Advanced

Multi-target Tracking Using on-line Viterbi Optimisation and Stochastic Modelling

Ardö, Håkan LU (2009)
Abstract
To study and compare the safety of intersection, traffic scientists today typically manually monitor the intersection

during several days and count how often certain events such as evasive manoeuvres occur. This is a

laboursome and costly procedure. The aim of this thesis is to provide tools that can reduce the amount of manual

labour required by using automated video analytics. Two methods for creating for such tools are presented.



The first method is a probabilistic background foreground segmentation that for each block of pixels calculate the

probability that this block currently views the static background or some moving foreground object. This is done

by deriving the... (More)
To study and compare the safety of intersection, traffic scientists today typically manually monitor the intersection

during several days and count how often certain events such as evasive manoeuvres occur. This is a

laboursome and costly procedure. The aim of this thesis is to provide tools that can reduce the amount of manual

labour required by using automated video analytics. Two methods for creating for such tools are presented.



The first method is a probabilistic background foreground segmentation that for each block of pixels calculate the

probability that this block currently views the static background or some moving foreground object. This is done

by deriving the probability distribution of the normalised cross correlation in the background and the foreground

case respectively. The background distribution depends on the amount of structure in the block.



The second method is a multi-target tracker that uses the probabilistic background foreground segmentation to

produce the trajectories of all objects in the scene. It operates online but with a few seconds delay in order to

incorporate information from both past and future frames when deciding on the current state. This means that the

output is guaranteed to be consistent, i.e. no jumping between different hypothesis, and the respect constrains

placed on the system such as "objects may not occupy the same space at the same time" or "objects may only

appear at the border of the image".



The methods have been tested both on synthetic and numerous sets of real data by implementing applications such

as people counting, loitering detection and traffic surveillance. The applications have been shown to perform very

well as long as the scene studied is not too large. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Dr Ferryman, James, University of Reading, UK
organization
publishing date
type
Thesis
publication status
published
subject
pages
171 pages
publisher
Centre for Mathematical Sciences, Lund University
defense location
Lecture Hall MH:C, Centre for Mathematical Sciences, Sölvegatan 18, Lund University, Faculty of Engineering
defense date
2009-02-13 13:15
ISSN
1404-0034
ISBN
978-628-7685-2
language
English
LU publication?
yes
id
290c25b3-674f-4a4e-819e-1ac3c34e79cb (old id 1278549)
date added to LUP
2009-01-19 12:42:28
date last changed
2016-09-19 08:44:45
@phdthesis{290c25b3-674f-4a4e-819e-1ac3c34e79cb,
  abstract     = {To study and compare the safety of intersection, traffic scientists today typically manually monitor the intersection<br/><br>
during several days and count how often certain events such as evasive manoeuvres occur. This is a<br/><br>
laboursome and costly procedure. The aim of this thesis is to provide tools that can reduce the amount of manual<br/><br>
labour required by using automated video analytics. Two methods for creating for such tools are presented.<br/><br>
<br/><br>
The first method is a probabilistic background foreground segmentation that for each block of pixels calculate the<br/><br>
probability that this block currently views the static background or some moving foreground object. This is done<br/><br>
by deriving the probability distribution of the normalised cross correlation in the background and the foreground<br/><br>
case respectively. The background distribution depends on the amount of structure in the block.<br/><br>
<br/><br>
The second method is a multi-target tracker that uses the probabilistic background foreground segmentation to<br/><br>
produce the trajectories of all objects in the scene. It operates online but with a few seconds delay in order to<br/><br>
incorporate information from both past and future frames when deciding on the current state. This means that the<br/><br>
output is guaranteed to be consistent, i.e. no jumping between different hypothesis, and the respect constrains<br/><br>
placed on the system such as "objects may not occupy the same space at the same time" or "objects may only<br/><br>
appear at the border of the image".<br/><br>
<br/><br>
The methods have been tested both on synthetic and numerous sets of real data by implementing applications such<br/><br>
as people counting, loitering detection and traffic surveillance. The applications have been shown to perform very<br/><br>
well as long as the scene studied is not too large.},
  author       = {Ardö, Håkan},
  isbn         = {978-628-7685-2},
  issn         = {1404-0034},
  language     = {eng},
  pages        = {171},
  publisher    = {Centre for Mathematical Sciences, Lund University},
  school       = {Lund University},
  title        = {Multi-target Tracking Using on-line Viterbi Optimisation and Stochastic Modelling},
  year         = {2009},
}