October 23, 2008
ECE graduate student Rohit Patnaik and his research advisor, Professor Dave Casasent, have won the 2008 Best Student Paper Award at the Automatic Target Recognition (ATR) Conference, part of SPIE Defense and Security Symposium that was held in Orlando last March. The paper is titled "Clutter performance and confuser rejection on infrared data using distortion-invariant filters for ATR."
Lockheed Martin Sponsored the award, which was judged by a panel headed by the ATR conference chairs.
In pattern recognition, when an object in a scene needs to be located and recognized fast, a category of classifiers known as the distortion-invariant filter (DIF) is attractive. One DIF handles different object distortions (e.g. pose, scale, and elevation angle). The paper shows that one filter per object can handle the full 360* range of pose variations, a +/-30% range of scale variations (two filters can handle a +/-50% range of scale variations), and a +/-2* range of elevation angle. The paper also addresses other important issues concerning DIFs.
Patnaik's work was supported by Raytheon Missile Systems, Tucson, Arizona.
Rohit Patnaik and Professor Dave Casasent, winners of the 2008 Best Student Paper Award at the Automatic Target Recognition (ATR) Conference