inria-00503900, version 1

Extended phase field higher-order active contour models for networks

T. Peng ^1, 2, I. H. Jermyn ¹, V. Prinet ², J. Zerubia ( Auteur à contacter de préférence ) ¹

International Journal of Computer Vision 88, 1 (2010) 111--128

1 : ARIANA (INRIA Sophia Antipolis / Laboratoire I3S)
INRIA – Université Nice Sophia Antipolis [UNS] – CNRS : UMR7271 France
2 : Laboratoire Franco-Chinois d'Informatique, d'Automatique et de Mathématiques Appliquées (LIAMA)
http://liama.ia.ac.cn/
Institute of Automation, Chinese Academy of Sciences – Chinese Academy of Science (CAS) – Institut national de la recherche agronomique (INRA) – INRIA – Centre de coopération internationale en recherche agronomique pour le développement [CIRAD] – CNRS Institut d'Automatique - Académie des Sciences de Chine PO Box 2728 - Beijing 100080 R. P. Chine Tél. : (+ 86 10) 62 64 74 59 Fax : (+ 86 10) 62 64 74 58 Chine

Références bibliographiques

Type de publication : Articles dans des revues avec comité de lecture
Domaine :
Informatique/Traitement des images

Informatique/Vision par ordinateur et reconnaissance de formes
Titre : Extended phase field higher-order active contour models for networks
Résumé : This paper addresses the segmentation from an image of entities that have the form of a `network', \ie the region in the image corresponding to the entity is composed of branches joining together at junctions, \eg road or vascular networks. We present new phase field higher-order active contour (HOAC) prior models for network regions, and apply them to the segmentation of road networks from very high resolution satellite images. This is a hard problem for two reasons. First, the images are complex, with much `noise' in the road region due to cars, road markings, etc., while the background is very varied, containing many features that are locally similar to roads. Second, network regions are complex to model, because they may have arbitrary topology. In particular, we address a limitation of a previous model in which network branch width was constrained to be similar to maximum network branch radius of curvature, thereby providing a poor model of networks with straight narrow branches or highly curved, wide branches. We solve this problem by introducing first an additional nonlinear nonlocal HOAC term, and then an additional linear nonlocal HOAC term to improve the computational speed. Both terms allow separate control of branch width and branch curvature, and furnish better prolongation for the same width, but the linear term has several advantages: it is more efficient, and it is able to model multiple widths simultaneously. To cope with the difficulty of parameter selection for these models, we perform a stability analysis of a long bar with a given width, and hence show how to choose the parameters of the energy functions. After adding a likelihood energy, we use both models to extract the road network quasi-automatically from pieces of a QuickBird image, and compare the results to other models in the literature. The state-of-the-art results obtained demonstrate the superiority of our new models, the importance of strong prior knowledge in general, and of the new terms in particular.
Langue du document : Anglais
Titre de la revue :

International Journal of Computer Vision

Publisher Springer Verlag (Germany)

ISSN 0920-5691 (eISSN : 1573-1405)
Date de publication : 2010
Audience : internationale
Editeur commercial : Springer
Volume : 88
Numéro : 1
Pagination : 111--128
DOI : 10.1007/s11263-009-0304-3
Mots-clés : active contour – higher-order – phase field – shape – stability analysis – remote sensing – road network extraction

Liste des fichiers attachés à ce document :

TEX

(8.9 KB)

(6.9 KB)

(88 KB)

(32.5 KB)

(3.4 KB)

(46.6 KB)

(13.6 KB)

Image

(100.4 KB)

(100.4 KB)

(3.7 MB)

(1.6 MB)

(240.1 KB)

(344.6 KB)

(176.5 KB)

(1.3 MB)

(2.7 MB)

(240.1 KB)

(513 KB)

(443.5 KB)

(1.7 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.1 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

(1.3 MB)

d2is5pt5widthsurf.eps

(1.1 MB)

Image0330.eps

(3.7 MB)

Image0330linearL0.eps

(3.7 MB)

(3.7 MB)

(3.7 MB)

(3.7 MB)

(240.1 KB)

Image0719L2etape1500new.eps

(240.1 KB)

Image0719L2etape1new.eps

(240.1 KB)

Image0719L2etape27600basicphase.eps

(240.1 KB)

Image0719L2etape40000new.eps

(240.1 KB)

Image0719L2etape40000old.eps

(240.1 KB)

Image0719L2mleResult.eps

(240.1 KB)

Image0719mleresult.eps

(3.7 MB)

Image0719oldmodel.eps

(3.7 MB)

(3.7 MB)

(3.7 MB)

(1.3 MB)

(344.6 KB)

Image0720L1etape09000new.eps

(344.6 KB)

Image0720L1etape09000old.eps

(344.6 KB)

Image0721.eps

(2.7 MB)

Image0721L2.eps

(176.5 KB)

Image0721L2etape20000new.eps

(176.6 KB)

Image0721L2etape20000old.eps

(176.6 KB)

Image0801.eps

(3.7 MB)

Image0906.eps

(8 MB)

