Advanced

Object Tracking Using Multiple Overhead Cameras

Rehnqvist, Erik LU (2017) In Master's Theses in Mathematical Sciences FMA820 20162
Mathematics (Faculty of Engineering)
Abstract
Retail intelligence is a big and upcoming thing in the brick and mortar-world of retail. This simply means that the resellers want to gather as much data about their products,and possibly also their customers, as possible. Some stores might want to know different information than others, but the fact of the matter is that gathering off-line data is slowly entering the world of business. There are of course different methods to use to actually extract that data. In this thesis, image analysis is used in order to track objects in an enclosed space. Doing this, it is shown that it is possible to extract information about the movement of objects moving over multiple overhead cameras. This can for instance be used to keep track of a specific... (More)
Retail intelligence is a big and upcoming thing in the brick and mortar-world of retail. This simply means that the resellers want to gather as much data about their products,and possibly also their customers, as possible. Some stores might want to know different information than others, but the fact of the matter is that gathering off-line data is slowly entering the world of business. There are of course different methods to use to actually extract that data. In this thesis, image analysis is used in order to track objects in an enclosed space. Doing this, it is shown that it is possible to extract information about the movement of objects moving over multiple overhead cameras. This can for instance be used to keep track of a specific customer in a store, even though the tracked customer is moving over a large area and thus switching camera, and still keep the integrity of the customer, because no information about that
customer is extracted. The only information needed is their position and their previous positions. A simple solution is proposed in order to see if this could actually be achieved in a satisfying way.

The results show that it is indeed possible to track objects using the methods described in this thesis. However, the results are situational and depends on various matters; one of which is lighting in the scene. (Less)
Popular Abstract (Swedish)
Att spåra kunder i online-butiker har länge varit en självklarhet för branschen. Det de flesta inte vet är att det även är möjligt att spåra kunder i fysiska butiker, utan att den personliga integriteten går förlorad.
Please use this url to cite or link to this publication:
author
Rehnqvist, Erik LU
supervisor
organization
alternative title
Objektspårning med hjälp av multipla kameror i ett ovanifrån perspektiv
course
FMA820 20162
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Image analysis, tracking, overhead
publication/series
Master's Theses in Mathematical Sciences
report number
LUTFMA-3331-2017
ISSN
1404-6342
other publication id
2017:E60
language
English
id
8926603
date added to LUP
2017-12-01 15:29:36
date last changed
2017-12-01 15:29:36
@misc{8926603,
  abstract     = {Retail intelligence is a big and upcoming thing in the brick and mortar-world of retail. This simply means that the resellers want to gather as much data about their products,and possibly also their customers, as possible. Some stores might want to know different information than others, but the fact of the matter is that gathering off-line data is slowly entering the world of business. There are of course different methods to use to actually extract that data. In this thesis, image analysis is used in order to track objects in an enclosed space. Doing this, it is shown that it is possible to extract information about the movement of objects moving over multiple overhead cameras. This can for instance be used to keep track of a specific customer in a store, even though the tracked customer is moving over a large area and thus switching camera, and still keep the integrity of the customer, because no information about that
customer is extracted. The only information needed is their position and their previous positions. A simple solution is proposed in order to see if this could actually be achieved in a satisfying way.

The results show that it is indeed possible to track objects using the methods described in this thesis. However, the results are situational and depends on various matters; one of which is lighting in the scene.},
  author       = {Rehnqvist, Erik},
  issn         = {1404-6342},
  keyword      = {Image analysis,tracking,overhead},
  language     = {eng},
  note         = {Student Paper},
  series       = {Master's Theses in Mathematical Sciences},
  title        = {Object Tracking Using Multiple Overhead Cameras},
  year         = {2017},
}