Tracking a large number of objects from multiple views

We propose a multi-object multi-camera framework for tracking large numbers of tightly-spaced objects that rapidly move in three dimensions. We formulate the prob-lem of finding correspondences across multiple views as a multidimensional assignment problem and use a greedy randomized adaptive search procedure to solve this NP-hard problem efficiently. To account for occlusions, we re-lax the one-to-one constraint that one measurement corre-sponds to one object and iteratively solve the relaxed as-signment problem. After correspondences are established, object trajectories are estimated by stereoscopic recon-struction using an epipolar-neighborhood search. We em-bedded our method into a tracker-to-tracker multi-view fu-sion system that not only obtains the three-dimensional tra-jectories of closely-moving objects but also accurately set-tles track uncertainties that could not be resolved from sin-gle views due to occlusion. We conducted experiments to validate our greedy assignment procedure and our tech-nique to recover from occlusions. We successfully track hundreds of flying bats and provide an analysis of their group behavior based on 150 reconstructed 3D trajectories. 1