Efficient Incorporation of Markov Random Fields in Change Detection

Henrik Aanæs; Allan Aasbjerg Nielsen; Jens Michael Carstensen; Rasmus Larsen; Bjarne Kjær Ersbøll

doi:10.1109/IGARSS.2009.5417856

Efficient Incorporation of Markov Random Fields in Change Detection

Henrik Aanæs, Allan Aasbjerg Nielsen, Jens Michael Carstensen, Rasmus Larsen, Bjarne Kjær Ersbøll

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

460 Downloads (Pure)

Abstract

Many change detection algorithms work by calculating the probability of change on a pixel-wise basis. This is a disadvantage since one is usually looking for regions of change, and such information is not used in pixel-wise classification - per definition. This issue becomes apparent in the face of noise, implying that the pixel-wise classifier is also noisy. There is thus a need for incorporating local homogeneity constraints into such a change detection framework. For this modelling task Markov Random Fields are suitable. Markov Random Fields have, however, previously been plagued by lack of efficient optimization methods or numerical solvers. We here address the issue of efficient incorporation of local homogeneity constraints into change detection algorithms. We do this by exploiting recent advances in graph based algorithms for Markov Random Fields. This is combined with an IR-MAD change detector, and demonstrated on real data with good results.

Original language	English
Title of host publication	IEEE International Geoscience and remote sensing symposium
Volume	3
Publisher	IEEE
Publication date	2009
ISBN (Print)	978-1-4244-3394-0
DOIs	https://doi.org/10.1109/IGARSS.2009.5417856
Publication status	Published - 2009
Event	2009 IEEE International Geoscience and Remote Sensing Symposium - University of Cape Town, Cape Town, South Africa Duration: 12 Jul 2009 → 17 Jul 2009 https://ieeexplore.ieee.org/xpl/conhome/5403226/proceeding

Conference

Conference	2009 IEEE International Geoscience and Remote Sensing Symposium
Location	University of Cape Town
Country/Territory	South Africa
City	Cape Town
Period	12/07/2009 → 17/07/2009
Internet address	https://ieeexplore.ieee.org/xpl/conhome/5403226/proceeding

Bibliographical note

Copyright 2009 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

Keywords

Markov Random Fields
IR-MAD
Homogeneity Constraints
Graph Based Algorithms
Change Detection

Access to Document

10.1109/IGARSS.2009.5417856

IGARSS2009 0300689Final published version, 549 KB

OpenUrl availability

Full text

Cite this

@inproceedings{66d201c5b99c466c9eb6a9e737ac2f28,

title = "Efficient Incorporation of Markov Random Fields in Change Detection",

abstract = "Many change detection algorithms work by calculating the probability of change on a pixel-wise basis. This is a disadvantage since one is usually looking for regions of change, and such information is not used in pixel-wise classification - per definition. This issue becomes apparent in the face of noise, implying that the pixel-wise classifier is also noisy. There is thus a need for incorporating local homogeneity constraints into such a change detection framework. For this modelling task Markov Random Fields are suitable. Markov Random Fields have, however, previously been plagued by lack of efficient optimization methods or numerical solvers. We here address the issue of efficient incorporation of local homogeneity constraints into change detection algorithms. We do this by exploiting recent advances in graph based algorithms for Markov Random Fields. This is combined with an IR-MAD change detector, and demonstrated on real data with good results.",

keywords = "Markov Random Fields, IR-MAD, Homogeneity Constraints, Graph Based Algorithms, Change Detection",

author = "Henrik Aan{\ae}s and Nielsen, {Allan Aasbjerg} and Carstensen, {Jens Michael} and Rasmus Larsen and Ersb{\o}ll, {Bjarne Kj{\ae}r}",

note = "Copyright 2009 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.; 2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 ; Conference date: 12-07-2009 Through 17-07-2009",

year = "2009",

doi = "10.1109/IGARSS.2009.5417856",

language = "English",

isbn = "978-1-4244-3394-0",

volume = "3",

booktitle = "IEEE International Geoscience and remote sensing symposium",

publisher = "IEEE",

address = "United States",

url = "https://ieeexplore.ieee.org/xpl/conhome/5403226/proceeding",

}

TY - GEN

T1 - Efficient Incorporation of Markov Random Fields in Change Detection

AU - Aanæs, Henrik

AU - Nielsen, Allan Aasbjerg

AU - Carstensen, Jens Michael

AU - Larsen, Rasmus

AU - Ersbøll, Bjarne Kjær

N1 - Copyright 2009 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

PY - 2009

Y1 - 2009

N2 - Many change detection algorithms work by calculating the probability of change on a pixel-wise basis. This is a disadvantage since one is usually looking for regions of change, and such information is not used in pixel-wise classification - per definition. This issue becomes apparent in the face of noise, implying that the pixel-wise classifier is also noisy. There is thus a need for incorporating local homogeneity constraints into such a change detection framework. For this modelling task Markov Random Fields are suitable. Markov Random Fields have, however, previously been plagued by lack of efficient optimization methods or numerical solvers. We here address the issue of efficient incorporation of local homogeneity constraints into change detection algorithms. We do this by exploiting recent advances in graph based algorithms for Markov Random Fields. This is combined with an IR-MAD change detector, and demonstrated on real data with good results.

AB - Many change detection algorithms work by calculating the probability of change on a pixel-wise basis. This is a disadvantage since one is usually looking for regions of change, and such information is not used in pixel-wise classification - per definition. This issue becomes apparent in the face of noise, implying that the pixel-wise classifier is also noisy. There is thus a need for incorporating local homogeneity constraints into such a change detection framework. For this modelling task Markov Random Fields are suitable. Markov Random Fields have, however, previously been plagued by lack of efficient optimization methods or numerical solvers. We here address the issue of efficient incorporation of local homogeneity constraints into change detection algorithms. We do this by exploiting recent advances in graph based algorithms for Markov Random Fields. This is combined with an IR-MAD change detector, and demonstrated on real data with good results.

KW - Markov Random Fields

KW - IR-MAD

KW - Homogeneity Constraints

KW - Graph Based Algorithms

KW - Change Detection

U2 - 10.1109/IGARSS.2009.5417856

DO - 10.1109/IGARSS.2009.5417856

M3 - Article in proceedings

SN - 978-1-4244-3394-0

VL - 3

BT - IEEE International Geoscience and remote sensing symposium

PB - IEEE

T2 - 2009 IEEE International Geoscience and Remote Sensing Symposium

Y2 - 12 July 2009 through 17 July 2009

ER -

Efficient Incorporation of Markov Random Fields in Change Detection

Abstract

Conference

Bibliographical note

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this