Abstract
The image registration problem of finding a mapping that matches data from multiple cameras is computationally intensive. Current solutions to this problem [3], [4], [5] tolerate Gaussian noise, but are unable to perform the underlying global optimization computation in real time. This paper expands these approaches to other noise models and proposes the Terminal Repeller Unconstrained Subenergy Tunneling (TRUST) method, originally introduced by Cetin et al. [7], as an appropriate global optimization method for image registration. TRUST avoids local minima entrapment, without resorting to exhaustive search by using subenergy-tunneling and terminal repellers. The TRUST method applied to the registration problem shows good convergence results to the global minimum. Experimental results show TRUST to be more computationally efficient than either tabu search or genetic algorithms.
Original language | English |
---|---|
Pages (from-to) | 79-92 |
Number of pages | 14 |
Journal | IEEE Transactions on Knowledge and Data Engineering |
Volume | 14 |
Issue number | 1 |
DOIs | |
State | Published - Jan 2002 |
Funding
Jacob Barhen received the DSc degree from the Technion-Israel Institute of Technology, Haifa in 1978. He is the director of the Center for Engineering Science Advanced Research (CESAR) at the Oak Ridge National Laboratory (ORNL). He is also the head of the Intelligent and Emerging Computational Systems section in the Computer Science and Mathematics division. From 1987 to 1994, he was the head of the Nonlinear Science Information Processing Group at the Jet Propulsion Laboratory (JPL), California Institute of Technology. From 1978 to 1987, he was with ORNL, where he headed the Machine Intelligence Group. He has been the principal investigator and manager of numerous basic and applied research projects funded by the US Departments of Defense and Energy, NASA, and other US government agencies. His research interests include global optimization, neural networks, intelligent systems, computational methods for seismic imaging, and quantum computing and communication. He has authored more than 150 scientific papers and serves on the editorial boards of Neural Networks, Neural Processing Letters, Mathematical and Computer Modeling, and Concurrency. He holds eight US patents. He has received the 1998 R and D 100 Award for his invention of the TRUST global optimization method (Science, vol. 276, 1094-1097, 1997). He was also the recipient of 12 NASA Awards for Technical Innovation and was honored in 1993 and 1995 with two NASA Major Monetary Awards for his contribution to the National Space Program. He is a member of the IEEE, SPIE, the Planetary Society, and the International Neural Networks Society. Part of this research was funded by ONR grant number N00015-94-I-0343 to Professor Iyengar. Research for S.S. Iyengar was partially funded by Oak Ridge National Laboratory. Research for N.S.V. Rao and J. Barhen was sponsored by the Engineering Research Program of the Office of Basic Energy Sciences, US Department of Energy, under contract no. DE-AC05-00OR22725 with UT Battelle, LLC. The authors would like to thank the anonymous reviewers whose comments improved the presentation of this paper.
Keywords
- Genetic algorithms
- Global optimization
- Image registration
- Sensor fusion
- Subenergy-tunneling
- Tabu search
- Terminal repellers