Comparing Bayesian and Montecarlo localization for a robot with local vision

Position estimation is one of the classical problems in mobile robotics. For instance, robots have to know where in the map they are in order to use maps in any task involving navigation. Even in highly dynamical environments such as the RoboCup competition the robot behaviour or attitude depends on its position in the playground. The goal of this paper is to compare two probabilistic localization methods based on local vision for a mobile robot. The experimental set up is based on the Aibo league of the RoboCup, where the robotic dogs major sensor is the on-board camera. Two localization algorithms, Bayesian and Montecarlo ones, have been implemented and compared, and their behaviour studied in several situations. A simulator has been developed which adds actuation noise to the commands ordered to the motors and sensor errors to the images perceived. This way both algorithms use exactly the same data collection to estimate the robot position.

[1]  A. Jiménez,et al.  Estimación de la posición de un robot móvil , 1996 .

[2]  Wolfram Burgard,et al.  Using the CONDENSATION algorithm for robust, vision-based mobile robot localization , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[3]  R. Simmons,et al.  Probabilistic Navigation in Partially Observable Environments , 1995 .

[4]  Wolfram Burgard,et al.  Particle Filters for Mobile Robot Localization , 2001, Sequential Monte Carlo Methods in Practice.

[5]  William Whittaker,et al.  Conditional particle filters for simultaneous mobile robot localization and people-tracking , 2002, Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292).

[6]  Michael Isard,et al.  CONDENSATION—Conditional Density Propagation for Visual Tracking , 1998, International Journal of Computer Vision.

[7]  Wolfram Burgard,et al.  Robust Monte Carlo localization for mobile robots , 2001, Artif. Intell..

[8]  Wolfram Burgard,et al.  Monte Carlo Localization: Efficient Position Estimation for Mobile Robots , 1999, AAAI/IAAI.

[9]  Greg Welch,et al.  An Introduction to Kalman Filter , 1995, SIGGRAPH 2001.

[10]  Dieter Fox,et al.  An experimental comparison of localization methods continued , 2002, IEEE/RSJ International Conference on Intelligent Robots and Systems.

[11]  W. Burgard,et al.  Markov Localization for Mobile Robots in Dynamic Environments , 1999, J. Artif. Intell. Res..

[12]  Sebastian Thrun,et al.  Learning to Locate an Object in 3D Space from a Sequence of Camera Images , 1998, ICML.

[13]  Sebastian Thrun,et al.  Probabilistic Algorithms in Robotics , 2000, AI Mag..

[14]  Pat Langley,et al.  Editorial: On Machine Learning , 1986, Machine Learning.

[15]  Reid G. Simmons,et al.  Probabilistic Robot Navigation in Partially Observable Environments , 1995, IJCAI.

[16]  M. Buehler,et al.  Three-state Extended Kalman Filter for Mobile Robot Localization , 2002 .

[17]  Alessandro Saffiotti,et al.  Fuzzy landmark-based localization for a legged robot , 2000, Proceedings. 2000 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2000) (Cat. No.00CH37113).