measurement and annotation, tracking changes in objects, object interrelationship visualization and analysis, volume image navigation, and generating and integrating multiple views of data to provide view focus and context.

The novelty of this work lies in the synthesis of state-of-the-art computer graphics and GPU-based data processing algorithms, recent and/or original interaction models (such as sketching), and modern input and display devices, all of which must be seamlessly woven together to create highly effective interactive data visualization systems.