The goal of this class is to read state-of-the-art research
papers on multi-dimensional signal processing and approximation
theory. We will spend some time on fundamentals in 1D, after
which we will focus on the intricacies of multi-dimensional
signals. While the standard approach of multi-dimensional
processing are tensor-product extensions of one-dimensional
concepts, the main goal of this class is to bring across that the
world of multi-D is much richer and more exciting than this.
You will be applying the learned concepts to applications from
image processing, graphics and visualization. As this is a seminar
course, I am also open to other application areas.
(Please be aware that my understanding of multi-D is at least 3D.:))
Some topics that will be covered include:
- classical derivations by Shannon and Peterson, Middleton
- fundamentals of multi-dimensional signal processing
- reconstruction in shift-invariant spaces
- Box-splines and applications