Non-supervised multiple-agent tracking is a complex task which demands a structured framework in order to accomplish it. Therefore, this proposal presents a system which is modular and hierarchically organised. It consists in several levels, working in cascade, which are defined according to the different functionalities to be performed. The goal of this work is to implement and experimentally verify a novel image-based algorithm which deals with serious segmentation difficulties, thereby being able to simultaneously perform a reliable tracking of several agents. As a result, agents’ trajectories are obtained, as well as quantitative information about their state at any time, such as their speed or size.

This work has been supported by EC grant IST-027110 for the HERMES project and by the Spanish MEC under projects TIC2003-08865 and DPI-2004-5414. J. Andrade and J. Gonzàlez also acknowledge the support of Juan de la Cierva Postdoctoral fellowships from the Spanish MEC.

This work was supported by the project 'Integration of robust perception, learning, and navigation systems in mobile robotics' (J-0929).

Presentado al 6th IASTED/VIIP celebrado en 2006 en Palma de Mallorca (España).

Peer Reviewed