Bilkent University
Department of Computer Engineering


Peano Curves and *-Trees for Computations on Large Images


Dr. Guna Seetharaman
Computing Technology Applications Branch Air Force Research Laboratory

There is has been a renewed interest in heterogeneous computing architecture, stimulated in part by the increased availability of graphics engines designed for video games, and high density FPGAs all designed to be fitted on standard desktop PCs. These architectural aspects of these processors differ significantly from that of the main host resulting in a data-reorganization step in a sequential computation designed to exploit the available resources. In addition, the multi-core and multi-threaded operations of the CPUs pose newer challenges. All these dictate a need for representation of large data that captures both locality in data and recursive decomposition of the data domain to suit divide and conquer paradigms. Peano curves -- a family of space filling curves that can be embedded in two and higher-dimensional bounded spaces – offer an attractive set of features. Image data compression and content based image retrieval among others have been tried successfully using a variety of tree-representations. The presentation will highlight the features, and the insights based on case studies.

Short bio: Dr. Guna Seetharaman, a member of the Science & Engineering personnel, is a Senior Computing Architectures Engineer, Computing Technology Applications Branch, AFRL, Rome, NY. Dr. Seetharaman spent 20 years as a professor of computer science and engineering, at the Air Force Institute of Technology and University of Louisiana at Lafayette. He was also a CNRS Invited professor at University of Paris XI on multiple tenures between1998-2005. He established and successfully ran the Computer Vision Laboratory, and Intelligent Robotics Laboratory (co-established) at The University of Louisiana at Lafayette. He also participated in the DARPA Grand Challenge as a charter member of Team CajunBot. He led the LiDAR data processing and obstacle detection efforts in Team CajunBot, demonstrated in 2005 and 2007 Grand Challenges. He has published more than 120 peer reviewed articles in: Computer Vision, low-altitude aerial imagery, SIMD-Parallel Computing, VLSI-signal processing, 3D Displays, Nano-Technology, micro-optics, and 3D Video analysis. He co-organized the DOE/ONR/NSF Sponsored Second International Workshop on Foundations of Decision and Information Fusion, in 1996 (Washington DC), and the IEEE Sixth International Workshop on Computer Architecture for Machine Perception, New Orleans, 2003. He guest edited IEEE COMPUTER special issue devoted to Unmanned Intelligent Autonomous Vehicles, Dec 2006. He also guest-edited a special issue of the EURASIP Journal on Embedded Computing in the topics of Intelligent Vehicles. He is an active member of the IEEE, and ACM. He is also a member of Tau Beta Pi, Eta Kappa Nu and Upsilon Pi Epsilon. He is a Paul Harris Fellow of the Rotary International. Dr. Seetharaman’s research interests focus on computational science and engineering, and advanced computing architectures. His efforts cut across high performance computing applications, embedded computing, computer vision, persistent, ubiquitous and net-centric image exploitation, micro-optics, and algorithm optimization.

Disclaimer: All views expressed in this presentation are that of the author and his collaborators, and do not represent or reflect the policies and priorities of his employer – US Air Force, Department Defense etc.

DATE: 7 October, 2010, Thursday @ 13:40