Publisher:Pasadena, CA : Jet Propulsion Laboratory, National Aeronautics and Space Administration, 2005.
Citation:International Symposium on Artificial Intelligence, Robotics, and Automation in Space, Munchen, Germany, September 5-9, 2005
Abstract:
In this paper, we evaluate our method on uncalibrated THEMIS data and find 1) agreement with manual cap edge identifications to within 28.2 km, and 2) high accuracy even with a reduced context window, yielding large reductions in memory requirements.