Trajectory Association across Non-overlapping Moving Cameras in Planar Scenes

The ability to associate objects across multiple views allows co-operative use of an ensemble cameras for scene understanding. In this paper, we present a principled solution to object association where both the scene and the object motion are modeled. By making the motion model of each object with respect to time explicit, we are able to solve the trajectory association problem in a unified framework for overlapping or non-overlapping cameras. We recover the assignment of associations while simultaneously computing the maximum likelihood estimates of the inter-camera homographies and the trajectory parameters using the expectation maximization algorithm. Quantitative results on simulations are reported along with several results on real data.

[1]  Yaakov Bar-Shalom,et al.  Multitarget-multisensor tracking: Advanced applications , 1989 .

[2]  Chris Stauffer,et al.  Automated multi-camera planar tracking correspondence modeling , 2003, 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings..

[3]  Amnon Shashua,et al.  Multiple View Geometry of General Algebraic Curves , 2004, International Journal of Computer Vision.

[4]  Jeff A. Bilmes,et al.  A gentle tutorial of the em algorithm and its application to parameter estimation for Gaussian mixture and hidden Markov models , 1998 .

[5]  Bernhard P. Wrobel,et al.  Multiple View Geometry in Computer Vision , 2001 .

[6]  Ramin Zabih,et al.  Bayesian multi-camera surveillance , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[7]  Ying Shan,et al.  Unsupervised learning of discriminative edge measures for vehicle matching between non-overlapping cameras , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[8]  Mubarak Shah,et al.  Appearance modeling for tracking in multiple non-overlapping cameras , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[9]  Tim J. Ellis,et al.  Bridging the gaps between cameras , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[10]  Takeo Kanade,et al.  Algorithms for cooperative multisensor surveillance , 2001, Proc. IEEE.

[11]  Robert B. Fisher.  Self-Organization of Randomly Placed Sensors , 2002, ECCV.

[12]  Peter F. Sturm.  Vision 3D non calibrée : contributions à la reconstruction projective et étude des mouvements critiques pour l'auto-calibrage. (Uncalibrated 3D Vision: Contributions to Projective Reconstruction and Study of the Critical Motions for Self-Calibration) , 1997 .

[13]  W. Eric L. Grimson,et al.  Inference of non-overlapping camera network topology by measuring statistical dependence , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[14]  Katsushi Ikeuchi,et al.  Toward automatic robot instruction from perception-mapping human grasps to manipulator grasps , 1997, IEEE Trans. Robotics Autom..

[15]  Norimichi Ukita,et al.  Real-time multitarget tracking by a cooperative distributed vision system , 2002, Proc. IEEE.

[16]  Andrew W. Fitzgibbon,et al.  A Buyer's Guide to Conic Fitting , 1995, BMVC.

[17]  Trevor Darrell,et al.  Simultaneous calibration and tracking with a network of non-overlapping sensors , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[18]  Yaser Sheikh,et al.  Object tracking across multiple independently moving airborne cameras , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.