Pedestrian tracking and stereo matching of tracklets for autonomous vehicles

Research output: Chapter in Book/Conference paperConference paper

Abstract

The prediction of the surrounding pedestrians' walking paths is a vital part for autonomous driving systems in the aspect of traffic safety. In this paper, we propose a pipeline which tracks pedestrians captured by a stereo camera system onboard a mobile vehicle, composes the pedestrian tracklets, clusters the tracklets to form trajectories, and matches the trajectories. The output 3D pedestrian trajectories can be used for further applications such as pedestrian trajectory prediction for driverless vehicles. Our algorithm has been compared with various state-of-art pedestrian tracking methods. Our experimental results show that the visual temporal features computed by our algorithm are effective for trajectory representation and that, by incorporating tracklet clustering into the pipeline, the pedestrian tracking performance is improved.

Original languageEnglish
Title of host publication2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings
PublisherIEEE, Institute of Electrical and Electronics Engineers
ISBN (Electronic)9781728112176
DOIs
Publication statusPublished - 2019
Event89th IEEE Vehicular Technology Conference, VTC Spring 2019 - Kuala Lumpur, Malaysia
Duration: 28 Apr 20191 May 2019

Publication series

NameIEEE Vehicular Technology Conference
Volume2019-April
ISSN (Print)1550-2252

Conference

Conference89th IEEE Vehicular Technology Conference, VTC Spring 2019
CountryMalaysia
CityKuala Lumpur
Period28/04/191/05/19

Fingerprint

Stereo Matching
Autonomous Vehicles
Trajectories
Trajectory
Pipelines
Prediction
Safety
Camera
Cameras
Traffic
Clustering
Path
Output
Experimental Results

Cite this

Xue, H., Huynh, D. Q., & Reynolds, M. (2019). Pedestrian tracking and stereo matching of tracklets for autonomous vehicles. In 2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings [8746329] (IEEE Vehicular Technology Conference; Vol. 2019-April). IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/VTCSpring.2019.8746329
Xue, Hao ; Huynh, Du Q. ; Reynolds, Mark. / Pedestrian tracking and stereo matching of tracklets for autonomous vehicles. 2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings. IEEE, Institute of Electrical and Electronics Engineers, 2019. (IEEE Vehicular Technology Conference).
@inproceedings{2ba38ad1ffcd476fa711c4ea630591d9,
title = "Pedestrian tracking and stereo matching of tracklets for autonomous vehicles",
abstract = "The prediction of the surrounding pedestrians' walking paths is a vital part for autonomous driving systems in the aspect of traffic safety. In this paper, we propose a pipeline which tracks pedestrians captured by a stereo camera system onboard a mobile vehicle, composes the pedestrian tracklets, clusters the tracklets to form trajectories, and matches the trajectories. The output 3D pedestrian trajectories can be used for further applications such as pedestrian trajectory prediction for driverless vehicles. Our algorithm has been compared with various state-of-art pedestrian tracking methods. Our experimental results show that the visual temporal features computed by our algorithm are effective for trajectory representation and that, by incorporating tracklet clustering into the pipeline, the pedestrian tracking performance is improved.",
keywords = "Autonomous vehicles, Pedestrian tracking, Stereo matching, Tracklet clustering, Trajectory reconstruction",
author = "Hao Xue and Huynh, {Du Q.} and Mark Reynolds",
year = "2019",
doi = "10.1109/VTCSpring.2019.8746329",
language = "English",
series = "IEEE Vehicular Technology Conference",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
booktitle = "2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings",
address = "United States",

}

Xue, H, Huynh, DQ & Reynolds, M 2019, Pedestrian tracking and stereo matching of tracklets for autonomous vehicles. in 2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings., 8746329, IEEE Vehicular Technology Conference, vol. 2019-April, IEEE, Institute of Electrical and Electronics Engineers, 89th IEEE Vehicular Technology Conference, VTC Spring 2019, Kuala Lumpur, Malaysia, 28/04/19. https://doi.org/10.1109/VTCSpring.2019.8746329

Pedestrian tracking and stereo matching of tracklets for autonomous vehicles. / Xue, Hao; Huynh, Du Q.; Reynolds, Mark.

2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings. IEEE, Institute of Electrical and Electronics Engineers, 2019. 8746329 (IEEE Vehicular Technology Conference; Vol. 2019-April).

Research output: Chapter in Book/Conference paperConference paper

TY - GEN

T1 - Pedestrian tracking and stereo matching of tracklets for autonomous vehicles

AU - Xue, Hao

AU - Huynh, Du Q.

AU - Reynolds, Mark

PY - 2019

Y1 - 2019

N2 - The prediction of the surrounding pedestrians' walking paths is a vital part for autonomous driving systems in the aspect of traffic safety. In this paper, we propose a pipeline which tracks pedestrians captured by a stereo camera system onboard a mobile vehicle, composes the pedestrian tracklets, clusters the tracklets to form trajectories, and matches the trajectories. The output 3D pedestrian trajectories can be used for further applications such as pedestrian trajectory prediction for driverless vehicles. Our algorithm has been compared with various state-of-art pedestrian tracking methods. Our experimental results show that the visual temporal features computed by our algorithm are effective for trajectory representation and that, by incorporating tracklet clustering into the pipeline, the pedestrian tracking performance is improved.

AB - The prediction of the surrounding pedestrians' walking paths is a vital part for autonomous driving systems in the aspect of traffic safety. In this paper, we propose a pipeline which tracks pedestrians captured by a stereo camera system onboard a mobile vehicle, composes the pedestrian tracklets, clusters the tracklets to form trajectories, and matches the trajectories. The output 3D pedestrian trajectories can be used for further applications such as pedestrian trajectory prediction for driverless vehicles. Our algorithm has been compared with various state-of-art pedestrian tracking methods. Our experimental results show that the visual temporal features computed by our algorithm are effective for trajectory representation and that, by incorporating tracklet clustering into the pipeline, the pedestrian tracking performance is improved.

KW - Autonomous vehicles

KW - Pedestrian tracking

KW - Stereo matching

KW - Tracklet clustering

KW - Trajectory reconstruction

UR - http://www.scopus.com/inward/record.url?scp=85068974362&partnerID=8YFLogxK

U2 - 10.1109/VTCSpring.2019.8746329

DO - 10.1109/VTCSpring.2019.8746329

M3 - Conference paper

T3 - IEEE Vehicular Technology Conference

BT - 2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings

PB - IEEE, Institute of Electrical and Electronics Engineers

ER -

Xue H, Huynh DQ, Reynolds M. Pedestrian tracking and stereo matching of tracklets for autonomous vehicles. In 2019 IEEE 89th Vehicular Technology Conference, VTC Spring 2019 - Proceedings. IEEE, Institute of Electrical and Electronics Engineers. 2019. 8746329. (IEEE Vehicular Technology Conference). https://doi.org/10.1109/VTCSpring.2019.8746329