January 19, 2009
ECE graduate student Nicholas O'Donoughue has won the Best Student Paper award at the 2008 Asilomar Conference on Signals, Systems, and Computers held in Pacific Grove, California in October. The paper is titled "Signal-Domain Registration for Change Detection in TR-SAR" and his coauthored with his research advisor, José M. F. Moura and Yuanwei Jin, an assistant professor at the University of Maryland Eastern Shore, who was a Project Scientist in our department for the past year.
Change detection is a method for comparing images taken from different time instances, in order to classify changes in the target area. Time-Reversal SAR is a processing method for Synthetic Aperture Radar that increases performance in highly cluttered environments, and utilizes change detection to search for targets. Conventional change detection performs registration at the end of the processing chain, on the images formed by SAR processing, in order to coherently compare the data. TR-SAR, however, requires that this registration be done before SAR image formation, in the signal domain. The paper imposes restrictions on the problem and presents a solution to the alignment of two SAR datasets, and presents the results of experiments on data collected by Raytheon Missile Systems, Tucson, Arizona.
O'Donoughue is supported by a National Defense Science and Engineering Graduate (NDSEG) Fellowship. This work was partially supported by the DARPA Mathematical Time-Reversal Methods program.
There were 108 entries to the contest, which was sponsored by Texas Instruments.
Headshot of Nicholas O'Donoughue