Multi-Object Tracking Algorithm Based on Spatial Constraints

DOI

10.2991/csic-15.2015.91How to use a DOI?

Keywords

Histogram of oriented gradients, Multi-object tracking, Spatial constraints, Online learning

Abstract

For multi-object tracking in complex environments, this paper presents an improved tracking algorithm based on spatial constraints. With basic framework of Dalal-Triggs detector (which uses HOG features to describe the image blocks and an SVM to predict the existence of objects), we use a graph structure model to constrain the spatial relationship between the multiple objects that are being tracked. Experiments show that spatial constraints among the objects make greatly improved performance of the tracker in multiple objects tracking. In the video of camera significant movement, fast-moving objects, objects change appearance and occlusion, the tracker performs well.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Yuanhang Cheng
AU  - Jing Wang
PY  - 2015/07
DA  - 2015/07
TI  - Multi-Object Tracking Algorithm Based on Spatial Constraints
BT  - Proceedings of the 2015 International Conference on Computer Science and Intelligent Communication
PB  - Atlantis Press
SP  - 379
EP  - 382
SN  - 2352-538X
UR  - https://doi.org/10.2991/csic-15.2015.91
DO  - 10.2991/csic-15.2015.91
ID  - Cheng2015/07
ER  -