University of Minnesota
University Relations
http://www.umn.edu/urelate
612-624-6868
myU OneStop


Go to unit's home.

Home | Seminars and Symposia | Past seminars/symposia: Wednesday, November 19, 2003

DTC Seminar Series

Design, Implementation and Evaluation of Novel Texture-Based Methods for the Effective Visualization of 3D Shape and Multivalued Data

by

Victoria Interrante
University of Minnesota
Computer Science and Engineering

Wednesday, November 19, 2003
1:00 pm

402 Walter Library

Victoria InterranteDownload slides (pdf 19.8 MB) In this talk, Dr. Interrante will present highlights of her research group's recent results, focusing primarily on two applications: conveying 3D shape with texture, and conveying multivalued flow data with color and texture. When people look at a 3D object in the real world with their two eyes, their visual system derives information about its intrinsic surface shape and position and orientation in space from a multitude of complementary cues. Some of these cues are inherent to the process of vision but others are affected by various external factors such as the surface material properties and the characteristics of the illumination. When creating computer-generated images, researchers have tremendous latitude in determining how to ‘set the scene’ so that the 3D shapes and spatial relationships of the objects in the scene can be easily, accurately and intuitively understood. For several years, Dr. Interrante and her students have been investigating methods for effectively using texture to convey this 3D shape information. Most recently their work has focused on investigating the question: if we could design the ideal texture to convey surface shape, what would the characteristics of that texture be? In the first part of this talk, Dr. Interrante will describe methods her group have recently developed for synthesizing custom-fitted shape-revealing textures over arbitrary doubly curved surfaces and will discuss the findings of their recent human observer studies investigating the impacts of various texture features on surface shape perception.

Interrante presenting

In the second part of this talk, Dr. Interrante will describe her recent work in developing methods for the effective visualization of multivalued flow data using color and texture. In particular, Dr. Interrante will discuss new techniques in color weaving and texture stitching for simultaneously conveying multiple scalar distributions superimposed over a common vector field in 2D, along with other related efforts in visualizing multivalued flow data using line integral convolution.

 

Dr. Interrante's research in computer graphics, visualization and perception focuses on the design and implementation of algorithms for making images that effectively communicate information. Her current projects involve feature extraction for the visualization and quantitative analysis of vortical structures in turbulent flows, texture synthesis for 3D shape representation and high quality realistic rendering, perceptually-based methods for the effective dynamic visualization of multivariate and uncertainty data, and gaze determination of infants from photographs. http://www.cs.umn.edu/people/faculty/index.php?id=145