Daniel P. Huttenlocher
Associate Professor
Cornell Weiss Presidential Fellow
Ph.D. MIT, 1988

My research area is computer vision, and my work focuses on the problems of model-based recognition, geometric shape comparison, and the computation of visual correspondence. My approach to these problems involves a combination of theoretical analysis, algorithm design, implementation, and experimentation. The overall goal of my work is to produce systems that work well in practice and have a firm underlying theoretical and algorithmic foundation. Recently we have begun to look at tasks in which two-dimensional shape matching methods can be used to solve problems that are three-dimensional in nature. The main question that we are investigating is whether purely two-dimensional representations are sufficient for solving certain tasks, such as motion tracking or visual navigation. The initial results are encouraging. For example we have a system that uses visual information to guide a mobile robot using only two-dimen-sional representations. The main focus of our work this past year has been in building systems in order to apply our theoretical and algorithmic results from the last few years to real problems. In addition to robotic applications, we are working on searching images of documents (in conjunction with the Cornell Digital Library initiative) and on using our motion tracking techniques to provide automatic structuring of video data-e.g. marking "sections" when a person enters or leaves the scene.


University Activities

Professional Activities



Return to:
1995-1996 Annual Report Home Page
Departmental Home Page

If you have questions or comments please contact: www@cs.cornell.edu.

Last modified: 2 November 1996 by Denise Moore (denise@cs.cornell.edu).