Advanced search
1 file | 416.89 KB

A canonical correlation analysis based motion model for probabilistic visual tracking

Author
Organization
Abstract
Particle filters are often used for tracking objects within a scene. As the prediction model of a particle filter is often implemented using basic movement predictions such as random walk, constant velocity or acceleration, these models will usually be incorrect. Therefore, this paper proposes a new approach, based on a Canonical Correlation Analysis (CCA) tracking method which provides an object specific motion model. This model is used to construct a proposal distribution of the prediction model which predicts new states, increasing the robustness of the particle filter. Results confirm an increase in accuracy compared to state-of-the-art methods
Keywords
particle filter, prediction model, Canonical Correlation Analysis, Object tracking

Downloads

  • paper VCIP12 25 final final.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 416.89 KB

Citation

Please use this url to cite or link to this publication:

Chicago
Heyman, Tom, Vincent Spruyt, Sebastian Grünwedel, Alessandro Ledda, and Wilfried Philips. 2012. “A Canonical Correlation Analysis Based Motion Model for Probabilistic Visual Tracking.” In 2012 IEEE VISUAL COMMUNICATIONS AND IMAGE PROCESSING (VCIP), HO2–25. San Diego, California: Proceedings of Visual Communications and Image Processing.
APA
Heyman, T., Spruyt, V., Grünwedel, S., Ledda, A., & Philips, W. (2012). A canonical correlation analysis based motion model for probabilistic visual tracking. 2012 IEEE VISUAL COMMUNICATIONS AND IMAGE PROCESSING (VCIP) (pp. HO2–25). Presented at the IEEE Visual Communications and Image Processing (VCIP), San Diego, California: Proceedings of Visual Communications and Image Processing.
Vancouver
1.
Heyman T, Spruyt V, Grünwedel S, Ledda A, Philips W. A canonical correlation analysis based motion model for probabilistic visual tracking. 2012 IEEE VISUAL COMMUNICATIONS AND IMAGE PROCESSING (VCIP). San Diego, California: Proceedings of Visual Communications and Image Processing; 2012. p. HO2–25.
MLA
Heyman, Tom, Vincent Spruyt, Sebastian Grünwedel, et al. “A Canonical Correlation Analysis Based Motion Model for Probabilistic Visual Tracking.” 2012 IEEE VISUAL COMMUNICATIONS AND IMAGE PROCESSING (VCIP). San Diego, California: Proceedings of Visual Communications and Image Processing, 2012. HO2–25. Print.
@inproceedings{4158504,
  abstract     = {Particle \unmatched{fb01}lters are often used for tracking objects within a scene. As the prediction model of a particle \unmatched{fb01}lter is often implemented using basic movement predictions such as random walk, constant velocity or acceleration, these models will usually be incorrect. Therefore, this paper proposes a new approach, based on a Canonical Correlation Analysis (CCA) tracking method which provides an object speci\unmatched{fb01}c motion model. This model is used to construct a proposal distribution of the prediction model which predicts new states, increasing the robustness of the particle \unmatched{fb01}lter. Results con\unmatched{fb01}rm an increase in accuracy compared to state-of-the-art methods},
  author       = {Heyman, Tom and Spruyt, Vincent and Gr{\"u}nwedel, Sebastian and Ledda, Alessandro and Philips, Wilfried},
  booktitle    = {2012 IEEE VISUAL COMMUNICATIONS AND IMAGE PROCESSING (VCIP)},
  isbn         = {9781467344050},
  keyword      = {particle filter,prediction model,Canonical Correlation Analysis,Object tracking},
  language     = {eng},
  location     = {San Diego, California},
  pages        = {HO2--25},
  publisher    = {Proceedings of Visual Communications and Image Processing},
  title        = {A canonical correlation analysis based motion model for probabilistic visual tracking},
  year         = {2012},
}

Web of Science
Times cited: