home · mobile · calendar · colloquia · 2004-2005 · 

Colloquium - Fan

Multi-Aspect Target Detection/Tracking Using Particle Filters
Oklahoma State University

A new approach for multi-aspect target detection/tracking is proposed based on the sequential Monte Carlo estimation method. In addition to the conventional position and velocity parameters, we want to estimate aspect parameters as continuous values, including rotation, shearing, and scaling. Two particle filtering algorithms are implemented, i.e., Sequential Importance Resampling (SIR) and Auxiliary Particle Filter (APF). The simulation is conducted on the simulated infrared image sequences where the spatial clutter is modeled as the Gaussian-Markov random field (GMrF). Simulation results show that the proposed methods can well detect/track multi-aspect targets with time-variant aspects in a case of low target-to-clutter ratio.

Hosted by Jane Mulligan.

Department of Computer Science
University of Colorado Boulder
Boulder, CO 80309-0430 USA
May 5, 2012 (14:13)