A Hierarchical Association Framework for Multi-Object Tracking in Airborne Videos
This publication appears in: Remote Sensing
Authors: T. Chen, A. Pennisi, L. Zhi, Y. Zhang and H. Sahli
Number of Pages: 26
Publication Year: 2018
Multi-Object Tracking (MOT) in airborne videos is a challenging problem due tothe uncertain airborne vehicle motion, vibrations of the mounted camera, unreliable detections, changes of size, appearance and motion of the moving objects and occlusions caused by the interaction between moving and static objects in the scene. To deal with these problems, this work proposes a four-stage hierarchical association framework for multiple object tracking in airborne video. The proposed framework combines Data Association-based Tracking (DAT) methods and target tracking using a compressive tracking approach, to robustly track objects in complex airborne surveillance scenes. In each association stage, different sets of tracklets and detections are associated o efficiently handle local tracklet generation, local trajectory construction, global drifting tracklet orrection and global fragmented tracklet linking. Experiments with challenging airborne videos how significant tracking improvement compared to existing state-of-the-art methods.