{"references": ["D. Gorodnichy, S. Malik and G. Roth, Affordable 3D face tracking using\nprojective vision, Proc. of Int. Conf. on Vision Interface, pp. 383-390,\nCalgary (Canada) May 2002.", "A.J. Davison, Y. Gonzalez-Cid and N. Kita, Real-time 3D SLAM\nwith wide-angle vision, 5th IFAC/EURON Symposium on Intelligent\nAutonomous Vehicles, July 2004.", "D. Margaritis and S. Thrun, Learning to locate an object in 3D space\nfrom a sequence of images. Proc. of Int. Conf. on Machine Learning, pp.\n332-340, 1998.", "D. Fox, W. Burgard, F. Dellaert and S. Thrun, Monte Carlo localization:\nefficient position estimation for mobile robots, In Proc. of 16th. AAAI\nNat. Conf. on Artificial Intelligenge, pp. 343-349, Orlando (USA), July\n1999", "S. Arulampalam, S. Maskell, N. Gordon and T. Clapp, A tutorial on\nparticle filters for on-line non-linear/non-gaussian bayesian tracking,\nIEEE Transactions on Signal Processing, vol. 50, no. 2, pp. 174-188,\n2002.", "A.M. Boumaza and J. Louchet, Mobile robot sensor fusion using flies\nIn G\u251c\u255dnter Raidl et.al. editors, Applications of Evolutionary Computing,\nEvo Workshops 2003, Lecture Notes in Computer Science, vol. 2611, pp.\n357-367, Springer, 2003.", "Jean Louchet, Using an individual evolution strategy for stereovision,\nGenetic Programming and Evolvable Machines, vol. 2, pp. 101-109, 2001", "M. Isard and A. Blake, CONDENSATION- conditional density propagation\nfor visual tracking, Int. Journal of Computer Vision, vol. 20, no. 1,\npp. 5-28, 1998.", "P. Perez, J. Vermaak and A. Blake, Data fusion for visual tracking with\nparticles, Proceedings of IEEE, vol. 92, no. 3, pp. 495-513, March 2004.\n[10] D. Mackay, Introduction to Monte Carlo methods, In M. Jordan editor,\nLearning in graphical models, pp. 175-204, MIT Press, 1999\n[11] A. J. Davison, Real-time simultaneous localisation and mapping with\na single camera, IEEE Int. Conf. on Computer Vision, ICCV-2003, pp.\n1403-1410, Nice (France), October 2003.\n[12] R.T. Collins, Multi-image focus of attention for rapid site model construction,\nIEEE Int. Conf. on Computer Vision and Pattern Recognition,\n1997, pp. 575-581, San Juan, Puerto Rico, June 1007."]}

This paper addresses the problem of determining the current 3D location of a moving object and robustly tracking it from a sequence of camera images. The approach presented here uses a particle filter and does not perform any explicit triangulation. Only the color of the object to be tracked is required, but not any precisemotion model. The observation model we have developed avoids the color filtering of the entire image. That and the Monte Carlotechniques inside the particle filter provide real time performance.Experiments with two real cameras are presented and lessons learned are commented. The approach scales easily to more than two cameras and new sensor cues.