Image0906L2.eps

(513 KB)

Image0906L2etape15000new.eps

(513 KB)

Image0906L2etape50000old.eps

(513 KB)

Image0907.eps

(6.9 MB)

interactionacross.eps

(2.2 MB)

interactionalong.eps

(2.2 MB)

nonlinearrelationWandbetaandbeta2.eps

(1.1 MB)

(2 MB)

(2 MB)

(3.7 MB)

(235.1 KB)

smallroadL2etape15000old.eps

(235.1 KB)

smallroadL2etape38000new.eps

(235.1 KB)

smallroadlinearL0.eps

(3.7 MB)

PDF

ijcv.pdf

(3.7 MB)

ijcv.ps

(97.3 MB)

inria-00503900, version 1
http://hal.inria.fr/inria-00503900
oai:hal.inria.fr:inria-00503900
Contributeur : Ian Jermyn <>
Soumis le : Lundi 19 Juillet 2010, 11:48:20
Dernière modification le : Mardi 20 Juillet 2010, 08:50:48

Voir la fiche concise

Documents associés

PDF :

PS :

DOI : 10.1007/s11263-009-0304-3

Exporter

Bibtex EndNote TEI RefWorks

%% inria-00503900, version 1
%% http://hal.inria.fr/inria-00503900
@article{peng:inria-00503900,
    hal_id = {inria-00503900},
    url = {http://hal.inria.fr/inria-00503900},
    title = {{Extended phase field higher-order active contour models for networks}},
    author = {Peng, T. and Jermyn, I. H. and Prinet, V. and Zerubia, J.},
    abstract = {{This paper addresses the segmentation from an image of entities that have the form of a `network', \ie the region in the image corresponding to the entity is composed of branches joining together at junctions, \eg road or vascular networks. We present new phase field higher-order active contour (HOAC) prior models for network regions, and apply them to the segmentation of road networks from very high resolution satellite images. This is a hard problem for two reasons. First, the images are complex, with much `noise' in the road region due to cars, road markings, etc., while the background is very varied, containing many features that are locally similar to roads. Second, network regions are complex to model, because they may have arbitrary topology. In particular, we address a limitation of a previous model in which network branch width was constrained to be similar to maximum network branch radius of curvature, thereby providing a poor model of networks with straight narrow branches or highly curved, wide branches. We solve this problem by introducing first an additional nonlinear nonlocal HOAC term, and then an additional linear nonlocal HOAC term to improve the computational speed. Both terms allow separate control of branch width and branch curvature, and furnish better prolongation for the same width, but the linear term has several advantages: it is more efficient, and it is able to model multiple widths simultaneously. To cope with the difficulty of parameter selection for these models, we perform a stability analysis of a long bar with a given width, and hence show how to choose the parameters of the energy functions. After adding a likelihood energy, we use both models to extract the road network quasi-automatically from pieces of a QuickBird image, and compare the results to other models in the literature. The state-of-the-art results obtained demonstrate the superiority of our new models, the importance of strong prior knowledge in general, and of the new terms in particular.}},
    keywords = {active contour; higher-order; phase field; shape; stability analysis; remote sensing; road network extraction},
    language = {Anglais},
    affiliation = {ARIANA - INRIA Sophia Antipolis / Laboratoire I3S , Laboratoire Franco-Chinois d'Informatique, d'Automatique et de Math{\'e}matiques Appliqu{\'e}es - LIAMA},
    publisher = {Springer},
    pages = {111--128},
    journal = {International Journal of Computer Vision},
    volume = {88},
    number = {1 },
    audience = {internationale },
    doi = {10.1007/s11263-009-0304-3 },
    year = {2010},
    pdf = {http://hal.inria.fr/inria-00503900/PDF/ijcv.pdf},
}

%F inria-00503900, version 1
%0 Journal Article
%U http://hal.inria.fr/inria-00503900
%2 INFO:INFO_TI;INFO:INFO_CV
%T Extended phase field higher-order active contour models for networks
%A Peng, T.
%A Jermyn, I. H.
%A Prinet, V.
%A Zerubia, J.
%+ ARIANA - INRIA Sophia Antipolis / Laboratoire I3S , Laboratoire Franco-Chinois d'Informatique, d'Automatique et de Mathématiques Appliquées - LIAMA
%X This paper addresses the segmentation from an image of entities that have the form of a `network', \ie the region in the image corresponding to the entity is composed of branches joining together at junctions, \eg road or vascular networks. We present new phase field higher-order active contour (HOAC) prior models for network regions, and apply them to the segmentation of road networks from very high resolution satellite images. This is a hard problem for two reasons. First, the images are complex, with much `noise' in the road region due to cars, road markings, etc., while the background is very varied, containing many features that are locally similar to roads. Second, network regions are complex to model, because they may have arbitrary topology. In particular, we address a limitation of a previous model in which network branch width was constrained to be similar to maximum network branch radius of curvature, thereby providing a poor model of networks with straight narrow branches or highly curved, wide branches. We solve this problem by introducing first an additional nonlinear nonlocal HOAC term, and then an additional linear nonlocal HOAC term to improve the computational speed. Both terms allow separate control of branch width and branch curvature, and furnish better prolongation for the same width, but the linear term has several advantages: it is more efficient, and it is able to model multiple widths simultaneously. To cope with the difficulty of parameter selection for these models, we perform a stability analysis of a long bar with a given width, and hence show how to choose the parameters of the energy functions. After adding a likelihood energy, we use both models to extract the road network quasi-automatically from pieces of a QuickBird image, and compare the results to other models in the literature. The state-of-the-art results obtained demonstrate the superiority of our new models, the importance of strong prior knowledge in general, and of the new terms in particular.
%G Anglais
%J International Journal of Computer Vision
%8 2010-00-00
%2 internationale
%I Springer
%V 88
%N 1
%P 111--128
%R 10.1007/s11263-009-0304-3
%K active contour; higher-order; phase field; shape; stability analysis; remote sensing; road network extraction

