Multiple objects tracking in the presence of long term occlusions

We present a robust object tracking algorithm that handles spatially extended and temporally long object occlusions. The proposed approach is based on the concept of ``object permanence'' which suggests that a totally occluded object will re-emerge near its occluder. The proposed method does not require prior training to account for differences in the shape, size, color or motion of the objects to be tracked. Instead, the method automatically and dynamically builds appropriate object representations that enable robust and effective tracking and occlusion reasoning. The proposed approach has been evaluated on several image sequences showing either complex object manipulation tasks or human activity in the context of surveillance applications. Experimental results demonstrate that the developed tracker is capable of handling several challenging situations, where the labels of objects are correctly identified and maintained over time, despite the complex interactions among the tracked objects that lead to several layers of occlusions.

For more details see:
http://www.ics.forth.gr/~argyros/research/occlusions.html

Reference:
V. Papadourakis, A.A. Argyros, "Multiple Objects Tracking in the Presence of Long-term Occlusions", in Computer Vision and Image Understanding, Elsevier, vol. 114, issue 7, pp. 835-846, July 2010.

Category:

Science & Technology

Tags:

• tracking
• occlusions
• object tracking
• occlusions handling

[AntonisArgyros]

Hello, no prior preparation on the objects is needed. A background model of the scene is needed, so a few views of the scene without the objects is required. The current, _unoptimized_ implementation processes a few (2-3) frames per second on an Intel Core Duo CPU P8700 @ 2.53GHz, 4 GB RAM, but this depends, of course, on image resolution, number of objects, sizes, interaction, etc.

[ahmedrefaat]

is the object tracking in real-time, and is there any prior preparation of the scene or objects?

[ss83032]

這實在是太厲害了