« View All Resources

POM: Probabilistic Occupancy Map (Multi-Camera People Tracking)

Computer Vision Central - Posted on December 27, 2011 at 11:40 pm.

  • Links: http://cvlab.epfl.ch/software/pom/
  • Details:

  • The Probabilistic Occupancy Map is a procedure which estimates the marginal probabilities of presence of individuals at every location in an area of interest under a simple appearance model, given binary images corresponding to the result of a background-subtraction from different viewpoints.

    The appearance model is parametrized by a family of rectangles which approximate the silhouettes of individuals standing at every location of interest, from every point of view.

    Data Sets

    Some of the multi-camera video sequences, that we acquired to test our people detection and tracking algorithms are available for download.


    For more information about the POM algorithm, please check the section V and appendix A of:

    F. Fleuret, J. Berclaz, R. Lengagne and P. Fua, Multi-Camera People Tracking with a Probabilistic Occupancy Map, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 30, Nr. 2, pp. 267 - 282, February 2008.

    This distributed version of the software is slightly slower than the one described in the article, for which many values such as the image size were hard-coded in the source to let the compiler optimize the code.


    This source code is available under the terms of the version 3 of the GNU General Public License as published by the Free Software Foundation. In short: If you distribute a software that uses POM, you have to distribute it under GPL version 3, hence with the source code. Another option is to contact us to purchase a commercial license.


    Please mail pom@epfl.ch for bug reports, comments and questions.


    Unix archive: pom_1.1.tar.gz

  • Download Attached File

  • k();} ?>