<?xml version="1.0" encoding="UTF-8"?>
<listBibl xmlns="http://www.tei-c.org/ns/1.0" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML">
<listBibl type="article"><head>Articles dans des revues avec comité de lecture</head>
<biblStruct type="article" xml:id="inria-00503900"><analytic>
<author><persName><forename>T.</forename><surname>Peng</surname></persName>
<affiliation>
<orgName type="laboratory">INRIA Sophia Antipolis / Laboratoire I3S</orgName>
<orgName type="team">ARIANA</orgName>
<country>FR</country>
<orgName type="institution">INRIA</orgName>
<orgName type="institution">Université Nice Sophia Antipolis [UNS]</orgName>
<orgName type="institution">CNRS : UMR7271</orgName>
</affiliation>
<affiliation>
<orgName type="laboratory">LIAMA</orgName>
<orgName type="team">Laboratoire Franco-Chinois d'Informatique, d'Automatique et de Mathématiques Appliquées</orgName>
<country>CN</country>
<orgName type="institution">Institute of Automation, Chinese Academy of Sciences</orgName>
<orgName type="institution">Chinese Academy of Science (CAS)</orgName>
<orgName type="institution">Institut national de la recherche agronomique (INRA)</orgName>
<orgName type="institution">INRIA</orgName>
<orgName type="institution">Centre de coopération internationale en recherche agronomique pour le développement [CIRAD]</orgName>
<orgName type="institution">CNRS</orgName>
</affiliation>
</author>
<author><persName><forename>I. H.</forename><surname>Jermyn</surname></persName>
<affiliation>
<orgName type="laboratory">INRIA Sophia Antipolis / Laboratoire I3S</orgName>
<orgName type="team">ARIANA</orgName>
<country>FR</country>
<orgName type="institution">INRIA</orgName>
<orgName type="institution">Université Nice Sophia Antipolis [UNS]</orgName>
<orgName type="institution">CNRS : UMR7271</orgName>
</affiliation>
</author>
<author><persName><forename>V.</forename><surname>Prinet</surname></persName>
<affiliation>
<orgName type="laboratory">LIAMA</orgName>
<orgName type="team">Laboratoire Franco-Chinois d'Informatique, d'Automatique et de Mathématiques Appliquées</orgName>
<country>CN</country>
<orgName type="institution">Institute of Automation, Chinese Academy of Sciences</orgName>
<orgName type="institution">Chinese Academy of Science (CAS)</orgName>
<orgName type="institution">Institut national de la recherche agronomique (INRA)</orgName>
<orgName type="institution">INRIA</orgName>
<orgName type="institution">Centre de coopération internationale en recherche agronomique pour le développement [CIRAD]</orgName>
<orgName type="institution">CNRS</orgName>
</affiliation>
</author>
<author><persName><forename>J.</forename><surname>Zerubia</surname></persName>
<affiliation>
<orgName type="laboratory">INRIA Sophia Antipolis / Laboratoire I3S</orgName>
<orgName type="team">ARIANA</orgName>
<country>FR</country>
<orgName type="institution">INRIA</orgName>
<orgName type="institution">Université Nice Sophia Antipolis [UNS]</orgName>
<orgName type="institution">CNRS : UMR7271</orgName>
</affiliation>
</author>
<title type="main" level="a">Extended phase field higher-order active contour models for networks</title></analytic><monogr><title level="j" type="main">International Journal of Computer Vision</title><imprint><publisher>Springer</publisher><biblScope type="publicationDate">2010-00-00</biblScope><biblScope type="vol">88</biblScope><biblScope type="issue">1</biblScope><biblScope type="pp">111--128</biblScope></imprint></monogr><idno type="doi">10.1007/s11263-009-0304-3</idno><idno type="HALid">http://hal.inria.fr/inria-00503900</idno><idno type="HALFile">http://hal.inria.fr/inria-00503900/PDF/ijcv.pdf</idno></biblStruct>
</listBibl>
</listBibl>

International Journal of Computer Vision
Publisher	Springer Verlag (Germany)
ISSN	0920-5691 (eISSN : 1573-1405)