Robust hierarchical multiple hypothesis tracker for multiple-object tracking

Research output: Contribution to journalArticle

26 Citations (Scopus)

Abstract

Multiple object tracking is a fundamental subsystem of many higher level applications such as traffic monitoring, people counting, robotic vision and many more. This paper explains in details the methodology of building a robust hierarchical multiple hypothesis tracker for tracking multiple objects in the videos. The main novelties of our approach are anchor-based track initialization, prediction assistance for unconfirmed track and two virtual measurements for confirmed track. The system is built mainly to deal with the problems of merge, split, fragments and occlusion. The system is divided into two levels where the first level obtains the measurement input from foreground segmentation and clustered optical flow. Only K-best hypothesis and one-to-one association are considered. Two more virtual measurements are constructed to help track retention rate for the second level, which are based on predicted state and division of occluded foreground segments. Track based K-best hypothesis with multiple associations are considered for more comprehensive observation assignment. Histogram intersection testing is performed to limit the tracker bounding box expansion. Simulation results show that all our algorithms perform well in the surroundings mentioned above. Two performance metrics are used; multiple-object tracking accuracy (MOTA) and multiple-object tracking precision (MOTP). Our tracker have performed the best compared to the benchmark trackers in both performance evaluation metrics. The main weakness of our algorithms is the heavy processing requirement.

Original languageEnglish
Pages (from-to)12319-12331
Number of pages13
JournalExpert Systems with Applications
Volume39
Issue number16
DOIs
Publication statusPublished - 15 Nov 2012

Fingerprint

Optical flows
Anchors
Robotics
Monitoring
Testing
Processing

Keywords

  • Hierarchical system
  • Histogram intersection
  • Multiple hypothesis tracker
  • Multiple object tracking
  • Occlusion prediction

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Engineering(all)

Cite this

Robust hierarchical multiple hypothesis tracker for multiple-object tracking. / Zulkifley, Mohd Asyraf; Moran, Bill.

In: Expert Systems with Applications, Vol. 39, No. 16, 15.11.2012, p. 12319-12331.

Research output: Contribution to journalArticle

@article{b0914b2735ad41e0bd8b0f5718941b7c,
title = "Robust hierarchical multiple hypothesis tracker for multiple-object tracking",
abstract = "Multiple object tracking is a fundamental subsystem of many higher level applications such as traffic monitoring, people counting, robotic vision and many more. This paper explains in details the methodology of building a robust hierarchical multiple hypothesis tracker for tracking multiple objects in the videos. The main novelties of our approach are anchor-based track initialization, prediction assistance for unconfirmed track and two virtual measurements for confirmed track. The system is built mainly to deal with the problems of merge, split, fragments and occlusion. The system is divided into two levels where the first level obtains the measurement input from foreground segmentation and clustered optical flow. Only K-best hypothesis and one-to-one association are considered. Two more virtual measurements are constructed to help track retention rate for the second level, which are based on predicted state and division of occluded foreground segments. Track based K-best hypothesis with multiple associations are considered for more comprehensive observation assignment. Histogram intersection testing is performed to limit the tracker bounding box expansion. Simulation results show that all our algorithms perform well in the surroundings mentioned above. Two performance metrics are used; multiple-object tracking accuracy (MOTA) and multiple-object tracking precision (MOTP). Our tracker have performed the best compared to the benchmark trackers in both performance evaluation metrics. The main weakness of our algorithms is the heavy processing requirement.",
keywords = "Hierarchical system, Histogram intersection, Multiple hypothesis tracker, Multiple object tracking, Occlusion prediction",
author = "Zulkifley, {Mohd Asyraf} and Bill Moran",
year = "2012",
month = "11",
day = "15",
doi = "10.1016/j.eswa.2012.03.004",
language = "English",
volume = "39",
pages = "12319--12331",
journal = "Expert Systems with Applications",
issn = "0957-4174",
publisher = "Elsevier Limited",
number = "16",

}

TY - JOUR

T1 - Robust hierarchical multiple hypothesis tracker for multiple-object tracking

AU - Zulkifley, Mohd Asyraf

AU - Moran, Bill

PY - 2012/11/15

Y1 - 2012/11/15

N2 - Multiple object tracking is a fundamental subsystem of many higher level applications such as traffic monitoring, people counting, robotic vision and many more. This paper explains in details the methodology of building a robust hierarchical multiple hypothesis tracker for tracking multiple objects in the videos. The main novelties of our approach are anchor-based track initialization, prediction assistance for unconfirmed track and two virtual measurements for confirmed track. The system is built mainly to deal with the problems of merge, split, fragments and occlusion. The system is divided into two levels where the first level obtains the measurement input from foreground segmentation and clustered optical flow. Only K-best hypothesis and one-to-one association are considered. Two more virtual measurements are constructed to help track retention rate for the second level, which are based on predicted state and division of occluded foreground segments. Track based K-best hypothesis with multiple associations are considered for more comprehensive observation assignment. Histogram intersection testing is performed to limit the tracker bounding box expansion. Simulation results show that all our algorithms perform well in the surroundings mentioned above. Two performance metrics are used; multiple-object tracking accuracy (MOTA) and multiple-object tracking precision (MOTP). Our tracker have performed the best compared to the benchmark trackers in both performance evaluation metrics. The main weakness of our algorithms is the heavy processing requirement.

AB - Multiple object tracking is a fundamental subsystem of many higher level applications such as traffic monitoring, people counting, robotic vision and many more. This paper explains in details the methodology of building a robust hierarchical multiple hypothesis tracker for tracking multiple objects in the videos. The main novelties of our approach are anchor-based track initialization, prediction assistance for unconfirmed track and two virtual measurements for confirmed track. The system is built mainly to deal with the problems of merge, split, fragments and occlusion. The system is divided into two levels where the first level obtains the measurement input from foreground segmentation and clustered optical flow. Only K-best hypothesis and one-to-one association are considered. Two more virtual measurements are constructed to help track retention rate for the second level, which are based on predicted state and division of occluded foreground segments. Track based K-best hypothesis with multiple associations are considered for more comprehensive observation assignment. Histogram intersection testing is performed to limit the tracker bounding box expansion. Simulation results show that all our algorithms perform well in the surroundings mentioned above. Two performance metrics are used; multiple-object tracking accuracy (MOTA) and multiple-object tracking precision (MOTP). Our tracker have performed the best compared to the benchmark trackers in both performance evaluation metrics. The main weakness of our algorithms is the heavy processing requirement.

KW - Hierarchical system

KW - Histogram intersection

KW - Multiple hypothesis tracker

KW - Multiple object tracking

KW - Occlusion prediction

UR - http://www.scopus.com/inward/record.url?scp=84864428138&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84864428138&partnerID=8YFLogxK

U2 - 10.1016/j.eswa.2012.03.004

DO - 10.1016/j.eswa.2012.03.004

M3 - Article

VL - 39

SP - 12319

EP - 12331

JO - Expert Systems with Applications

JF - Expert Systems with Applications

SN - 0957-4174

IS - 16

ER -