Controlled Active Vision with Various Applications

Tuesday, February 1, 2011 - 11:00am - 12:00pm EST

  • Location:Dibner Building, LC400
    Five MetroTech Center, Brooklyn, New York

Speaker: Professor Allen R. Tannenbaum

Faculty Host: Professor Jonathan Chao

Abstract
In this talk, we will describe some of the key issues in controlled active vision, namely the utilization of visual information in a feedback loop. The applications range from visual tracking (e.g., laser tracking in turbulence, flying in formation of UAVs, etc.), nanoparticle ow control, and various medical imaging applications (image guided therapy and surgery). Accordingly, we will describe several models of active contours for which both local (edge-based) and global (statistics-based) information may be included for various segmentation tasks. We will indicate how statistical estimation and prediction ideas (e.g., particle ltering) may be naturally combined with this methodology. A novel model of directional active contour models for path-planning will be considered.
In addition to segmentation, the second key component of many active vision tasks is registration. The registration problem (especially in the elastic case) is still one of the great challenges in vision and image processing. Registration is the process of establishing a common geometric reference frame between two or more data sets obtained by possibly different imaging modalities. Registration has a substantial literature devoted to it, with numerous approaches ranging from optical flow to computational fluid dynamics. For this purpose, we propose using ideas from optimal mass transport. The mass transport problem was first formulated by Gaspar Monge in 1781, and concerned finding the optimal way, in the sense of minimal transportation cost, of moving a pile of soil from one site to another. This problem was given a modern formulation in the work of Kantorovich, and is now known as the \Monge{Kantorovich problem." The optimal mass transport approach has strong connections to optimal control, and can be the basis for a geometric observer theory for target tracking in which shape information is explicitly taken into account. Finally, we will describe how mass transport ideas may be utilized in order to formulate a novel distance on the space of probability measures with applications to spectral estimation and information geometry. The talk is designed to be accessible to a general audience with an interest in vision, control, and image processing. We will demonstrate our techniques on a wide variety of data sets both military and medical.

About the Speaker
Dr. Tannenbaum obtained his Ph.D. in mathematics from Harvard in 1976. He has held faculty positions at the Weizmann Institute of Science, McGill University, ETH in Zurich, Technion, Ben-Gurion University of the Negev, and University of Minnesota. He is presently Julian Hightower Professor of Electrical and Biomedical Engineering at Georgia Tech/Emory. Dr. Tannenbaum has authored or co-authored about 400 research papers, and is the author or co-author of three books: Invariance and Systems Theory, Feedback Control Theory (with J. Doyle and B. Francis), and Robust Control of Innite Dimensional Systems (with C. Foias and H. Ozbay). He also edited two volumes with Feedback Control, Nonlinear Systems, and Complexity (with B. Francis), and Mathematical Methods in Computer Vision (with Peter Olver). He also has four patents in computer vision and medical imaging. Dr. Tannenbaum has been an Associate Editor of several journals including SIAM J. Control and Optimization, Systems and Control Letters, Int. Robust and Nonlinear Control, SIAM Journal Imaging Science. He has won several awards including the Kennedy Research Prize, George Taylor Research Award, IEEE Fellow, SICE Best Paper Award, Foams 2000 Best Paper Award, MICCAI Best Paper Award, and Hugo Schuck Award (Best Paper at ACC). He has given a number of plenary talks including at the American Mathematical Society, SIAM, IEEE CDC 2000, MTNS, and SCICADE. He has done research in image processing, medical imaging, computer vision, robust control, systems theory, robotics, semiconductor process control, operator theory, functional analysis, cryptography, algebraic geometry, and invariant theory.