Online clustering-based multi-camera vehicle tracking in scenarios with overlapping FOVs

Abstract Multi-Target Multi-Camera (MTMC) vehicle tracking is an essential task of visual traffic monitoring, one the main research fields Intelligent Transportation Systems. Several offline approaches have been proposed to address this task; however, they are not compatible with real-world applications due their high latency and post-processing requirements. This lack suitable motivates our proposal: A new low-latency online approach for MTMC in scenarios partially overlapping view (FOVs), such as road intersections. Firstly, detects vehicles at each camera. Then, detections merged between cameras by applying cross-camera clustering based on appearance location. Lastly, clusters containing different same temporally associated compute tracks a frame-by-frame basis. The experiments show promising results while addressing challenges priori unknown time-varying number targets continuous state estimation them without performing any trajectories. Our code available http://www-vpu.eps.uam.es/publications/Online-MTMC-Tracking .