Online selecting discriminative tracking features using particle filter

The paper proposes a method to keep the tracker robust to background clutters by online selecting discriminative features from a large feature space. Furthermore, the feature selection procedure is embedded into the particle filtering process with the aid of existed "background" particles. Feature values from background patches and object observations are sampled during tracking and Fisher discriminant is employed to rank the classification capacity of each feature based on sampled values. Top-ranked discriminative features are selected into the appearance model and simultaneously invalid features are removed out to adjust the object representation adaptively. The implemented tracker with online discriminative feature selection module embedded shows promising results on experimental video sequences.

[1]  Dorin Comaniciu,et al.  Real-time tracking of non-rigid objects using mean shift , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[2]  Patrick Pérez,et al.  Maintaining multimodality through mixture tracking , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.

[3]  Stanley T. Birchfield,et al.  Elliptical head tracking using intensity gradients and color histograms , 1998, Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231).

[4]  David G. Stork,et al.  Pattern classification, 2nd Edition , 2000 .

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

[6]  Gregory D. Hager,et al.  Efficient Region Tracking With Parametric Models of Geometry and Illumination , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  R. Collins,et al.  On-line selection of discriminative tracking features , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.

[8]  Richard O. Duda,et al.  Pattern Classification by Iteratively Determined Linear and Piecewise Linear Discriminant Functions , 1966, IEEE Trans. Electron. Comput..

[9]  Yong Rui,et al.  Better proposal distributions: object tracking using unscented particle filter , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[10]  Shigeo Abe DrEng.  Pattern Classification , 2001, Springer London.

[11]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[12]  Arnold W. M. Smeulders,et al.  Tracking Aspects of the Foreground against the Background , 2004, ECCV.

[13]  Helman Stern,et al.  Adaptive color space switching for face tracking in multi-colored lighting environments , 2002, Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition.