Robust Multi-Object Tracking Under a Wide Range of Real-World Conditions

Abstract

In this paper we propose a new paradigm for solving the correspondence problem and then determination of a motion trajectory based on a trisectional structure. The paradigm distinguishes between real world objects; extracts image features such as motion blobs and color patches; and abstract objects such as meta objects that denote real-world physical objects. The efficiency of the proposed method for determining the motion trajectories of moving objects are demonstrated in this paper on the basis of the analysis of real image sequences that are subject to severe disturbances (e.g. congestion and lighting transitions)