Trajectory clustering based customer movement tracking in a supermarket

Xi WANG,Wei WU,Yuntao QIAN

DOI: https://doi.org/10.3969/j.issn.1673-4785.201401002

2015-01-01

Abstract:Tracking the moving targets in complex scenarios such as supermarkets can be a challenging task. This paper proposes a method to track moving customers in a supermarket by clustering the trajectories of the targets. In this method, all the background and short?time feature points are removed in the preprocessing step in order to re?fine the feature points, which were detected and tracked by the Kanade?Lucas?Tomasi ( KLT) algorithm. The occlu?sion problem of single frame static feature point clustering is solved by applying the mean shift algorithm to the traj?ectories of moving objects. Finally, the full trajectories of moving customers are generated by the matching algorithm of movement tracking. The algorithm tackles the stable tracking problem by optimally matching the feature point clusters between successive frames when the target goes across the boundary of the video region or has a complex trajectory . Experimental results showed that the proposed method can successfully track the trajectories of customers in various typical regions of the supermarket such as entrance, fresh area and checkout stand. This method is robust under partial occlusion, complex trajectory and asynchronous moving.