July 21, 1998 1

Centre for Intelligent Machines
McGill University
3480 University St.
Montréal (Québec) H3A 2A7, Canada
Tel. : (514) 933-5795

simra@@cim.mcgill.ca the Faculty of Graduate Studies and Research Department of Computer Science Faculty of Graduate Studies and Research Master of Science Computer Science

Robert Sim MCMXCVIII

Mobile Robot Localisation Using Learned Landmarks

Robert Sim

July 1998

 
• Contents
• List of Figures
• Introduction
  • Problem Statement
  • Approach
  • Applications
  • Outline
• Previous Work
  • Triangulation Methods
  • Kalman Filtering
  • Feature-Based Methods
  • Sensor Inversion
  • Visual Attention
• Learning Landmarks
  • Edge Detection
  • Feature Interpretation
  • Landmark Detection
• Visual Tracking
  • Landmark Recognition
  • Landmark Tracking
  • Example: A Small Database
• Position Estimation
  • Estimation by Linear Combination
  • Robust Estimate Combination
    • Estimating Error
    • Removing Outliers
• Experimental Results
  • A Simple Scene
  • Parameter Variation
    • Appearance-only Pose Estimates
    • Using the Edge Distribution
  • A Larger Scene
  • Two Indoor Scenes
    • A Laboratory Environment
    • Laboratory Environment Revisited
  • Recovering Orientation
• Discussion and Conclusions
  • Overview
  • Landmarks
  • Pose Estimation
  • Experimental Results
  • Future Work
    • Visual Attention
    • Visual Tracking
    • Parameterisation Properties
  • Conclusion
• References
• About this document ...