My dissertation, "Visualization and Analysis of Diffusion Tensor Fields", is
here. The visualization component of this work describes
an intuitive barycentric space for
describing the range of tensor shape, superquadric tensor glyphs,
fiber tracking based on "deflection", and the generalization of volume
rendering to diffusion tensor fields. The analysis component of this
work presents a mathematical framework for decomposing the gradient of
a symmetric second-order tensor field (a third-order tensor) into six
orthogonal components, composed of three degrees of freedom in shape
change, and three degrees of freedom in orientation change, which facilitates
a generalization of edge detection to tensor-valued images.
I'm interested in combining scientific visualization with image
processing and analysis, while guided by basic aspects of human visual
perception. Some research along these lines has
been completed. Some
software has also been written.