An Efficient Edge Artificial Intelligence MultiPedestrian Tracking Method With Rank Constraint

Characterized by the ability to handle varying number of objects, tracking by detection framework becomes increasingly popular in multiobject tracking (MOT) problem. However, the tracking performance heavily depends on the object detector. Considering that data association optimization and association affinity model are two key parts in MOT, an online multipedestrian tracking method is proposed to formulate a more effective association affinity model. It includes a two-step data association taking advantage of rank-based dynamic motion affinity model. The rank-based dynamic motion affinity model is used to estimate the object state and refine the trajectory for each of target to achieve the noiseless trajectory. Both strategies are beneficial to eliminate ambiguous detection responses during association. To fairly verify the proposed method, three public datasets are adopted. Both qualitative and quantitative experiment results demonstrate the superiorities of the proposed tracking algorithm in comparison with its counterparts.

I. Introduction

In last decades, multiobjects tracking (MOT) is an important research field in computer vision due to its wide applications in visual surveillance, traffic safety, automatic video content analysis, and virtual reality. We mainly focus on pedestrians tracking in this paper. This is a challenging problem since the multipedestrian in crowded scene always has similar appearance with occlusions, intersected trajectories, missing data, and camera motion. MOT methods based on tracking-by-detection (TBD) framework are the state-of-the-art method due to the development of pedestrian detectors [1], [2]. There are mainly two kinds of TBD tracking methods: the online tracking and the batch tracking [3]–[16]. For these methods, data association is a key issue, in which the association affinity model and the association optimization model are two main components.