3D Shape Recovery: Questions and Answers, and More Questions ...

Dr. Gerda Kamberova
Department of Computer Science
Hofstra University

We will present ongoing work on recovery of 3D shape invariants for computer vision and computer graphics. We are dealing with discrete, and in many cases, noisy data. We will review existing techniques for computing curvatures and curvature lines from discrete data. We will compare our results with experimental results presented in the literature. We will discuss directions for future research, in particular, the derivation of confidence measures of the shape invariants. This is a joint research with G. Kamberov from Stevens Institute of Technology.