Multicamera People Tracking with a Probabilistic Occupancy Map

TLDR

The study aims to combine a generative model with dynamic programming to track up to six people across thousands of frames from multiple synchronized cameras despite occlusions and lighting changes. The method uses a generative model paired with dynamic programming to produce metrically accurate trajectories for each tracked individual. The results show that the generative model can handle occlusions with only background subtraction and unknown person counts, and that processing trajectories separately with a ranking heuristic yields reliable multi‑person tracking over long sequences.

Abstract

Given two to four synchronized video streams taken at eye level and from different angles, we show that we can effectively combine a generative model with dynamic programming to accurately follow up to six individuals across thousands of frames in spite of significant occlusions and lighting changes. In addition, we also derive metrically accurate trajectories for each one of them. Our contribution is twofold. First, we demonstrate that our generative model can effectively handle occlusions in each time frame independently, even when the only data available comes from the output of a simple background subtraction algorithm and when the number of individuals is unknown a priori. Second, we show that multi-person tracking can be reliably achieved by processing individual trajectories separately over long sequences, provided that a reasonable heuristic is used to rank these individuals and avoid confusing them with one another.