         
|
GRASP Lab Seminar 2003-2004
April 23, 1:00 PM, Levine Hall 307, hosted by Kostas Daniilidis.
Peter
Corke
CSIRO
Vision-based Control for Field Robots
Abstract: Vision-based control, or visual-servoing is the use of low-level visual features, such as contours or moments, that can be computed quickly to provide input to a control system. Vision is used as a sensor in a similar way to how we would use an encoder or a gyroscope. Visual servoing is a planning free approach to control, that provides a practical alternative to the sense, model, plan and act paradigm. Visual servoing provides a control solution that avoids solving difficult problems such as scene interpretation and pose estimation. In this this talk I will briefly introduce the main approaches to vision-based control and then present some applications to field robotic systems: large scale systems for the mining industry, stabilization of a small-scale helicopter and pose control of a non-holonomic outdoor mobile robot.
Biography: Peter Corke is a Senior Principal Research Scientist at CSIRO, and his primary research focus has been on vision-based robot control. He holds B.Eng and M.Eng.Sc degrees, both in electrical engineering, and a PhD, all from the University of Melbourne. He is an adjunct professor at the Australian National University and the University of Queensland, and is a member of the editorial board of the International Journal of Robotics Research. He has held visiting positions at CMU's Robotics Institute, the Coordinated Science Laboratory at UIUC and at the GRASP Lab of the University of Pennsylvania.
